<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Liz Dugan on Medium]]></title>
        <description><![CDATA[Stories by Liz Dugan on Medium]]></description>
        <link>https://medium.com/@eadugan28?source=rss-4fa13b03a8f6------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*w0OJnYPobLsNjUukQgmmWQ.jpeg</url>
            <title>Stories by Liz Dugan on Medium</title>
            <link>https://medium.com/@eadugan28?source=rss-4fa13b03a8f6------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 07:20:02 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@eadugan28/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[The quiet parts, out loud]]></title>
            <link>https://medium.com/design-bootcamp/the-quiet-parts-out-loud-b309da8f9189?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/b309da8f9189</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[ux-research]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Mon, 16 Nov 2026 16:26:23 GMT</pubDate>
            <atom:updated>2026-11-16T16:26:23.604Z</atom:updated>
            <content:encoded><![CDATA[<h4>On the inbox after the essay, and why the ‘thank you’s’ arrive in private</h4><p>Every time I publish something about the underside of work (the job search that won’t end, the layoff that arrives on a Tuesday, what months of rejection do to a body), a second conversation starts. It doesn’t happen in the comments. It happens offstage, in my DMs.</p><p>The messages are kind. Most say some version of the same thing: thank you for saying the quiet parts out loud.</p><p>I’m not going to pretend that <strong><em>isn’t</em></strong> flattering. It is. It’s lovely to be told you put words to something someone else had been carrying around wordlessly. It’s a strange and sick comfort to know I am not alone in this– that you aren’t alone in this, either.</p><p>But I was a researcher long before I wrote essays like this, and researchers get a little suspicious when the most important data comes in through a side door. If so many people agree, why does all of the agreement arrive privately?</p><p>I’ve been sitting with that question, and I think the answer is more interesting, and more uncomfortable, than shame.</p><h3>Two feelings in one envelope</h3><p>Each of those DMs carries two things at once.</p><p>The first is a kind of appointment. Without meaning to, the sender makes me akin to a champion: the person who said the thing, who stood up in the theater and spoke while everyone else stayed in their seats.</p><p>The second is the reason they’re in my DMs and not my comments. It’s fear. Not shame exactly, though shame is usually in the room. Shame says <em>something is wrong with me.</em> Fear says <em>something will happen to me if the wrong person finds out.</em> Those are different feelings, and they call for different responses. You can talk someone out of shame. You can’t talk them out of a risk that’s real.</p><p>And this one, unfortunately, <strong><em>is</em></strong> real.</p><h3>The fear isn’t paranoia; it’s good math.</h3><p>I’d love to tell the people in my inbox they’re overthinking it. The literature won’t let me.</p><p>In a large field experiment, economists Kory Kroft, Fabian Lange, and Matthew Notowidigdo sent roughly 12,000 fictitious resumes to real job postings across 100 U.S. cities, randomly varying how long each applicant had been out of work. Callbacks for interviews dropped as the unemployment spell got longer, with most of the drop happening in the first eight months. <strong>Employers seemed to treat time out of work as a signal of something wrong with the worker</strong>, even when nothing else on the resume said so (Kroft et al., 2013).</p><p>That’s the hiring side. The human side is just as well documented. Karsten Paul and Klaus Moser pooled 237 cross-sectional and 87 longitudinal studies and found that unemployment doesn’t merely travel alongside distress; the longitudinal evidence supports the conclusion that it causes it. The effect was also stronger in countries with weaker unemployment protections and more unequal incomes (Paul &amp; Moser, 2009). <strong>I’ll let you decide where the U.S. falls on that spectrum.</strong></p><p>So when someone reads my essay, nods along, and then writes to me privately instead of commenting where a recruiter might see, they aren’t being cowardly. They’re running a cost-benefit analysis, and the numbers support their conclusion. Honestly, as a researcher, I <em>respect</em> the rigor.</p><h3>The hidden cost of staying hidden</h3><p>Here’s the catch: <em>hiding</em> isn’t free either.</p><p>Unemployment is what psychologists call a <em>concealable</em> stigma. It doesn’t show on your face, so you’re always the one deciding who knows. John Pachankis (2007) laid out what that does to a person: the ambiguity of every social situation, plus the ongoing threat of being discovered, produces <em>real</em> cognitive, emotional, and behavioral strain. <strong>You’re never fully in the conversation, because part of you is managing the conversation.</strong></p><p>Diane Quinn and Stephenie Chaudoir (2009) studied people living with a range of concealable stigmatized identities and found that <em>anticipating</em> being devalued predicted greater psychological distress, and was tied to worse self-reported health.</p><p>You don’t have to be <em>judged</em> to pay the toll. You just have to <em>expect</em> it.</p><p>The finding that stopped me cold comes from a German panel study of unemployed adults. Krug et al. (2019) found that <strong>the people most conscious of the stigma of unemployment had lower well-being and worse health</strong>. They didn’t give up, either; they searched <em>harder</em> than other unemployed people. But all that extra effort didn’t improve their odds of getting hired. If you’ve read any of my other articles, you already know why this one crawled into my brain and set up shop alongside my nervous system.</p><p>Read that again, because I think a lot of people in my DMs are living it: working twice as hard to look like you’re fine, and getting nothing back for the overtime.</p><h3>Why a public comment feels like a press conference</h3><p>There’s a reason the reply goes to a private message and not a public comment, and it has a name.</p><p>Alice Marwick and Danah Boyd (2011) described <em>context collapse</em>: social platforms flatten many audiences into one, which breaks the everyday skill of talking differently to different people. Offline, you’d say one thing to your best friend, another to your old manager, and a third to the recruiter you’re hoping will call back. Online, they’re all in the same row, reading the same sentence.</p><p>So a comment under my essay isn’t really a comment. It’s a statement made in front of your former boss, your future boss, your aunt, and a hiring manager you haven’t met yet. A DM rebuilds a room with a door on it.</p><p><strong>I don’t blame anyone for wanting the door.</strong></p><h3>Everyone in the audience thinks they’re the only one crying</h3><p>This is where my inbox gets <em>strange</em>.</p><p>In a classic set of studies, Deborah Prentice and Dale Miller (1993) found that college students consistently believed they were less comfortable with campus drinking than the average student. Almost everyone felt like an outlier. And the students who felt most out of step reported feeling more alienated from campus life, even though their sense of being different was an illusion. Psychologists call this <em>pluralistic ignorance</em>: privately, the majority dissents, but everyone assumes they’re alone, so the public norm stays put.</p><p>I think that’s what’s happening with work, worth, and unemployment. Each person who writes to me believes they’re unusually rattled, unusually tired, and/or unusually afraid. My inbox is evidence that they aren’t. But that evidence lives in my inbox, where no one else can see it.</p><p>That’s the odd position of the champion. I’m the only one who can see the full house, and I’m holding proof that the norm is a bluff, and I can’t show anyone the receipts. They were handed to me in confidence, and <strong>I’m going to keep them that way.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fjYJ5VoVv-RgnX7Gv188ag.png" /><figcaption><em>The room only looks empty from where you’re sitting.</em></figcaption></figure><h3>Who our silence works for</h3><p>This is the part I keep turning over.</p><p>When I publish one of these pieces, I get praise from two directions that never meet. My colleagues and peers cheer, quietly. The other audience, the people with hiring budgets and background checks, say nothing at all. I never learn what they think. I just get to wonder which pile I’ve moved myself into.</p><p>That’s the strange economy of saying the quiet part out loud. The people with seemingly the least power tell you you’re brave. The people with the most power stay silent, and their silence decides whether you were.</p><p>I should know. I’ve spoken up at every job I’ve ever worked. I honestly don’t know if that’s a lack of self-preservation, an internal need to maintain my own integrity, or some weird, uncomfortable balance of the two.</p><p>Organizational researchers named this a long time ago. Elizabeth Morrison and Frances Milliken (2000) argued that powerful forces inside many organizations lead employees to widely withhold what they know about problems. In their account, silence isn’t one person’s private choice. It’s something the system produces.</p><p>It isn’t purely rational, either. Kish-Gephart et al. (2009) describe the fear of challenging people above us as a <em>prepared fear</em>, in the same family as our fear of heights or snakes: quick to learn and stubborn to unlearn. Even when we stop to deliberate, fear tilts the math toward overestimating the risk. Repeated often enough, silence becomes a habit that no longer needs a fresh scare, and it can be mistaken for simple resignation.</p><p>Elisabeth Noelle-Neumann (1974) went further. She described public opinion as what can be voiced in public without fear of sanctions. When people believe they hold the minority view, they tend to go quiet. That makes the “majority” view look bigger than it is, which makes the next person even less likely to speak. She called it the <em>spiral of silence</em>.</p><p>Put those together and you get the loop. Every time one of us keeps it backstage, the official story goes unchallenged for another round: a long gap means something is wrong with you, everyone else is fine, it’s about performance. The people that the story protects never have to defend it. They don’t have to manufacture our dread. They only have to leave it alone, and pick it up when it’s useful.</p><p><em>My</em> silence protects <em>me</em>. <em>Our</em> silence protects <em>them</em>.</p><h3>What saying it out loud actually buys</h3><p>So is disclosure worth it? The honest answer is <em>it depends</em>, and the research is clear on <em>what it depends on</em>.</p><p>Chaudoir and Fisher’s (2010) disclosure processes model treats telling people as a process, not a single brave moment. Whether disclosure helps depends on why you’re doing it and how people respond. When it goes well, the benefits flow through a few routes, including relief from the work of holding it in and access to social support. When it goes badly, you take the hit. That’s why the first few disclosures matter so much, and why choosing your audience carefully is a <em>strategy</em>, not a weakness.</p><p>There’s also an older study I return to over and over again. Spera et al. (1994) worked with 63 recently laid-off professionals. The ones assigned to write about their deepest thoughts and feelings about losing their jobs were reemployed more quickly than those who wrote about neutral topics or didn’t write at all. The writing didn’t seem to make them more motivated; it seemed to change how they felt about the old job and the new search.</p><p>I have to be careful here, though, because that writing was <em>private</em>. The study doesn’t tell us that publishing your feelings with your name attached helps you get hired. Nobody has handed me that guarantee, and I’m not going to pretend otherwise.</p><p>Which means I’m running the same cost-benefit analysis as everyone in my DMs. My name is on every one of these essays. I’m a job seeker too, and the same callback math applies to me. <strong>I write them anyway</strong>, partly because the writing helps me think, and partly because I’d rather say it out loud than keep wondering if I’m the only one. Every time I do, a few more people find out they aren’t, either.</p><p>And I’d be lying if I said that was purely noble. Some days it’s stubbornness. Some days it’s just that I’d rather be seen than managed.</p><h3>To the people in my inbox</h3><p>You don’t owe the public your story. Choosing who gets to know is not cowardice. The research says it’s a reasonable response to a real risk.</p><p>But I want you to know what I know, because you can’t see it from your seat: the room you think is empty is actually quite full. There are so many of you. <strong>There are so many of us.</strong> You’re not the outlier. You’re the quiet majority, each of you sitting very still so no one notices you’re shaking.</p><p>Keep sending the messages. I’ll keep reading them, and I’ll keep them safe.</p><p>If this landed for you, you don’t have to say why. Share it, if that feels safe. Send it privately to one person who needs it. Or just tap the little clap button so the next person can see they’re not the only one. Every quiet signal counts.</p><p>And if you ever decide to say it out loud, you won’t be the only one standing. I’m already up here, and frankly, I could use the company.</p><p><em>Why do so many people thank me privately for writing about unemployment, but never in public? An essay about the DMs that arrive after every piece, the fear behind them, and what the research says about hiding, disclosing, and the quiet majority who think they’re alone.</em></p><p><em>I’m a UX researcher with a master’s in cognitive psychology and over a decade of leading research in complex, high-stakes environments. Lately, my inbox has become its own kind of field study. If you’re hiring, or you’ve been here too, I’d like to hear from you.</em></p><h3>References</h3><p>Chaudoir, S. R., &amp; Fisher, J. D. (2010). The disclosure processes model: Understanding disclosure decision making and postdisclosure outcomes among people living with a concealable stigmatized identity. <em>Psychological Bulletin, 136</em>(2), 236–256. <a href="https://doi.org/10.1037/a0018193">https://doi.org/10.1037/a0018193</a></p><p>Kish-Gephart, J. J., Detert, J. R., Treviño, L. K., &amp; Edmondson, A. C. (2009). Silenced by fear: The nature, sources, and consequences of fear at work. <em>Research in Organizational Behavior, 29</em>, 163–193. <a href="https://doi.org/10.1016/j.riob.2009.07.002">https://doi.org/10.1016/j.riob.2009.07.002</a></p><p>Kroft, K., Lange, F., &amp; Notowidigdo, M. J. (2013). Duration dependence and labor market conditions: Evidence from a field experiment. <em>The Quarterly Journal of Economics, 128</em>(3), 1123–1167. <a href="https://doi.org/10.1093/qje/qjt015">https://doi.org/10.1093/qje/qjt015</a></p><p>Krug, G., Drasch, K., &amp; Jungbauer-Gans, M. (2019). The social stigma of unemployment: Consequences of stigma consciousness on job search attitudes, behaviour and success. <em>Journal for Labour Market Research, 53</em>(1), Article 11. <a href="https://doi.org/10.1186/s12651-019-0261-4">https://doi.org/10.1186/s12651-019-0261-4</a></p><p>Marwick, A. E., &amp; boyd, d. (2011). I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. <em>New Media &amp; Society, 13</em>(1), 114–133. <a href="https://doi.org/10.1177/1461444810365313">https://doi.org/10.1177/1461444810365313</a></p><p>Morrison, E. W., &amp; Milliken, F. J. (2000). Organizational silence: A barrier to change and development in a pluralistic world. <em>Academy of Management Review, 25</em>(4), 706–725. <a href="https://doi.org/10.5465/amr.2000.3707697">https://doi.org/10.5465/amr.2000.3707697</a></p><p>Noelle-Neumann, E. (1974). The spiral of silence: A theory of public opinion. <em>Journal of Communication, 24</em>(2), 43–51. <a href="https://doi.org/10.1111/j.1460-2466.1974.tb00367.x">https://doi.org/10.1111/j.1460-2466.1974.tb00367.x</a></p><p>Pachankis, J. E. (2007). The psychological implications of concealing a stigma: A cognitive-affective-behavioral model. <em>Psychological Bulletin, 133</em>(2), 328–345. <a href="https://doi.org/10.1037/0033-2909.133.2.328">https://doi.org/10.1037/0033-2909.133.2.328</a></p><p>Paul, K. I., &amp; Moser, K. (2009). Unemployment impairs mental health: Meta-analyses. <em>Journal of Vocational Behavior, 74</em>(3), 264–282. <a href="https://doi.org/10.1016/j.jvb.2009.01.001">https://doi.org/10.1016/j.jvb.2009.01.001</a></p><p>Prentice, D. A., &amp; Miller, D. T. (1993). Pluralistic ignorance and alcohol use on campus: Some consequences of misperceiving the social norm. <em>Journal of Personality and Social Psychology, 64</em>(2), 243–256. <a href="https://doi.org/10.1037/0022-3514.64.2.243">https://doi.org/10.1037/0022-3514.64.2.243</a></p><p>Quinn, D. M., &amp; Chaudoir, S. R. (2009). Living with a concealable stigmatized identity: The impact of anticipated stigma, centrality, salience, and cultural stigma on psychological distress and health. <em>Journal of Personality and Social Psychology, 97</em>(4), 634–651. <a href="https://doi.org/10.1037/a0015815">https://doi.org/10.1037/a0015815</a></p><p>Spera, S. P., Buhrfeind, E. D., &amp; Pennebaker, J. W. (1994). Expressive writing and coping with job loss. <em>Academy of Management Journal, 37</em>(3), 722–733. <a href="https://doi.org/10.5465/256708">https://doi.org/10.5465/256708</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b309da8f9189" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/the-quiet-parts-out-loud-b309da8f9189">The quiet parts, out loud</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The gasp]]></title>
            <link>https://medium.com/design-bootcamp/the-gasp-80a22423cdd7?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/80a22423cdd7</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[work]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[ux-research]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Tue, 10 Nov 2026 17:42:40 GMT</pubDate>
            <atom:updated>2026-11-13T17:07:26.617Z</atom:updated>
            <content:encoded><![CDATA[<h3>The Gasp: <em>Update 11.12.26</em></h3><h4>204 days to get in. Less than three months to get out. No severance, no savings, and the question my kid asked me.</h4><p>There’s a reflex your body has when it hits cold water.</p><p>You gasp. You don’t decide to. Your lungs just pull in as much air as they can, all at once, because somewhere deep in the wiring your body has concluded that something has gone very wrong and it might need that air later.</p><p>The last thing I wrote was about an <em>exhale</em>. About 204 days of holding my breath, and the fever that came when I finally let it go.</p><p>This one is about <em>the gasp.</em></p><p>I was fired on a Tuesday. Last Tuesday.</p><h3>The <em>word</em></h3><p>Not laid off. Not restructured. Not “impacted.”</p><p><em>Fired.</em></p><p><strong>I have never been fired. </strong>I’ve been working since I was old enough to work, and in all that time I have never once been told I don’t at least meet expectations. Most of my reviews say something closer to <em>sets a new standard.</em> I’ve been the person brought in to build the thing from scratch.</p><p>The reasons I was given don’t hold up, and I’m not going to litigate them here. I’ll just say this: you can be told something about yourself that doesn’t match a single other piece of evidence you’ve collected over an entire career, and it still becomes true on paper.</p><p>There’s a specific kind of vertigo that comes from being handed a word that has never applied to you. You keep turning it over. You keep checking it against everything you know about yourself. It doesn’t fit.</p><p><strong>It doesn’t matter that it doesn’t fit. It’s yours now.</strong></p><h3>The difference</h3><p>When you get laid off, it happens to a group. There’s a list, and you’re on it, and so is half your team, and so is the person who wrote the onboarding doc everyone still uses. The language is passive on purpose. Roles were <em>eliminated.</em> Teams were <em>affected.</em> Nobody did anything. Something was done to everyone.</p><p>And there’s a script for that. People know their lines. <em>That’s awful. Their loss. Let me know how I can help.</em> A layoff is weather. Nobody asks you what you did to make it rain.</p><p>Getting fired doesn’t come with a script.</p><p>It’s singular. It’s active. Someone looked at you, specifically, and decided. And the word comes with a question attached, one that everyone fills in on their own, whether they say it out loud or not.</p><p><em>What did you do?</em></p><p>My son said it out loud. He’s nine. He doesn’t know yet that grown-ups are only supposed to think it.</p><p>I’m a cognitive psychologist. I know that one of the most reliable findings in social psychology is how quickly we explain someone else’s misfortune by pointing at the person instead of the situation (Ross, 1977). Laid off? The economy. Fired? Must be something about them. I know the bias by name. I know the mechanism.</p><p>Knowing it doesn’t stop me from running it on myself.</p><p>There’s a difference between guilt and shame. Guilt says <em>I did a bad thing.</em> Shame says <em>I am a bad thing</em> (Tangney et al., 2007). Guilt at least hands you something to fix. Shame just sits on your chest and waits.</p><p>I don’t feel guilty. I don’t believe I did something wrong. What I feel is the shame anyway: the borrowed kind, the kind that comes pre-attached to the word and doesn’t wait for you to agree with it.</p><p>A layoff is a story about a company.</p><p>Getting fired is a story about you.</p><p>Even when it isn’t.</p><h3>The dream</h3><p>Here’s the part I’m embarrassed by.</p><p>I let myself believe it.</p><p>After 204 days, after 460 roles and 582 applications and sixty-one percent silence, I let myself believe the hard part was over. I let myself think about the future in units bigger than weeks.</p><p>Hope came back. I didn’t go looking for it. It just walked back into the room and sat down, and I was too tired to ask it to leave.</p><p>I thought I’d woken up from a nightmare.</p><p>I had actually just fallen asleep inside of one.</p><h3>Everything that saved us last time</h3><p>In <a href="https://medium.com/design-bootcamp/the-exhale-4ff6bb562e95?sharedUserId=eadugan28"><em>The Exhale</em></a>, I wrote about privilege. About how we only survived 204 days because of things most people don’t have. I listed them.</p><p>I want to go back through that list now, because I think it’s the clearest way to show you where we are.</p><p><strong>Savings.</strong> Gone. We used all of it getting to August. I told you that last time.</p><p><strong>Credit.</strong> Maxed. Also getting to August. A few months of paychecks doesn’t undo eight months of minimum payments. You don’t climb out of a hole that deep in one season. You just stop digging for a little while.</p><p><strong>Severance.</strong> Last time, I had a small one. It bought us weeks when weeks were the unit of measurement that mattered. I called it a bed sheet you cling to on the way down.</p><p>This time there is no severance. The count is no longer weeks; it’s days.</p><p>There is no bed sheet. There’s just the way down.</p><p><strong>Insurance.</strong> Last time, I ended the piece by telling you I still didn’t have health insurance. And then, for a little while, I did.</p><p>That’s gone too. We’re back to being a family that can’t afford to get sick, living in bodies that spent the last year learning exactly how stress makes you sick.</p><p>That cushion was the whole point of that section. The cushion was the difference between us and the people who don’t make it.</p><p>The cushion is gone.</p><p>We’re the other people now.</p><h3>What did you do wrong?</h3><p>When I told my son I lost my job, he had a small panic attack.</p><p>He’s nine. He was there for the last search. He was there when our dog died in the middle of it. Kids feel the floor move even when nobody says it out loud.</p><p>And when he could talk again, he asked me what I did wrong.</p><p>I need you to understand that it wasn’t an accusation. It was the most reasonable question in the world. He’s nine. In his world, when you get in trouble, it’s because you did something. That’s how the rules work. That’s how they’re <em>supposed</em> to work. Do your work, be kind, try hard, and you don’t get sent to the principal’s office.</p><p>How do you explain to a nine-year-old that sometimes things happen that you don’t agree with, and you had no control over, and it still doesn’t mean you did something wrong?</p><p>I’m still looking for the words. I’m not sure they exist in a size that fits a nine-year-old. I’m not sure they exist in a size that fits me.</p><p>Last time, I wrote that your kid needs to believe you’re standing. This time he watched me get knocked down, and I can’t take that back.</p><p>The best I can do now is let him watch me get up.</p><figure><img alt="An illustration of me at a cluttered desk at dusk, one hand pushed into my messy silver hair, glasses on, floral tattoo sleeves on both arms, staring at a laptop with a worn-out expression. Behind me, a wall calendar has nearly every day crossed out, with one circled in red. An open notebook shows a checklist beside a coffee mug, crumpled paper, and stacks of binders bristling with sticky notes. A desk lamp gives off the only warm light in the room." src="https://cdn-images-1.medium.com/max/1024/1*HSwSEJu_BzWkPdIhtwu31w.png" /><figcaption>‘What did you do wrong, Mama?’</figcaption></figure><h3>Below zero</h3><p>Last time, I said we were rebuilding from below zero. I didn’t know yet that there was a basement under the basement.</p><p>The first search had a runway. Short, and getting shorter, but a runway.</p><p>This one starts at the end of it.</p><p>No savings to burn. No credit to lean on. No severance to stretch. No insurance to fall back on. And a body that never really finished coming down from the first time.</p><p>I wrote about being afraid of walking into a new job pre-burned. I never got the chance to find out if I’d heal. Now I’m walking back into the search already burned, which is a different thing entirely. The exhaustion didn’t leave. It took a short vacation, and now it’s back, and it brought friends.</p><p>And there’s something new in my chest that wasn’t there before, which is the word. My tracker never had a column for <em>fired.</em> Now there’s a question I’ll have to decide how to answer in every recruiter screen, in a market that was already finding reasons not to call me back.</p><p>The tracker is still there.</p><p>Row 461 is waiting.</p><h3>What I need you to know</h3><p>The arc didn’t resolve. I thought it did. I wrote an ending.</p><p>Here’s what I’ve learned since: an ending is only an ending if the next chapter doesn’t start. Mine started on a Tuesday, without warning, and without a soft place to land.</p><p>I don’t know what comes next. I know I’m going to do the thing I always do, which is get back up, open the spreadsheet, and start again. Not because I’m brave. Because there’s nothing else on the menu.</p><p>But I’m tired. <strong>I’m <em>so</em> tired.</strong></p><p>And I want the people reading this, especially the ones doing the hiring, to understand that a lot of us are chasing with less every time we have to start over. Less money. Less sleep. Less of whatever it is that makes you sound bright and easy in a thirty-minute screen.</p><p>We are not less qualified.</p><p>We are just carrying more.</p><p>Last time, I ended by telling you I still didn’t have health insurance, and that I started next week.</p><p>I don’t have health insurance.</p><p>I started over, again, on November 3rd.</p><figure><img alt="Illustration of me and my spouse slumped at a cluttered desk at dusk, both holding our heads as we stare at a laptop. I have short silver hair, glasses, and tattoos. My spouse is bald, with glasses, a red-and-gray beard and a flannel shirt. We are surrounded by mycold coffee mug, crumpled paper, a notebooks of checked-off boxes, stacks of tabbed binders, and a wall calendar full of X’s with one date circled in red." src="https://cdn-images-1.medium.com/max/1024/1*XCUMgS0jL89pET31zQ8NvA.png" /><figcaption><em>Update: Nine days apart. Two adults. Zero jobs.</em></figcaption></figure><h3>Unfortunate update: November 12th</h3><p><strong>Three days</strong> after this essay was posted, my spouse was laid off.</p><p>Three days. THREE days.</p><p>Laid off. Not fired. I spent a whole section up there explaining the difference, so I’ll keep this short: a layoff is weather. This time it rained on the other adult in our house.</p><p>Nine days apart. Two adults. Zero jobs. One panicked nine year old.</p><p>Nobody here is on the clock anymore. If you squint, you could argue that it’s <em>technically</em> retirement.</p><p>So we cashed out mine.</p><p>It wasn’t on the list above. I never thought of it as a cushion. I thought of it as the floor: the one thing you don’t touch, the money that belonged to a later version of me who’d finally get to be tired on purpose.</p><p>Future me just bailed us out. I’d say thank you, but future me is broke now.</p><p>Last week I found the basement under the basement. This week we pulled up the foundation and sold it.</p><p>We’re still getting up. There’s just less underneath us every time.</p><p><em>I’m a UX researcher with a master’s in cognitive psychology and over a decade of leading research in complex, high-stakes environments. I spent 204 days learning what a spreadsheet looks like when you’re the data point, and less than three months finding out it wasn’t finished with me. If you’re hiring, or you’ve been here too, I’d like to hear from you.</em></p><h3>References</h3><p>Ross, L. (1977). The intuitive psychologist and his shortcomings: Distortions in the attribution process. <em>Advances in Experimental Social Psychology</em>, <em>10</em>, 173–220.</p><p>Tangney, J. P., Stuewig, J., &amp; Mashek, D. J. (2007). Moral emotions and moral behavior. <em>Annual Review of Psychology</em>, <em>58</em>, 345–372.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=80a22423cdd7" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/the-gasp-80a22423cdd7">The gasp</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What else does the simple king do?]]></title>
            <link>https://eadugan28.medium.com/what-else-does-the-simple-king-do-56d82c287c66?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/56d82c287c66</guid>
            <category><![CDATA[work]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[literature]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[ux-design]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Thu, 05 Nov 2026 18:06:52 GMT</pubDate>
            <atom:updated>2026-11-05T23:16:08.697Z</atom:updated>
            <content:encoded><![CDATA[<h4>On Camelot, toothless lions, and writing yourself out of rescue.</h4><p>I have loved <em>Camelot</em> for longer than I&#39;ve understood it.</p><p>For most of my life, I loved it the way you love a musical. The pageantry. The yearning. Lancelot singing an entire song about how perfect he is, with a straight face. Then, recently, something else snuck up on me. It wasn&#39;t the affair. It wasn&#39;t the betrayal.</p><p>It was the look on Arthur&#39;s face when he realizes the thing he built is the thing holding him down.</p><h4>A quick refresher, for anyone not raised on cast albums and English tales of olde.</h4><p>Arthur never expected to <em>matter</em>. In T.H. White&#39;s telling, which the musical is based on, he&#39;s a boy called Wart, raised in someone else&#39;s castle. He pulls a sword from a stone and is suddenly in charge of a country that settles every argument with a sword.</p><p>So he spends Act One trying to fix that. First the Round Table: might <em>for</em> right, instead of might <em>is</em> right. Then something bigger and stranger. Law. Courts. Trials where the outcome doesn&#39;t depend on who swings hardest.</p><p>And it works. That&#39;s the part we skip past. It works for the people it was built for: the ones who never had a knight in their corner.</p><p>Then Act Two arrives, the way Act Twos do. Mordred catches Guenevere and Lancelot. Arthur&#39;s own law says the queen must be tried. And burned.</p><p>The most powerful man in England cannot save his wife.</p><p><em>He wrote the rule that says so.</em></p><h4>I&#39;m not the first person to notice this. Not even close.</h4><p>In Thomas Malory&#39;s <em>Le Morte d&#39;Arthur</em>, written around 1470 by a knight who spent much of his life in prison, Arthur condemns Guinevere himself. His reasoning fits in five words: &quot;she shall have the law.&quot;</p><p>The scholar <a href="https://www.cambridge.org/core/books/arthurian-literature/evil-will-and-shameful-death-revisiting-law-in-malorys-morte-darthur/3940DFE9A239FB6097EAC44A264E70B0">Elizabeth Edwards</a> points out that <em>law</em> is a word Malory barely uses. Which is exactly why that line lands like a dropped sword.</p><p>Then Malory gives Arthur a line I have honestly never forgiven him for.</p><p>Grieving the war that follows, Arthur says he&#39;s sorrier to lose his knights than his queen, &quot;for queens I might have [enough].&quot;</p><p>Sir. <em>Sir.</em></p><p>White, writing his version in the shadow of the Second World War, kept the trap and sharpened it. In &quot;The Candle in the Wind,&quot; Mordred and Agravaine don&#39;t beat Arthur with an army. They beat him with <em>his</em> own justice.</p><p>They know he can&#39;t make an exception for the people he loves without turning his law into a joke and himself into a tyrant. So they ask him, very politely, to follow his own rules.</p><h4>And here&#39;s the detail that breaks my heart.</h4><p>White&#39;s Arthur agrees to their plan while hoping out loud that Lancelot will kill his accusers.</p><p>The king is rooting for someone to break the law he can&#39;t.</p><p>Lerner and Loewe put all this on a stage in 1960, and it became Arthur&#39;s Act Two crisis: his wife or his law. (A grim little footnote: a <a href="https://etheses.whiterose.ac.uk/31305">2022 study of the show&#39;s development</a> found that during out-of-town tryouts, <em>Camelot</em> was trimmed to center Arthur and his idealism, at the expense of Guenevere&#39;s character.</p><p>Even offstage, the queen got cut down to make room for the system.)</p><p>So the irony isn&#39;t <em>new</em>. What I keep circling is something smaller and more uncomfortable.</p><p>Arthur doesn&#39;t just <em>grieve</em> Guenevere. He seems to <em>regret</em> the law.</p><p>Not the idea of justice. He never stops believing in that. But there&#39;s a moment where you can watch him realize what he traded to get it.</p><p><strong>Here&#39;s my theory.</strong> Arthur didn&#39;t build his law out of nothing. He built it to <em>replace</em> something.</p><p>The Round Table ran on loyalty. Knights kept faith with each other because they chose to. Edwards draws this line well: chivalry was a code you took on from the inside, while law is imposed from the outside and backed by force.</p><p>She argues that when Malory&#39;s Camelot reaches for law, it has already stepped down from what the Round Table was meant to be.</p><p>That landed <strong>hard</strong> for me.</p><p>Because Arthur couldn&#39;t make people loyal. Loyalty doesn&#39;t <strong><em>scale</em></strong>. You can&#39;t legislate it, and you definitely can&#39;t legislate it into Mordred. So he did the next best thing. He made people <em>accountable</em>.</p><p>And accountability, it turns out, has no exceptions clause for love.</p><h4>This is where I think the musical is wiser than it gets credit for.</h4><p>The law worked for the common folk because it didn&#39;t care who they were. A farmer could get a trial instead of a beating. A nobody could stand in the king&#39;s court and be treated like a somebody.</p><p>But a law that doesn&#39;t care who you are also doesn&#39;t care who you love.</p><p>That&#39;s not a flaw. That&#39;s the whole design.</p><p>And Arthur, the boy who never expected to matter, wrote himself into the category of &quot;anyone.&quot; He&#39;s still the king. He still has the crown, the castle, the roar. But the teeth (the power to look at one person and say <em>this one is mine, and I will protect her</em>) he pulled those himself.</p><p>On purpose. As policy.</p><p>A lion with no teeth isn&#39;t a coward. He&#39;s just a lion who decided biting was beneath the kingdom he wanted to build.</p><h4>So: is Arthur responsible for his own powerlessness?</h4><p>Yes. I think he is.</p><p>But here&#39;s what makes it a tragedy instead of a cautionary tale. I don&#39;t think he&#39;d undo it.</p><p>Offer him the old world, where the strongest arm wins and a king can save anyone he likes, or the new one, where he can&#39;t save the people he loves most. I think he builds the law again.</p><p>He&#39;d just know, this time, what it would cost him.</p><p>There&#39;s a song in Act Two called &quot;What Do the Simple Folk Do?&quot; The king and queen, both miserable, wonder how ordinary people get through sadness. It&#39;s played light.</p><p>It is <em>not</em> light.</p><p>Arthur built a world that protected the simple folk. By the end, he envies them. They get to choose each other over the rules. He doesn&#39;t.</p><p><strong>Here&#39;s why this one gets under my skin</strong>.</p><p>I&#39;m a researcher. For a long time, a big part of my job was being the bridge between the people who use a product and the people who build it. If you wanted to know what users actually needed, you came to me.</p><p>And I wanted that to change.</p><p>Not the bridge. The line in front of it. I wanted every team to reach what we&#39;d learned without <em>waiting</em> on me. So I built AI automations to carry the work: to help facilitate the research, analyze what came back, and put the findings where any function could get to them on their own.</p><p>It worked. That&#39;s the part I keep skipping past, too.</p><p>The research reached more people. The answers stopped depending on who had a meeting on my calendar. By any honest measure, it was more fair.</p><p>And somewhere in there, I essentially wrote myself out of a job.</p><figure><img alt="An illustration of a person with silver hair, glasses, and floral tattoo sleeves slumped across a stone throne in a candlelit hall, hand in their hair, staring wearily at a laptop. A handwritten scroll pours from it down the throne steps. A gold crown dangles forgotten from one hand. Beside them, a stone lion roars with no teeth. Through an arched window, a castle glows at sunset: the kingdom that runs fine on its own now, because the person on the throne wrote it that way." src="https://cdn-images-1.medium.com/max/1024/1*Eiril0rSxszRDuev5jEC7g.png" /><figcaption><em>Heavy is the head that automated the crown.</em></figcaption></figure><p>I&#39;m not telling you that for sympathy. I&#39;d build it again. Which is exactly why I believe Arthur would, too.</p><p>When you build something to serve everyone, you build it so it doesn&#39;t need you.</p><p>That&#39;s the point. That&#39;s the whole design. You don&#39;t get to act surprised when it works.</p><p>Arthur pulled his own teeth on purpose. So did I. I still can&#39;t decide whether that makes us fools, or the only people in the room who meant it.</p><p>The ending gets me every time.</p><p>The night before his last battle, Arthur meets a boy who wants to be a knight. In the musical, Arthur knights him and sends him away from the fighting. Not to fight.</p><p>To remember, and to tell the story.</p><p>In White&#39;s book, that boy is a young Thomas Malory.</p><p>I love that. After everything, what survives isn&#39;t the statute. It isn&#39;t the court, or the trial, or the law that couldn&#39;t bend. <strong>It&#39;s one person who cared enough to carry the story forward.</strong></p><p>Which is to say: loyalty. The thing Arthur thought he had to replace.</p><p>Turns out it was the only thing that lasted.</p><p><em>I’m a UX researcher with over 11 years of experience across agencies, big tech, startups, and enterprise. Lately I’ve been building systems that make research available to everyone, even when that means they need me less. If this resonated, or if you’re a hiring manager who’d rather keep a researcher with all their teeth, I’d love to hear from you.</em></p><p><em>All illustrations are the product of AI. Illustrations containing words are not guaranteed to have accurate spelling or number of limbs.</em></p><h4>Sources</h4><p><strong>Used in the essay (verified to exist; claims checked against publisher abstracts or the text):</strong></p><ul><li>Malory, Thomas. <em>Le Morte d&#39;Arthur</em>. Caxton, 1485. Book XX. Both quotes are public domain. <a href="https://standardebooks.org/ebooks/thomas-malory/le-morte-darthur/text/chapter-20-9">Book XX, ch. 9 (Standard Ebooks)</a></li><li>Edwards, Elizabeth. &quot;Evil Will and Shameful Death: Revisiting Law in Malory&#39;s <em>Morte Darthur</em>.&quot; In <em>Arthurian Literature XXXIX: A Celebration of Elizabeth Archibald</em>, ed. Megan G. Leitch and K.S. Whetter. D.S. Brewer, 2024, pp. 139–155. <a href="https://www.cambridge.org/core/books/arthurian-literature/evil-will-and-shameful-death-revisiting-law-in-malorys-morte-darthur/3940DFE9A239FB6097EAC44A264E70B0">Cambridge Core</a> · <a href="https://boydellandbrewer.com/?p=12395">Contents</a>. I only read the abstract; the full chapter is paywalled.</li><li>Jayasuriya, Kathryn. &quot;&#39;One Brief Shining Moment&#39;: A Study of the Genesis and Development of Lerner and Loewe&#39;s <em>Camelot</em> (1960).&quot; PhD thesis, University of Sheffield, 2022. <a href="https://etheses.whiterose.ac.uk/31305">White Rose eTheses</a></li><li>White, T.H. <em>The Once and Future King</em>. Collins, 1958; Book 4, &quot;The Candle in the Wind.&quot; Plot details (Arthur hoping Lancelot kills the accusers; exceptions making the law a joke) checked against <a href="https://www.sparknotes.com/lit/futureking/section16/">SparkNotes</a> and <a href="https://cliffsnotes-v1.prod.webpr.hmhco.com/literature/o/the-once-and-future-king/overview-the-candle-in-the-wind">CliffsNotes</a> summaries. Worth confirming against your own copy before publishing.</li><li>Brewer, Elisabeth. <em>T.H. White&#39;s The Once and Future King</em>. Arthurian Studies 30. D.S. Brewer, 1993. <a href="https://boydellandbrewer.com/9780859913935/t-h-whites-the-once-and-future-king">Publisher</a>. Supports the wartime context (White seeking an antidote to war).</li><li>Woller, Megan. <em>From Camelot to Spamalot: Musical Retellings of Arthurian Legend on Stage and Screen</em>. Oxford UP, 2021. Ch. 3, &quot;Interpretation and Characterization in Lerner and Loewe&#39;s <em>Camelot</em>.&quot; <a href="https://lib.ecu.edu/catalog-preview/catalog/5686726">Catalog</a></li><li>Coleman, Dwayne C. &quot;Murder, Manslaughter and Reputation: Killing in Malory&#39;s <em>Le Morte Darthur</em>.&quot; In <em>Medieval and Early Modern Murder</em>, Boydell &amp; Brewer, 2018. Notes that Arthur&#39;s knights are rarely tried for murder while Guinevere is. <a href="https://www.cambridge.org/core/books/medieval-and-early-modern-murder/murder-manslaughter-and-reputation-killing-in-malorys-le-morte-darthur/F3135BC44CF6EBCCAE3FAA38C6E17E6B">Cambridge Core</a></li></ul><p><strong>Deliberately not quoted:</strong> Lerner&#39;s lyrics and script. Song titles only.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=56d82c287c66" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Oil paint stays wet.]]></title>
            <link>https://eadugan28.medium.com/oil-paint-stays-wet-596dde340709?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/596dde340709</guid>
            <category><![CDATA[life]]></category>
            <category><![CDATA[disability]]></category>
            <category><![CDATA[health]]></category>
            <category><![CDATA[painting]]></category>
            <category><![CDATA[makers]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Sun, 04 Oct 2026 00:21:01 GMT</pubDate>
            <atom:updated>2026-10-22T16:24:31.830Z</atom:updated>
            <content:encoded><![CDATA[<h4>On acrylic’s deadline, watercolor’s opinions, the pastel that turned all of my clothing into ‘art clothes’, and why the only medium that ever let me be slow is the one that agreed to wait.</h4><p>Oil paint does not dry.</p><p>I mean — eventually. Eventually it does, in roughly the way continents move. But on the timescale of a person standing in front of a canvas trying to work out what they actually meant, oil paint just sits there. <strong>Wet. Patient. Available. For days.</strong></p><p>You can walk away. You can go to bed. You can come back on Thursday having thought about it in the shower, and the painting is still <em>open</em> — still soft, still arguable, still entirely willing to be talked out of whatever it was doing on Tuesday.</p><p>I did not know that was a thing I needed until I had it.</p><h3>The four that didn’t work.</h3><p>Everyone tries the other methods first, because the others are cheaper. And, frankly, none of the others also require you to own solvents — which you then have to explain to your family and friends why you suddenly have so many bottles of highly flamable liquids in your office.</p><p><strong>Acrylic dries in minutes.</strong> <em>Minutes.</em> You lay down a color, you turn to load your brush with the color you were going to blend into it, and by the time you turn back the window has closed. It is a medium with a deadline built into every single stroke, and the deadline is <em>right now.</em></p><p>Acrylic never asks what you meant. Acrylic asks whether you were <strong>fast enough.</strong> And I want to be fair to it — that constraint makes some people brilliant. Some people need a clock. Some people do their finest work at the exact edge of panic.</p><p><strong>I am not one of those people.</strong></p><p>I have an entire nervous system organized around not being those people. Acrylic and I looked at each other and mutually agreed to see other media– though we do hook up on occasion.</p><p>Watercolor is worse, and I say that with love, because watercolor is beautiful and I covet it to an unreasonable degree.</p><p>Watercolor is also where I started. Badly.</p><p>But watercolor isn’t a medium, it’s a negotiation with water — and water does not negotiate. Water has plans. You put down a wash with a clear intention and the pigment does something completely unrelated to your intention. It blooms. It backruns. It creeps into the fibers of the paper and reorganizes itself while you stand there, helpless, holding a brush.</p><p>People call this “<em>letting go.</em>”</p><p>And maybe that’s exactly why I loved it at the time. The whole act was a teenager loosening their grip on things teenagers don’t yet know they’re gripping — and doing it with the total, unearned confidence that only a teenager has. I was not <em>releasing</em> control. <strong>I had no control</strong>. I just happened to have found the one medium where that looked like a technique.</p><p>It also teaches you that an hour of work can be undone by humidity, which is a lesson I had already received, repeatedly, from my own body.</p><p><strong>Pastels get everywhere.</strong> This is not a metaphor and I refuse to make it one. Pastel goes on the paper, and then on your hands, and then on your face, and then on the cabinet you touched on the way to the sink, and then on the dog, and then — through a mechanism I have genuinely never been able to reconstruct — into a room you were not in. Most of my pants have colorful flecks that I hope come across as ‘<em>artsy</em>’ and ‘<em>aloof</em>’ instead of ‘<em>slob</em>’ and ‘<em>dirty</em>.’</p><p>And the smudging isn’t only a housekeeping complaint. Everything you make in pastel is <strong>provisional forever.</strong> It’s held on by friction and optimism. One careless sleeve and a face you spent four hours on becomes a weather system.</p><p><strong>Colored pencil </strong>deserves a brief note for the record, because it looked like the answer and wasn’t. Colored pencils are slow, which I appreciated, and unforgiving, which is the worst possible pairing. Nine hours building up a value, one wrong line, and there’s no recovery — only surrender.</p><p><strong>Slow and permanent is how you manufacture a person who’s afraid to make a mark.</strong></p><p>And yet….</p><p>Soft colored pencils have been quietly working their way back into my life, which I offer as evidence that I am not a person who learns.</p><p><strong>Soft pencils are worse</strong>. More pigment, more saturation, more going down per stroke — which means every wrong line I leave is bigger, brighter, and significantly harder to look away from. I have raised my own stakes for no reason I can defend to you or to myself. The mistakes aren’t just permanent now.</p><p>They’re loud.</p><h3>Fat over lean.</h3><p>Then <strong><em>oil</em></strong>.</p><p>And the first thing oil taught me is a rule that sounds like bureaucracy and turns out to be a philosophy.</p><p><strong>Fat over lean.</strong></p><p>Layers with more oil in them have to go on top of layers with less. Not for aesthetic reasons. For <em>structural</em> ones — and the actual mechanism is better than the version I used to repeat.</p><p>It isn’t really about which layer dries first. It’s about <strong>flexibility.</strong> Lean paint — less oil, more pigment — dries faster and cures <strong>brittle.</strong> Fat paint — more oil — dries slower and stays <strong>flexible</strong>, and keeps moving, subtly, for years. So if you put the brittle layer on top of the flexible one, you’ve laid a rigid skin over something that has not finished moving.</p><p>The skin can’t accommodate the movement. The skin <strong>cracks.</strong></p><p>I need you to appreciate that this is a <em>physical</em> <em>law</em> about the consequences of <em>finishing too early, </em>on top of something that isn’t done moving yet<em>.</em> It’s printed in every technical manual on the subject. It’s just sitting there. Nobody comments on it.</p><h3>The happy accident.</h3><p>The second thing oil taught me is that you can go <em>back.</em></p><p>Not “<em>you can cover it up</em>” — every medium lets you cover things up, at least, if only poorly. I mean you can go back <strong>into</strong> it. A shape that’s wrong isn’t a wrong shape you now have to hide for the rest of the painting’s life. It’s a wrong shape you can soften at the edge, pull half of into its neighbor, lift out with a rag, and reintroduce as something else entirely.</p><p>The error becomes <strong>material.</strong> It’s still under there, doing work — adding a warmth or a weight you would never have chosen on purpose and could not have gotten any other way.</p><p>Bob Ross called those ‘<em>happy accidents</em>,’ and I’ll allow it, though I want to note for the pedants in the audience that Ross painted <strong>wet-on-wet</strong>, at speed, completing entire landscapes inside a half-hour television slot. He is a genuinely awkward patron saint for an essay about slowness.</p><p>He’s the patron saint of <strong>not panicking</strong>, which — fine. That turns out to be the part I actually needed.</p><h3>What oil actually taught me.</h3><p>Here’s the real lesson, and it’s narrower than “<em>slow is better,</em>” because slow is <em>not</em> automatically better. Slow can just be <em>avoidance</em> wearing a nicer outfit.</p><p>The lesson is that <strong>oil separates the making from the judging.</strong></p><p>When a medium dries fast, you’re <em>forced to evaluate the work</em> while you’re still standing inside the feeling that made it. And the feeling that made it is a <strong>liar.</strong> Every <em>single</em> time. In the moment, anything I paint is either the best thing I’ve ever done or unsalvageable garbage, and those two verdicts can arrive forty seconds apart, and <em>neither one of them is information.</em></p><p><strong>Oil makes you wait out the feeling.</strong></p><p>You come back on day three and the thing you were certain about on day one is just — a painting. Sitting there. Some of it works. Some of it doesn’t. And now you can finally see which is which, because <strong>the person doing the looking is no longer the person who was doing the wanting.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YTWpYoMzd3OzvvshYRvO0g.png" /><figcaption>Arms cross is the technique. Day three. The bottom right corner and I are still negotiating. Half of this is finished. The other half is still allowed to change it’s mind.</figcaption></figure><p>There’s research that rhymes with this in a way I find almost annoying. Nelson and Dunlosky found that people’s judgments about how well they’ve learned something are usually poor predictors of what they’ll actually remember — <em>unless</em> the judgment is made after a short delay rather than immediately. Same person, same material, same question. <strong>Only the timing changed</strong> (Nelson &amp; Dunlosky, 1991). And Grossmann and Kross found that people reason more wisely about somebody else’s problem than their own, and that deliberately taking a distanced perspective on your <em>own</em> problem closes the gap entirely (Grossmann &amp; Kross, 2014).</p><p>Neither study is about painting. I want to be honest that I’m borrowing them as analogies rather than evidence. But the shape is the same shape: <strong>you are not a reliable instrument while the event is still happening</strong>, and the only fix anybody has ever found is <em>distance</em>.</p><h3>The body reason.</h3><p>And I’ll be honest about the other reason, which is less noble and considerably more load-bearing.</p><p><strong>I’m chronically ill</strong>. I paint in short stretches, with breaks, on a schedule my body sets and does not consult me about. There are weeks I don’t get to the easel at all. There are times when I leave work untouched for months.</p><p><strong>Every fast medium punishes that.</strong> Acrylic punishes it by closing the window. Watercolor punishes it by drying into a hard edge exactly where I stopped — so the painting keeps a permanent record of the moment my hands gave out. That edge is <em>visible.</em> It’s a scar that says <strong>here is where they had to go lie down.</strong></p><p>Oil doesn’t do that. Oil holds the state.</p><p>I can leave in the middle of a passage — mid-blend, mid-thought, brush down, done for the day — and come back after however long it takes and pick up the same wet edge like nothing happened.</p><p>It is the only medium I have ever used that <strong>doesn’t ask me to have a different body than the one I have.</strong></p><h3>What I’m not saying.</h3><p>I don’t want to turn this into a tidy lesson about how everything good takes time, because that isn’t true and I don’t believe it. Plenty of good things are fast. Some of my best thinking arrives at a genuinely inconvenient speed and I’d be a fool to slow it down.</p><p><strong>The research doesn’t say slow-is-better either, for what it’s worth</strong>. The big meta-analysis on incubation found that stepping away from a problem <em>does</em> improve solutions — but only after you’ve already put in sustained work first, and what you do during the break changes the size of the effect (Sio &amp; Ormerod, 2009). <strong>The break isn’t magic. The break is what you earn by having already tried.</strong></p><p>But the <em>finishing</em> — the deciding, the calling-it-done, the judging of the thing you made — that part I now refuse to do at the speed of the feeling.</p><p>Oil taught me that “done” is not something you <strong>feel.</strong> It’s something you <strong>decide</strong>, later, from a distance, in better light, after the wanting has drained out and you can finally see what you actually built.</p><p>The paint will wait.</p><p>It’s honestly the only thing in my life that does.</p><p><em>I am a non-binary mom, AuDHD, and chronically ill. I write essays about the small, devastating, faintly ridiculous moments that shape how we see ourselves — in the hope someone else finds themselves in them, too. Mostly, though, I’m just here for the snacks.</em></p><h4>References</h4><p>Grossmann, I., &amp; Kross, E. (2014). Exploring Solomon’s paradox: Self-distancing eliminates the self-other asymmetry in wise reasoning about close relationships in younger and older adults. Psychological Science, 25(8), 1571–1580. <a href="https://doi.org/10.1177/0956797614535400">https://doi.org/10.1177/0956797614535400</a></p><p>Nelson, T. O., &amp; Dunlosky, J. (1991). When people’s judgments of learning (JOLs) are extremely accurate at predicting subsequent recall: The “delayed-JOL effect.” Psychological Science, 2(4), 267–271. <a href="https://doi.org/10.1111/j.1467-9280.1991.tb00147.x">https://doi.org/10.1111/j.1467-9280.1991.tb00147.x</a></p><p>Sio, U. N., &amp; Ormerod, T. C. (2009). Does incubation enhance problem solving? A meta-analytic review. Psychological Bulletin, 135(1), 94–120. <a href="https://doi.org/10.1037/a0014212">https://doi.org/10.1037/a0014212</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=596dde340709" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[He waited years to tell me about the cheese.]]></title>
            <link>https://eadugan28.medium.com/he-waited-years-to-tell-me-about-the-cheese-c14d7c75c3f1?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/c14d7c75c3f1</guid>
            <category><![CDATA[habits]]></category>
            <category><![CDATA[neuroscience]]></category>
            <category><![CDATA[relationships]]></category>
            <category><![CDATA[communication]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Wed, 19 Aug 2026 21:03:30 GMT</pubDate>
            <atom:updated>2026-08-19T21:03:30.508Z</atom:updated>
            <content:encoded><![CDATA[<h4>On sharp cheddar, the compounding interest on an unsaid thing, and what gets lost when you keep waiting for a better time to say it.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y_ZiJno4Avvou5ahB7PZPg.png" /><figcaption><em>The four seconds this entire essay is about.</em></figcaption></figure><p>We were standing in the cheese section of a grocery store, which is where I now believe most relationships quietly reveal their structural problems, and I reached for the sharp cheddar.</p><p>I want to describe that reach accurately, because the reach is the whole essay.</p><p>It was not a decision. It was not a moment of consideration between two options on a shelf. It was the physical motion of a person who has performed a task several hundred times and has stopped rendering it consciously — hand out, block off shelf, block into cart, keep moving. <strong>The cart was already moving.</strong> That’s how automatic it was. My body had committed to the next aisle before my hand had finished the transaction.</p><p>And my fiancé said, “Hey — could we get mild instead?”</p><p>I stopped. I turned around. I said, with what I <em>hope</em> was neutrality but was almost certainly not neutrality, “We always get sharp. Sharp tastes better than mild.”</p><p>Which is true. It is still true. Mild cheddar is cheese that comes to a party and then stands at a wall doing nothing but staring– not enough to get asked to leave, but enough that everyone makes mental notes to avoid that person.</p><p>But I heard myself say it, and I heard how it landed, and then I did the one thing I try very hard to always do, which is <strong>ask the question instead of defending the position.</strong></p><p>I said: “Why?”</p><p>And he said: “I don’t like sharp cheddar.”</p><h3>The math.</h3><p>Reader, we had been together for <em>years</em> at this point.</p><p>We were engaged. We were living together — same lease, same kitchen, same refrigerator, same shelf in that refrigerator. We had a shared grocery routine so established it had its own choreography. We had done this specific errand more times than either of us could have counted.</p><p>Every block of cheddar in that entire history was sharp. <em>Every single one</em>. I bought them. He <em>watched</em> me buy them. He carried them to and from car.</p><p><strong>And he ate all of them. </strong>He ate them with a silence that filled the void I didn’t know existed.</p><h3>The thing that actually happened in my chest.</h3><p>I’ve told this story a lot of times, and people always laugh in exactly the same place, which is the reveal. And the reveal <em>is</em> funny. It has the shape of a joke — long setup, short punchline, cheese.</p><p>But what happened in my body, standing in front of a refrigerated case with a cart I’d already started pushing, was not funny even slightly. It was <strong>vertigo.</strong></p><p>Because the immediate question is not ‘<em>what about cheese?’</em> The immediate question is:</p><p><strong><em>What else?</em></strong></p><p>What else has been <em>quietly</em> tolerated. What else did<strong> <em>I</em></strong> establish as a preference of <strong><em>ours</em></strong> when it was only ever a preference of <strong><em>mine</em></strong><em>.</em> How many of the things I think of as <strong>how we do it</strong> are actually <em>how I did it once, early, before anybody had settled in, and then nobody said anything.</em></p><p>I went through our whole apartment in my head, right there, holding a block of cheese. The side of the bed. The thermostat. The restaurant we defaulted to when neither of us wanted to decide. How we spent Sundays. Every one of those had an origin story I could not actually recall agreeing on — because <strong>you don’t agree on those.</strong> You drift into them. One person’s shrug becomes household policy, and the policy runs long enough that it starts to pass for a shared value.</p><p>I still don’t know how many of them were his shrug. That’s the part I never got to the bottom of. He told me about the cheese. The cheese happened to be the one that came up.</p><h3>The compounding interest.</h3><p>Here is what actually gets me, all these years later, and it isn’t dishonesty — nobody lied to me. It’s the <strong>cost accounting.</strong></p><p>If he’d said it in month two — <em>“hey, I’m more of a mild guy”</em> — that conversation is four seconds long. I say “oh, weird, okay,” possibly with a visible amount of judgment, and then we buy a block of medium, or we alternate, or I find out I don’t care nearly as much as I thought I did. It resolves. It costs nothing. <strong>It is nothing.</strong></p><p>But every week he didn’t say it, the price went up.</p><p>At month two it’s a <em>preference.</em> At year two it’s a <em>correction.</em> By the time we were engaged it had become a <strong>revelation</strong> — it’s <em>“actually, this whole time”</em> — and now the conversation isn’t about cheese at all. <strong>Now the conversation is about the fact that he sat on it</strong>. And why he sat on it. And what that implies about the rest of the inventory. And whether I am, apparently, the kind of person you can’t just tell things to.</p><p>That’s the trap, and it’s a beautiful one: <strong>the longer you wait, the more expensive the truth gets, and the expense is precisely the reason you keep waiting.</strong></p><p>It compounds. It’s the only debt I know of where the interest is entirely psychological and entirely self-issued.</p><p>He waited long enough that a four-second correction turned into a moment I am still writing about.</p><h3>The decision he made about me.</h3><p>Now the ungenerous part, which I think is also the true part.</p><p>Somewhere in there, <strong>he made a decision about who I was.</strong></p><p>Not a cruel one–at least at that moment in time. Probably not even a conscious one. But at some point his brain ran the calculation — <em>if I say I don’t like this, it will be a whole thing</em> — and it returned an answer, and the answer was <strong><em>they’re not someone I can just tell.</em></strong></p><p>I never got a vote on that.</p><p>I never got to <em>be</em> the person who says “oh! sure!” and reaches for the mild. That version of me was never allowed onto the field. He decided who I would be in that conversation and then acted on the decision without ever once running the experiment — and I want you to sit with how disorienting it is to discover, years in, that someone has been quietly protecting you from a version of yourself you don’t recognize and never got the chance to defend.</p><p><strong>I would have bought the mild cheese.</strong></p><p>That’s it. That’s the entire tragedy of it. I would have grumbled, and I would have bought the mild cheese, and it would have cost me approximately nothing — and instead he ate cheese he didn’t like for years to avoid a conversation that was never going to be the size he thought it was.</p><h3>The ledger.</h3><p>We did get married, eventually. We aren’t married now.</p><p>That is not because of the cheese. Obviously it is not because of the cheese.</p><p>But the cheese was the <strong>mechanism.</strong> The cheese was a very small, very legible sample of a much larger process in which one of us was quietly absorbing things and the other one had no idea there was anything to absorb. And by the time the absorbing finally became visible, it wasn’t a list of preferences anymore.</p><p>It was a <strong>ledger.</strong></p><p>And you cannot pay down a ledger that was compiled without you. You can’t even see the line items. You just get handed a total.</p><h3>What I ask now.</h3><p>I think about that a lot, in every relationship I have. Not <em>“are you happy”</em> — that question is too big and it invites a reflexive yes from anyone who loves you. The question I actually ask is smaller and stranger and infinitely more useful:</p><p><strong><em>What are you doing my way that you’d rather do differently?</em></strong></p><p>People are terrible at answering it the first time. They say “nothing, I’m easy.” <strong>Ask again in three months.</strong> Ask like you mean it. Ask in a way that makes unmistakably clear that the answer is not a complaint, it’s <em>information</em>, and you would genuinely like to have it now, while it’s still cheap.</p><p>And I still buy sharp.</p><p>I’m not going to pretend I had a conversion in the dairy aisle and became a mild cheddar person. I did not. Mild cheddar tastes like the <em>concept</em> of cheese, described to someone over the phone, by a stranger.</p><p>But I ask now. Every time, with everyone, at a frequency my partners and friends probably find <em>mildly</em> exhausting. What do you actually want? What are we doing because we decided together, and what are we doing because I did it once, years ago, and nobody stopped me?</p><p>Tell me early. Tell me while it’s still a preference and not yet a confession.</p><p><strong>I promise you I am cheaper to talk to than you think.</strong></p><p><em>I am a non-binary mom, AuDHD, and chronically ill. I write essays about the small, devastating, faintly ridiculous moments that shape how we see ourselves — in the hope someone else finds themselves in them, too. Mostly, though, I’m just here for the snacks.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c14d7c75c3f1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Exhale]]></title>
            <link>https://medium.com/design-bootcamp/the-exhale-4ff6bb562e95?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/4ff6bb562e95</guid>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[health]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[mental-health]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Mon, 03 Aug 2026 17:54:49 GMT</pubDate>
            <atom:updated>2026-08-03T22:54:08.887Z</atom:updated>
            <content:encoded><![CDATA[<h4>582 total submissions. 460 jobs. One offer. And a fever the day I signed it.</h4><p>The day I signed my offer letter, my body gave up.</p><p>Not in a dramatic, cinematic way. Not a collapse on the kitchen floor while the credits roll. I just got a fever. A real one — the kind that lasts for days, the kind where your body decides it’s been holding something for you and it’s done now, thanks, you can have it back.</p><p><strong>I had been job hunting for 204 days.</strong></p><p>I felt lighter. I felt heavier. I felt both of those things at the same time, and I don’t know how to describe that to someone who hasn’t been there except to say: imagine holding your breath for eight months and then being told you can exhale, and the exhale itself knocks you down. That’s what it was. The relief didn’t feel like relief. It felt like my nervous system was <em>finally</em> processing eight months of backlog — the cortisol, the adrenaline, the vigilance — and presenting me with the bill.</p><p>I signed the letter and then I couldn’t get out of bed for five days. My body felt like it had been poured out of something. The muscles that had been clenched for months — jaw, shoulders, the ones across your chest that you don’t even know you’re using until they stop — just released. All at once. And the release didn’t feel like peace. It felt cramped and sore. It felt like being t-boned at an intersection. The adrenaline continues to sputter.</p><p>Here is what 204 days of unemployment looks like when you write it down.</p><h3>The Spreadsheet</h3><p>I kept a tracker. Of course I kept a tracker. I’m a researcher — the compulsion to quantify my own misery was practically occupational.</p><p>I applied to 460 unique roles. I submitted 582 applications, because when you’re desperate enough, you apply to the same company multiple times — different roles, different teams, different job boards, different postings of the same position that somehow appeared on three separate platforms with slightly different titles. You hope that maybe this time the system that auto-rejected you on Tuesday will let a human see your name on Thursday. You tell yourself you’re being strategic. <strong>You’re not being strategic</strong>. You’re being thorough in the way that a person is thorough when thoroughness is the only thing left they <em>can</em> control. One company alone got nineteen separate applications from me. Nineteen versions of <em>I am qualified, I am here, please look at me.</em> Nineteen times I tailored a cover letter, adjusted keywords to get past the ATS, and clicked submit. Nineteen times I heard nothing back.</p><p>I applied to 323 different companies. <strong>Three hundred and twenty-three.</strong></p><p>I want that number to land the way it <em>felt</em>. Three hundred and twenty-three different career pages. Three hundred and twenty-three logins created on three hundred and twenty-three applicant tracking systems, each with its own password requirements and its own inexplicable insistence on re-entering information already on my resume. Three hundred and twenty-three times I typed my name into a field like I was introducing myself at a party where nobody wanted to talk to me.</p><p>I got rejected from 155 of them.</p><p>And I want to be clear about something: getting rejected was not the worst outcome. Getting rejected meant that somewhere in the process, a human — or at least a system — <strong>acknowledged that I existed</strong>. A rejection email, even the formulaic kind that starts with “After careful consideration” and you know no consideration occurred, is at least a closed loop. You can <em>grieve</em> a rejection. You can move past it.</p><p><strong>What you can’t move past is nothing.</strong></p><p>And here’s the number that will stay with me longer than any of the others: 281 of those 460 applications — 61% — received no response at all. Not a rejection. Not a form letter. Not even the courtesy of an automated email telling me the position had been filled. I sent my resume, my cover letter, my carefully constructed case studies into a system that couldn’t even be bothered to tell me no. <strong>Sixty-one percent of the time, I was talking to a wall.</strong> I just didn’t know which 61% it was while I was doing it.</p><p>Twenty-four roles generated some form of interview activity (not counting the AI ‘recruiter’ calls, because those deserve their own specialized hell and article). Twenty-four out of 460. A 5.2% hit rate. And “interview activity” is doing a lot of work in that sentence — it includes recruiter screens that got canceled the day of with no explanation, hiring managers who rescheduled three, four times before going silent, processes that “paused” indefinitely with polite reassurances that it was absolutely not a reflection of my candidacy — just internal considerations impacting timing. It includes companies that moved me forward <strong>just far enough to feel hope</strong> — shortlisted, scheduled for a video interview — and then pulled the rug. One company rescheduled my interview repeatedly over the course of a month before I finally withdrew because I’d signed elsewhere. Another ghosted me after a recruiter screen and a hiring manager conversation, after telling me they’d be in touch “very soon.”</p><p>I had a scheduled task — an automated one, running every single morning — that scanned job boards, checked my email for responses, compared new postings against what I’d already applied to, and updated my tracker. Every day I woke up and the first thing waiting for me was a summary of how many new roles existed, how many rejections had come in overnight, and how many of yesterday’s applications had been swallowed by the void.</p><p><strong>I built a system to make my own rejection more efficient.</strong></p><p>I <em>optimized</em> the pipeline of my own despair because that is what this market does to you — it turns you into a machine for processing your own disappointment, and you let it, because the alternative is letting things slip through the cracks, and you can’t afford cracks when you’re 200 applications deep and your checking account is a number you’ve stopped looking at.</p><p>460 applications. 582 submissions. 323 companies. 155 rejections. 24 interviews. <strong>One offer.</strong></p><h4><strong>One.</strong></h4><h3>The Math Behind the Math</h3><p>Those are <em>my</em> numbers. Here are the ones that don’t fit in a spreadsheet.</p><p>My spouse was laid off last October (2025). Five months before my search started. We knew it was coming — you always kind of know — but knowing doesn’t actually prepare you for the morning when the income stops. For a stretch, we had one income — mine — and we adjusted. We budgeted. We were careful. And then January came and we had none.</p><p>Two people in one household of 4, both unemployed. Both filing for unemployment, which, if you’ve never had the experience, involves a weekly ritual of logging in to confirm that yes, you are still unemployed, yes, you are still looking, yes, you applied to the requisite number of jobs, yes, you are still a person who wants to work and has not simply decided to stop.</p><p>The checks were not enough. They are never enough — they’re a fraction of what you earned, with a cap that hasn’t kept pace with the cost of living in any major metro area, and they run out. The gap between the last check and whatever came next was a gap we had to fill with money that was already running low.</p><p>I picked up a small contract — twelve to fourteen weeks of research work that felt like a life raft thrown just close enough to reach. It didn’t pay what a full-time role would have, and it didn’t come with benefits. But it paid, and for those weeks I could focus on doing good work instead of recalculating how many months we had left before the math stopped working. Now, the unemployment office is trying to claim I was ‘self-employed.’ I am fighting them about whether I need to pay them my claims back for that entire time. Those claims, which total over $12,000, is money that I don’t have (because food and shelter) and I don’t owe– I found this out the morning of finishing this article. I am in appeals and it’s still stressful.</p><p>For a few months, I got to be a researcher again instead of a full-time applicant who occasionally did research. And then the contract ended, the way contracts do, and the math started again.</p><p><strong>We used all of our savings.</strong></p><p>I need you to sit with that sentence for a moment, because it goes by fast and it sounds like a thing people say. We used <em>all</em> of our savings. The emergency fund that took years to build. The buffer that was supposed to be there for exactly this scenario — the one we’d contributed to every month, the one that represented discipline and foresight and the belief that if we did things right, we’d be okay. Gone. Not gradually. Gradually is what happens in the first two months, when you’re still optimistic and you’re being careful and you’re telling yourself this is temporary. You cut subscriptions. You stop eating out. You tell yourself you’re being smart.</p><p>What happens after that is faster and <em>uglier</em>. The emergency fund becomes the operating account. You stop tracking individual expenses because the account balance is the only number that matters and it only moves in one direction. You start making decisions you never thought you’d make — which bills can be paid late without consequences, which payments have grace periods, which ones will hit your credit if you miss them. These are not abstractions. These are the actual calculations that occupied my evenings while I was supposed to be resting for another day of applications.</p><p>We virtually maxed out our credit cards.<strong> All of them.</strong></p><p>The ones we’d been disciplined about for years, some with the zero balances that we were proud of. The ones that represented the financial stability we’d built together as a household that knew what it was like to not have money and had worked very hard to be past that — loaded to their limits. We were paying minimums on debt we’d accrued to eat, to keep the lights on, to pay for the things that children need regardless of whether their parents are employed.</p><p>By July, we were getting ready to borrow against our 401ks. September rent was the upcoming wall we were running toward, and the only money left was retirement money. Money you’re not <em>supposed</em> to touch. Money that comes with penalties and tax hits and the implicit admission that the future you were saving for has been mortgaged to survive the present. We were going to take it, <strong>because the alternative was not having a place to live.</strong></p><p>That is a sentence I never thought I would write.</p><p>I have a master’s degree. I have over a decade of experience. And I don’t say that because I think anyone without degrees deserve less. I mean it in the “I have a special set of skills that make me a nightmare for people like…” I have led research programs at companies you’ve heard of. I have shipped work that changed how products were built, that influenced strategy, that mattered. And I was three months away from borrowing my retirement to pay rent.</p><p>My spouse just found work this month. July. A contract role — not permanent, not salaried, no benefits. After nine months of searching. Nine months of the same applications, the same silence, the same nothing. It’s a job he was/is excited for because he is working on a product our family loves, and is within the industries he wants to work for. We are both employed now, technically. We still don’t have health insurance.</p><p>Let me say that again, because it’s easy to miss the momentum of a sentence: <strong>we still don’t have health insurance. </strong>Two employed adults, one child, and no coverage. The offer I signed doesn’t start benefits immediately. The contract my spouse landed doesn’t include them at all. We are working and uninsured, which is a thing that is allowed to happen in this country, and it is happening to us right now.</p><h3>The Thing About Privilege</h3><p>Here’s something I’ve been sitting with, and I don’t think I’ve fully processed it yet: this was the first time in my life I truly understood what people mean when they say <em>privilege</em>.</p><p>Not in the abstract, academic way I’d encountered the word before. In the way where you look at your situation — your terrible, grinding, unsustainable situation — and you realize that the only reason you survived it was because of advantages most people don’t have. That I didn’t grow up having. That I never <em>knew</em> could exist.</p><p>My spouse is better with money than I am. That’s not a character flaw confession — it’s a statement of fact that saved us. Years of their financial discipline is the reason we had savings at all. The reason the emergency fund existed. The reason we lasted as long as we did before the credit cards came out. Without that, we would have hit the wall months earlier. Without that, the 401k conversation would have happened in April, not July.</p><p>I received a small severance package when I was let go. It wasn’t life-changing money. But it bought us a few extra months at a point where weeks were the unit of measurement that mattered. Weeks were the difference between making rent and not making rent. Between keeping the lights on this month and calling the utility company to negotiate a payment plan. A few weeks of severance, in a market like this one, is not a golden parachute. It’s a bed sheet you cling to in the false hope that you can use it to slow your fall.</p><p>Those two things — a partner who budgets well and a modest severance — are the difference between what happened to us and what happens to people who don’t have either. We had a cushion, and we burned through every inch of it, and we still almost didn’t make it. That means someone without those advantages would have been in crisis months before I was.</p><p>That thought hasn’t left me. <strong>I don’t think it should.</strong></p><p>I have always been the person who lands on their feet. <em>Always</em>. I’ve been working since I was old enough to work. When I finished my master’s, I found a role. When I left a job, I either had the next one lined up or I needed a month — maybe two — to find something. Sometimes, the ‘something’ came with a dramatic pay cut, but it also came with really good colleagues, and I made it work. That was the rhythm. I always landed. Landing was the thing I did.</p><p>This was the longest I have ever been unemployed. Longer than the gap after my master’s. Longer than any transition, any pivot, any market downturn I’d weathered before. Eight months. And at no point during those eight months did it feel like I was going to land. It felt like I was going to keep falling.</p><p>Eight months is long enough to change who you are.</p><h3>The House</h3><p>In the middle of all of this, we lost our dog.</p><p>I’m not going to write a lot about it. Not because it doesn’t matter — it matters so much that I can’t hold it and the rest of this at the same time, not yet, not in a piece of writing that’s already asking me to open more doors than I usually leave unlocked. But I need you to know it happened, because it is part of the fabric of this period and leaving it out would be dishonest.</p><p>You are already not sleeping. You are already running on adrenaline and whatever your body manufactures when it’s been in fight-or-flight for months and doesn’t know how to stop. Your mental health is so drained that you already haven’t spoken to some of the most important people in your life for months, maybe even longer. And then your dog dies, and your kid is devastated — the kind of devastated that only a child can be, the total-body kind, the kind that doesn’t know how to be contained because they haven’t learned yet that you’re supposed to contain it — and you have to show up for that. You have to hold space for a small person’s grief while your own grief is sitting right there next to it, asking to be let in, and you can’t, not right now, because you also need to submit more applications before the end of the day because the market does not pause for your loss. The market doesn’t know your dog died. The market doesn’t know your kid cried themselves to sleep. The market is an ATS with a 61% silence rate and it won’t pause for your loss– <strong>it doesn’t care.</strong></p><p>Helping a child through grief when you are also grieving and also terrified about money and also sleep-deprived and also trying to maintain enough composure to sound enthusiastic in a recruiter call the next morning — there is a specific kind of exhaustion that comes from that. I don’t have a word for it. I don’t think one exists. It is not burnout. Burnout is what happens when you’ve been working too hard for too long. <strong>This is what happens when the ground disappears and you have to fall gracefully while convincing everyone around you that you’re still standing</strong>. Your kid <em>needs</em> to believe you’re standing. Your spouse needs to believe it on the days when they can’t stand either. The recruiter on the other end of a call <em>absolutely</em> needs to believe it, because nobody hires the person who sounds like they’re drowning, even when they are.</p><p><strong>The nights were the worst.</strong> Not figuratively. The actual nights — the hours between midnight and five a.m. — were when everything I’d been holding during the day came loose. The anxiety about money. The <em>guilt</em> about not providing. The <em>anger</em> at a market that had apparently decided a decade of experience wasn’t worth a phone screen. The <em>fear</em> that this was it, that I was done, that the industry had moved on and I hadn’t been invited. The restlessness that isn’t quite insomnia but is close enough that the distinction stops mattering at 3 a.m. when you’re staring at the ceiling doing math you’ve already done, running numbers that don’t change no matter how many times you run them. <strong>And when you do sleep it’s nightmares.</strong> Sometimes, those were worth the price of getting any rest, but some were way more than a restless brain and heart could handle.</p><p>Some nights it was the applications that kept me up — replaying cover letters in my head, wondering if the phrasing on a case study had cost me the interview, second-guessing whether I should have applied to that role as an associate instead of waiting for the Staff-level or managerial posting that might never come. Some nights it was the money. Some nights it was the quiet, <em>terrible</em> awareness that my kid was asleep just steps away and <strong>I didn’t know if I could keep this up long enough to make sure they never had to know how close we were to the edge.</strong></p><p>My spouse was going through the same thing. Not beside me — through it, simultaneously, in their own body, with their own version of the same dread. We were taking turns being the strong one. On the good days, one of us could hold it together enough to carry the other through, which usually landed on my spouse. On the bad days, neither of us could, but we had to. Those were the quiet days — the ones where the house was too still and we moved around each other carefully, like everything was fragile, <em>because it was.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xUvAt5wKibhBpTJ_XGFoXQ.png" /><figcaption>The spreadsheet doesn’t have a column for this.</figcaption></figure><p>We didn’t fight. We didn’t have the energy for it. We just existed in the same space, two people trying very hard not to crack in front of each other, and sometimes cracking anyway, and the cracking being okay because at least it was honest.</p><h3>What Your Body Remembers</h3><p>I want to talk about the physical part, because nobody does, and I think it matters more than any of the numbers on my spreadsheet.</p><p>Eight months of job searching is not a professional challenge. <strong>It is a physical event. </strong>It lives in your body the way chronic stress always does — quietly, persistently, in the systems you don’t think about until they start breaking. My body was in a state of constant, low-grade activation for the better part of a year. Not the acute kind, not the kind that spikes and resolves. The kind that <em>hums</em>. The kind that sits underneath everything you do and everything you feel, turning every mundane moment into a threat assessment. This was a new addition to the multitude of other stresses my body was already struggling to suppress or resolve.</p><p>Every email notification was a potential rejection. Every LinkedIn ping was a potential opportunity that would probably go nowhere but that you couldn’t afford to ignore, because what if this one was different, what if this one was the one. Every morning started with a scan of the damage — the tracker update, the inbox check, the recalibration of what I was allowed to hope for based on what had happened overnight, which was almost always nothing.</p><p>You develop a hypervigilance in new ways that you don’t notice until it’s been there for so long that its <em>absence would feel like something is wrong</em>. You check your phone compulsively. You refresh your email. You craft follow-up messages in your head at 2 a.m. and spend twenty minutes deciding whether “just checking in” sounds too desperate. You start reading into the timing of responses — if they haven’t replied in three days, it’s dead; if the recruiter said “early next week” and it’s now Thursday, you compose the follow-up forty times before deciding not to send it. Your nervous system doesn’t know the difference between a tiger and an unanswered email from a hiring manager. It produces the same cortisol either way. And after eight months of cortisol, your body doesn’t know how to make anything else.</p><p><strong>At least the tiger is more efficient.</strong></p><p>The physical toll is cumulative and invisible. You lose weight or gain it. You clench your jaw in your sleep and wake up with headaches. Your shoulders live somewhere near your ears. Your chest is tight in a way that isn’t a medical emergency but feels like one. You get sick more often because your immune system is even more suppressed, and you push through because you don’t have health insurance and you can’t afford to be sick and you also can’t afford to miss a day of applications because what if today was the day the right posting went up and you missed it.</p><p>By the time I signed that offer letter, I had been running on emergency infrastructure for so long that emergency was just how things were. There was no other baseline to return to. My version of ‘normal’ had been gone for months. I didn’t remember what it felt like to check my email without bracing for impact.</p><p>And then the offer came, and I signed it, and my body said: <em>okay, we’re done.</em></p><p>And then that part of my life stopped and I had to brace for impact.</p><p>The fever lasted for days. Not a cold, not an illness I could point to — just heat, and bone-deep exhaustion, and the strange sensation of a system shutting down and rebooting at the same time. I know enough about stress physiology to understand the mechanism: the immune system suppresses prolonged stress and rebounds when the threat passes (Segerstrom &amp; Miller, 2004), and the rebound feels like getting sick because, functionally, it is. Your body, the moment it no longer has to hold the line, processes everything it deferred. Every skipped immune response. Every inflammatory cascade it postponed because there were more important things to fight. It all comes due at once. Researchers call this ‘leisure sickness’ — the documented tendency to fall ill the moment a sustained stressor ends (Vingerhoets et al., 2002).</p><p><strong>Knowing the mechanism didn’t make the fever lower.</strong></p><p>I felt lighter because the weight was finally gone. I felt heavier because I could finally feel how much it had cost me to carry it. Both of those things were true at the same time and I don’t think they’ve resolved. I think I’m going to be carrying this for a while — the residue of it, the way your body remembers a thing even after the thing is over.</p><p>Here’s what I’m afraid of now: walking into a new role already depleted. There’s a concept in burnout research — you can arrive at a job pre-burned (Lim et al., 2016). The exhaustion, the cynicism, the reduced sense of your own competence — the three dimensions of burnout, now validated for job seekers specifically (Lee et al., 2022). You can bring all of that with you from the search itself, and the new role doesn’t heal it. It just gives it a desk. The adrenaline doesn’t know it’s supposed to stop. <strong>The hypervigilance doesn’t know the emergency is over. </strong>You show up on day one with a body that’s been in survival mode for 204 days and you ask it to switch to learning mode, collaboration mode, creative mode, and it looks at you like <em>you’ve lost your mind.</em></p><p>I’ve been trying, in these last few days, to prevent that. To let my body come down. To sleep without calculating. To sit still without checking my email. To remember what it feels like to think about something that isn’t survival.</p><p>It’s not <em>going</em> <em>well</em> yet. But it’s <em>going</em>.</p><h3>What I Need You to Know</h3><p>This is not a success story.</p><p>I mean — it is. I got the offer. I signed the letter. I start soon. On paper, the arc resolves. The spreadsheet has an ending. The numbers stop accumulating.</p><p>But I need you to understand what it cost, because the narrative of “I got through it” flattens everything that happened inside the getting-through. It turns eight+ months of financial terror and sleepless nights and a dead dog and a grieving kid and multiple, maxed credit cards and a retirement account we almost raided and two people taking turns being the strong one and neither of them being strong enough into a redemption arc.</p><p>And it wasn’t a redemption arc.<strong> It was survival.</strong></p><p>Messy, unglamorous, deeply exhausting survival that only worked because of luck and a partner who’s better with money and a short-term contract and unemployment that wasn’t enough and a severance that bought us a few months we wouldn’t have had otherwise.</p><p>If any of those things had been different — if my spouse hadn’t been disciplined with savings, if the severance hadn’t come, if the contract hadn’t materialized when it did — I don’t know where we’d be. That’s not modesty. That’s the math. The margin was<em> that thin</em>. For eight+ months, we walked a line that got narrower every week, and we made it to the other side, and <strong>I am not going to pretend that making it means the line wasn’t there.</strong></p><p>I applied to 460 jobs. I submitted 582 applications. Sixty-one percent of them disappeared into nothing. I got one offer. And the day I accepted it, my body broke.</p><p>My spouse and I are both working now. We are rebuilding from zero — from below zero, actually, from the negative space of maxed credit cards and depleted accounts and the particular kind of financial exhaustion that makes you flinch when your phone buzzes because it might be another bill. We’ll be climbing back for a long time. That’s the part of the story that doesn’t fit in the arc.</p><p>I still don’t have health insurance.</p><p>I start this week.</p><p><em>Liz Dugan is a senior UX researcher with a master’s in cognitive psychology. They have spent the last decade leading research in complex, high-stakes environments and the last eight months learning what a spreadsheet looks like when you’re the data point. They remain both lighter and heavier. Find their other writing on Medium.</em></p><h4>References</h4><p>Lee, T., Ko, H., Lee, S., &amp; Lee, S. M. (2022). Validation of Maslach Burnout Inventory–Job-Seekers Survey. <em>Journal of Employment Counseling</em>, <em>59</em>(1). <a href="https://doi.org/10.1002/joec.12179">https://doi.org/10.1002/joec.12179</a></p><p>Lim, V. K. G., Chen, D., Aw, S. S. Y., &amp; Tan, M. (2016). Unemployed and exhausted? Job-search fatigue and reemployment quality. <em>Journal of Vocational Behavior</em>, <em>92</em>, 68–78.</p><p>Segerstrom, S. C., &amp; Miller, G. E. (2004). Psychological stress and the human immune system: A meta-analytic study of 30 years of inquiry. <em>Psychological Bulletin</em>, <em>130</em>(4), 601–630.</p><p>Vingerhoets, A. J. J. M., Van Huijgevoort, M., &amp; Van Heck, G. L. (2002). Leisure sickness: A pilot study on its prevalence, phenomenology, and background. <em>Psychotherapy and Psychosomatics</em>, <em>71</em>(6), 311–317.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4ff6bb562e95" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/the-exhale-4ff6bb562e95">The Exhale</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Your AI Interviewer Can’t Pass Its Own Test]]></title>
            <link>https://medium.com/design-bootcamp/your-ai-interviewer-cant-pass-its-own-test-944f4972290b?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/944f4972290b</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Mon, 20 Jul 2026 15:36:57 GMT</pubDate>
            <atom:updated>2026-07-20T21:56:44.265Z</atom:updated>
            <content:encoded><![CDATA[<h4><em>We’re using AI to measure people. We haven’t even figured out how to measure AI.</em></h4><p>Here is something that should terrify you.</p><p>Right now, AI systems are screening your job applications. They’re scoring your personality. They’re evaluating your “culture fit,” your cognitive ability, your potential. Companies are using large language models to generate psychometric assessments, administer them, score them, and make hiring decisions based on the results. In clinical settings, LLMs are being tested for psychological assessment, mental health screening, and diagnostic support.</p><p>These are psychometric tasks. They require instruments that are valid, reliable, and fair. That’s not a preference — it’s a scientific and legal standard that has been refined over seven decades of measurement research.</p><p>And the AI systems performing these tasks? We can’t even <strong><em>measure them</em></strong> properly.</p><h3>The Broken Ruler</h3><p>In 1955, Lee Cronbach and Paul Meehl published a paper that changed how science thinks about measurement. “Construct Validity in Psychological Tests” (Cronbach &amp; Meehl, 1955) asked a deceptively simple question: how do you know your test is measuring what you think it’s measuring?</p><p>Not whether it produces numbers. Not whether those numbers change. Whether the thing you’re quantifying — intelligence, anxiety, reasoning ability — is actually captured by the instrument you built to capture it. They called this <em>construct validity</em>, and they wrote about it because psychology had a problem: the field was full of tests that produced scores, and nobody could prove the scores meant what they claimed.</p><p>Seventy years later, AI has the same problem. Except now the broken tests aren’t just <em>mischaracterizing</em> a model’s capability. They’re being used to <em>make decisions about people’s careers, health, and lives.</em></p><h3>How Bad Is It?</h3><p>MMLU is one of the most widely cited benchmarks for large language model capability. It shows up in press releases, investor decks, and technical reports. Here is what we know about it:</p><p>Nearly a third of its test items — 29.1% — have leaked into model training data (Deng et al., 2024). Models score up to 44.5 percentage points higher on contaminated items than on clean ones. An additional 6.5% of its questions contain outright errors in their answer keys (Gema et al., 2024). And frontier models now score above 88%, making the benchmark functionally saturated — score differences at the top are statistically meaningless.</p><p><strong>In any other measurement-dependent field, this instrument would be recalled</strong>.</p><p>A clinical assessment with a 29% contamination rate would be pulled from use. A hiring test with a 6.5% error rate in its answer key would trigger a regulatory investigation. A standardized test where scores are inflated by 44 points due to prior exposure would be a scandal.</p><p>In AI, it’s the industry <em>standard</em>.</p><p>MMLU is not an outlier. A systematic review of 445 LLM benchmarks, presented at NeurIPS 2025, found “pervasive failures in basic construct validity” and “a systemic inability to map test items to well-defined capabilities robustly” (Bean et al., 2025).</p><p>The field is producing hundreds of benchmarks; the benchmarks are <strong><em>not measuring what they claim to measure.</em></strong></p><h3>What Psychometrics Already Solved</h3><p>Here’s what’s maddening for anyone trained in measurement science: these are not <em>new</em> problems. They are <em>solved</em> problems. Psychology encountered every single one of them and built frameworks to address them. Those frameworks are sitting in textbooks, peer-reviewed and replicated, gathering dust while the AI field reinvents them from scratch.</p><p><strong>The contamination problem</strong> is teaching to the test. Educational measurement has spent decades developing countermeasures — item banks, adaptive testing, form equivalence, secure test administration — to prevent exactly this. The AI field is encountering it in 2025 as though <em>nobody has ever seen it before.</em></p><p><strong>The saturation problem</strong> is a ceiling effect. When your instrument can’t differentiate between high performers, the test is too easy. You redesign it.<strong> This is day one of test design.</strong> Humanity’s Last Exam — a benchmark created by nearly 1,000 subject matter experts and published in <em>Nature</em> — exists because the entire prior generation of benchmarks had this problem (Center for AI Safety, Scale AI, &amp; HLE Contributors Consortium, 2026). Frontier models score 35–40% on it. Human experts average 90%. The gap tells you how much the previous benchmarks were <em>overstating</em>.</p><p><strong>The validity problem</strong> was solved in 1955. Cronbach and Meehl defined <em>construct validity</em>. Campbell and Fiske (1959) built the <em>multitrait-multimethod matrix</em> (MTMM) to separate what you’re measuring (the trait) from how you’re measuring it (the method). If two different methods of testing the same ability produce correlated scores, you have<em> convergent validity.</em> If tests of different abilities <em>don’t</em> correlate, you have <em>discriminant validity. </em>Together, they tell you whether your instrument is capturing the real construct or just capturing the test format.</p><p><strong>Almost no AI benchmarks collect this evidence.</strong></p><p>When Timo Freiesleben argued in March 2026 that LLM benchmarks require nomological networks — the framework Cronbach and Meehl introduced seven decades earlier — to establish construct validity, it was treated as a <strong>novel contribution </strong>(Freiesleben, 2026). When researchers proposed applying Campbell and Fiske’s MTMM matrix to LLM evaluation in May 2026, noting that current leaderboards function as “one-dimensional vectors” assessing “a single trait using a single method,” they presented it as an <em>insight</em> (Sakhawat et al., 2026). Stanford launched an entire course on “AI Measurement Science” in 2026 (Truong &amp; Koyejo, 2026).</p><p>The work is good. The concepts are right. <strong>But they already had names</strong>. They already had seventy years of refinement. And the AI field is rebuilding them from first principles because the people building the benchmarks were trained in engineering, not measurement.</p><h3>Now Give That Broken Ruler to HR</h3><p>Here’s where it gets <strong>dangerous</strong>.</p><p>While the AI evaluation field can’t even measure its own models properly, <strong>those same models are being deployed to measure <em>people</em></strong>. AI-powered psychometric assessments are now integrated into recruitment platforms, “delivering insights in minutes” about your cognitive abilities, personality traits, behavioral tendencies, and job-relevant competencies (Premuzic, 2026). Companies are marketing AI that generates, administers, and scores personality tests. LLMs are being used for clinical mental health screening.</p><p><strong>These applications inherit every measurement failure described above — and add new ones.</strong></p><p>LLMs exhibit<em> social desirability bias</em>, skewing personality scores toward favorable traits (Ye et al., 2025). They show <em>demographic biases</em>, generating different empathy levels for different races. <em>Cross-lingual psychometric tests </em>reveal notable inconsistencies — an LLM might assess you differently depending on what language you take the test in (Ye et al., 2025). <strong>When tested with clinical vignettes containing even one incorrect detail, hallucination rates in leading LLMs reach 50–82%</strong> (Qazi et al., 2025)<strong>.</strong></p><p>And the 2026 Stanford study analyzing 4 million real job applications found that when employers share the same algorithmic hiring tool, applicants rejected by one are rejected by all — a phenomenon researchers call <em>“algorithmic monoculture”</em> — <strong>with clear racial disparities in who gets screened out </strong>(Bommasani et al., 2026).</p><p>Think about what this means. We have AI systems that:</p><ul><li>Cannot be reliably measured by their own benchmark</li><li>Have not been validated using the psychometric standards that apply to any human assessment instrument</li><li>Exhibit documented biases across race, language, and demographic categories</li><li>Hallucinate clinical information at rates between 50% and 82%</li></ul><p><strong>And we are handing them a clipboard and telling them to assess people.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*B-oFG5Rc9Adprj3GsTyCrw.png" /><figcaption>Seventy years of measurement science on one side. A conveyor belt of unvalidated assessments on the other. Guess which one is deciding your future.</figcaption></figure><h3>The Circular Problem</h3><p>This is what keeps me up at night: the <em>circularity</em>.</p><p>We need valid benchmarks to know whether AI is capable enough to perform psychometric tasks. But the benchmarks themselves fail basic psychometric standards. So we’re using <em>invalid</em> instruments to <em>validate</em> the deployment of AI in contexts that require <em>valid</em> instruments.</p><p><strong>The ruler we’re using to measure the ruler is also broken.</strong></p><p>NYC Local Law 144 now requires bias audits for automated employment decision tools — but a bias audit performed by an AI system that hasn’t itself been validated for construct validity is an audit in name only. Illinois’s HB 3773, effective January 2026, prohibits employers from using AI in ways that produce discriminatory effects. The <em>legal</em> infrastructure is being built. The <em>measurement</em> infrastructure is not.</p><p>A personality test used in hiring must demonstrate <em>criterion validity</em> (does it predict job performance?), <em>construct validity</em> (does it measure what it claims to measure?), and<em> fairness across protected groups</em>.</p><p><strong>These aren’t suggestions.</strong></p><p>They’re requirements under the Uniform Guidelines on Employee Selection Procedures. They’re enforceable under Title VII.</p><p><strong>No AI-generated psychometric assessment has been validated to these standards.</strong> Not because it can’t be done. Because nobody is <em>requiring</em> it, and the field hasn’t built the frameworks to do it properly.</p><h3>What Needs to Happen</h3><p>Three things, and <strong>none</strong> of them are optional.</p><ol><li><strong>Define the construct before you build the test.</strong> In psychometrics, you don’t build the instrument first and then figure out what it measures. You define the construct — what is “reasoning”? What observable behaviors should it produce? — and then build items that operationalize that definition. Most AI benchmarks work backwards: someone builds a test, models take it, and the field assigns a capability label to whatever the scores seem to reflect. This is how you end up with “reasoning benchmarks” that can be passed through memorization.</li><li><strong>Validate AI assessment tools to the same standards as human assessment tools.</strong> If an AI system is making hiring decisions, it needs to meet the same criterion validity, construct validity, and adverse impact standards as any other selection instrument. <strong>Period</strong>. “It’s AI” is not an exemption from Title VII or the Uniform Guidelines. The measurement bar doesn’t drop because the instrument runs on a GPU.</li><li><strong>Bring measurement scientists to the table.</strong> The AI evaluation community needs psychometricians, I/O psychologists, and educational measurement experts — not as consultants who get called in after the benchmark is built, but as the people who design the measurement framework in the first place. Stanford’s AI Measurement Science course is a start (Truong &amp; Koyejo, 2026). <strong>But a course is not a field. </strong>The field needs to be built, and it needs to be built by people who understand measurement, not people who are learning measurement on the job.</li></ol><h3>The Stakes</h3><p>I want to be direct about what’s at risk here.</p><p>When a benchmark says a model can reason and the model can’t,<strong> that’s embarrassing</strong> for the company. When an AI hiring tool says a candidate lacks leadership potential and the assessment was generated by a model that hallucinates 50–82% of the time,<strong> that’s someone’s career.</strong> When an AI mental health screener misses a crisis because the model exhibits different empathy levels depending on the patient’s race, <strong>that’s someone’s life.</strong></p><p>Psychometrics exists because we decided, as a field, that measuring people requires rigor. That you can’t just build an instrument, run it, and trust the scores.</p><p><strong>That validity is earned, not assumed. </strong>That fairness isn’t a feature — it’s a prerequisite.</p><p>The AI field has <em>skipped</em> all of that. It is building measurement tools for humans using technology it cannot accurately measure itself, deploying them at a scale no psychometrician has ever imagined, and calling it innovation.</p><p>It’s not innovation.<strong> It’s malpractice with a marketing budget.</strong></p><p><em>Liz Dugan is a senior UX researcher with a master’s in cognitive psychology and a decade of experience building measurement frameworks in regulated, high-stakes environments. They have opinions about construct validity and a diminishing amount of patience for systems that measure people without meeting the standards required to measure people. Find their other writing on Medium.</em></p><h4>References</h4><ol><li>Bean, A. M., Kearns, R. O., Romanou, A., Hafner, F., Mayne, H., Batzner, J., Foroutan, N., et al. (2025). Measuring what matters: Construct validity in large language model benchmarks. In <em>Proceedings of the 39th Conference on Neural Information Processing Systems (NeurIPS 2025), Datasets and Benchmarks Track</em>. <a href="https://arxiv.org/abs/2511.04703">https://arxiv.org/abs/2511.04703</a></li><li>Bommasani, R., Bana, S. H., Creel, K. A., Jurafsky, D., &amp; Liang, P. (2026). Algorithmic monocultures in hiring. In <em>Proceedings of the 2026 ACM Conference on Fairness, Accountability, and Transparency (FAccT ‘26)</em>, Montreal, QC, Canada. <a href="https://doi.org/10.1145/3805689.3812400">https://doi.org/10.1145/3805689.3812400</a></li><li>Campbell, D. T., &amp; Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. <em>Psychological Bulletin</em>, <em>56</em>(2), 81–105. <a href="https://doi.org/10.1037/h0046016">https://doi.org/10.1037/h0046016</a></li><li>Center for AI Safety, Scale AI, &amp; HLE Contributors Consortium. (2026). A benchmark of expert-level academic questions to assess AI capabilities. <em>Nature</em>, <em>649</em>, 1139–1146. <a href="https://doi.org/10.1038/s41586-025-09962-4">https://doi.org/10.1038/s41586-025-09962-4</a></li><li>Cronbach, L. J., &amp; Meehl, P. E. (1955). Construct validity in psychological tests. <em>Psychological Bulletin</em>, <em>52</em>(4), 281–302. <a href="https://doi.org/10.1037/h0040957">https://doi.org/10.1037/h0040957</a></li><li>Deng, C., Zhao, Y., Tang, X., Gerstein, M., &amp; Cohan, A. (2024). Investigating data contamination in modern benchmarks for large language models. In <em>Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)</em> (pp. 8706–8719). Association for Computational Linguistics. <a href="https://doi.org/10.18653/v1/2024.naacl-long.482">https://doi.org/10.18653/v1/2024.naacl-long.482</a></li><li>Freiesleben, T. (2026). Establishing construct validity in LLM capability benchmarks requires nomological networks. <em>arXiv preprint arXiv:2603.15121</em>. <a href="https://arxiv.org/abs/2603.15121">https://arxiv.org/abs/2603.15121</a></li><li>Gema, A. P., Leang, J. O. J., Hong, G., Devoto, A., Mancino, A. C. M., Saxena, R., He, X., Zhao, Y., Du, X., Ghasemi Madani, M. R., Barale, C., McHardy, R., Harris, J., Kaddour, J., van Krieken, E., &amp; Minervini, P. (2024). Are we done with MMLU? In <em>Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2024)</em>. <a href="https://arxiv.org/abs/2406.04127">https://arxiv.org/abs/2406.04127</a></li><li>Premuzic, T. (2026, April 23). What’s new in assessment? <em>Forbes</em>. <a href="https://www.forbes.com/sites/tomaspremuzic/2026/04/23/whats-new-in-assessment/">https://www.forbes.com/sites/tomaspremuzic/2026/04/23/whats-new-in-assessment/</a></li><li>Qazi, I. A., Ali, S. A., Zia, H. B., Tahir, H., &amp; Raza, S. (2025). Automation bias in large language model assisted diagnostic reasoning among AI-trained physicians. <em>medRxiv</em>. <a href="https://doi.org/10.1101/2025.08.23.25334280">https://doi.org/10.1101/2025.08.23.25334280</a></li><li>Sakhawat, A., Islam, T., Farhin, T., Raiyan, S. R., Mahmud, H., &amp; Hasan, M. K. (2026). Coordinates of capability: A unified MTMM-geometric framework for LLM evaluation. <em>arXiv preprint arXiv:2605.08522</em>. <a href="https://arxiv.org/abs/2605.08522">https://arxiv.org/abs/2605.08522</a></li><li>Truong, S., &amp; Koyejo, S. (2026). <em>AI Measurement Science</em> [Textbook]. Stanford University. <a href="https://aimslab.stanford.edu/textbook/">https://aimslab.stanford.edu/textbook/</a></li><li>Ye, H., Jin, J., Xie, Y., Zhang, X., &amp; Song, G. (2025). Large language model psychometrics: A systematic review of evaluation, validation, and enhancement. <em>arXiv preprint arXiv:2505.08245</em>. <a href="https://arxiv.org/abs/2505.08245">https://arxiv.org/abs/2505.08245</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=944f4972290b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/your-ai-interviewer-cant-pass-its-own-test-944f4972290b">Your AI Interviewer Can’t Pass Its Own Test</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Portfolio Paradox]]></title>
            <link>https://medium.com/design-bootcamp/the-portfolio-paradox-e456b9a5c9d1?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/e456b9a5c9d1</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[hiring]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Mon, 13 Jul 2026 16:31:01 GMT</pubDate>
            <atom:updated>2026-07-13T22:08:42.019Z</atom:updated>
            <content:encoded><![CDATA[<h4>If you’re going to make me build it, the least you could do is look at it.</h4><p>Let me describe a ritual that every UX researcher knows intimately and nobody talks about honestly.</p><p>You find a job posting. It’s a good one — the kind where the role description actually sounds like a real job and not a wish list written by someone who thinks “UX researcher” means “designer who also does strategy who also does data science who also brings snacks.” You’re qualified. You’re <em>excited</em>. You start the application.</p><p>And then you hit the field: <em>Please provide a link to your online portfolio.</em></p><p>So you do. <strong>You build one</strong>, or you update the one you already have, which — if you’re being honest — hasn’t been touched since the last time you were job hunting, because maintaining a portfolio while actively employed at a job that keeps you busy enough to have portfolio-worthy work is a paradox that nobody in hiring has ever been asked to reconcile.</p><p>You <em>agonize</em> over case studies. You write up projects, carefully navigating the minefield of what you can and can’t share because half your best work is under NDA and the other half requires so much context that explaining it to a stranger feels like narrating a season of television starting at episode six. You pick fonts. You choose a platform. You wonder if Notion is too casual, if a custom site is try-hard, if a PDF is giving 2014. You spend weeks — not days, weeks — building a document that represents the best of what you’ve done, how you think, and why you’re good at this.</p><p>You submit it. You get the interview. You go through the loop — four, five, six conversations with people who are evaluating whether you belong.</p><h4><strong>And not a single one of them asks you about your portfolio.</strong></h4><h3>The Question Nobody’s Asking</h3><p>Here is the thing that has been bothering me for years, and I’ve finally decided to be loud about it: if the portfolio is the price of admission, why doesn’t it come up during the show?</p><p>In most UX research hiring loops I’ve been through — and I’ve been through a lot of them — the portfolio serves two functions. First, it’s a screening tool. The recruiter or hiring manager glances at it (and I do mean <em>glances</em> — we’ll get to that) to decide whether you’re worth a phone call. Second, it gets dropped into a calendar invite or Slack thread for the interview panel to “review before the loop.” And then you sit in five back-to-back conversations where people ask you behavioral questions, hypothetical questions, and the occasional unhinged whiteboard exercise, and your portfolio — the thing you spent weeks building — sits untouched in a browser tab that nobody opened.</p><p><strong>I’m not speculating about this.</strong> I have been in interview loops where I could tell, with the certainty of someone who has spent a decade studying how humans interact with information, that the panelist had not looked at my portfolio. Not skimmed it. Not opened it. The questions they asked had no relationship to the work I’d presented. The follow-ups were generic. The whole thing could have proceeded identically if my portfolio link had been a recipe for banana bread.</p><p><strong>So why did I build it?</strong></p><h3>The Screening Myth</h3><p>The charitable answer is that portfolios help with screening. In a market where a single UX researcher posting can attract 300+ applicants — and that’s not an exaggeration, that’s what hiring managers have been reporting since the market collapsed in 2023 — you need a way to narrow the field. A portfolio gives the recruiter something to evaluate beyond a resume.</p><p>But let’s be honest about what “evaluate” means in this context. The Ladders conducted an eye-tracking study in 2018 and found that recruiters spend an average of 7.4 seconds on an initial resume screen.</p><p><strong>Seven point four seconds.</strong></p><p>That’s not a review. That’s a vibe check at highway speed. And while no equivalent study exists for portfolio review — which is itself telling — I have a hard time believing that the person triaging 300 applications is giving my carefully constructed case studies a substantially different treatment than my resume.</p><p>So the portfolio is being used as a filter, but it’s being filtered at a resolution so low that the effort-to-signal ratio is absurd. I spent weeks building something that gets evaluated in less time it takes to microwave a Hot Pocket. And the candidates who make it past that filter don’t get to engage with their portfolio work in any meaningful way during the actual evaluation.</p><p>This is not a system designed for accuracy. This is a system designed for convenience — the hiring team’s convenience, subsidized by the candidate’s labor.</p><h3>What the Research Says (and Doesn’t Say)</h3><p>Here’s the part where I’d normally cite a stack of peer-reviewed studies about the predictive validity of portfolio review in hiring.</p><p>I <em>can’t</em>, because <strong>they don’t exist.</strong></p><p>There is no published research — <em>none</em> — measuring whether portfolio review predicts job performance for UX researchers, designers, or any adjacent role. Zero studies. Zero validity coefficients. Zero evidence that the thing companies require from every candidate actually tells them anything useful about how that candidate will perform on the job.</p><p>Compare that to structured interviews, which Schmidt and Hunter’s 1998 meta-analysis in <em>Psychological Bulletin</em> validated at r = .51 across 85 years of research — one of the strongest predictors of job performance in the entire I/O psychology literature. Sackett and colleagues revisited those estimates in 2022 with corrected methods and revised the figure to r = .42 — lower, but still among the strongest predictors in personnel selection. Structured interviews work. Structured interviews <em>work*</em>. They predict real performance. We have nearly a century of data saying so.</p><p>Portfolios? We have a tradition. We have an expectation. We have a field that says “this is how it’s done” without ever checking whether “how it’s done” actually does anything. And the entire burden of maintaining that unvalidated tradition falls on the people who can least afford it: the candidates.</p><h3>The Effort Tax</h3><p>Let’s talk about who’s actually paying for this system, because it’s not the companies.</p><p>The UX job market has been <em>brutal</em>. Tech layoffs hit roughly 262,000 workers in 2023, another 153,000 in 2024, and the bleeding hasn’t stopped in 2026. I was hit by both the 2023 and the 2026 layoffs. UX job openings dropped 71% between 2022 and 2023. Fewer than half of designers who lost their jobs in 2024 found new work within three months. The math is not subtle: there are a lot of very qualified people competing for a very small number of roles.</p><p>And every single one of them is being asked to build a portfolio.</p><p>That means people who are already under financial pressure, already managing the emotional toll of a prolonged job search, already applying to dozens of roles that may or may not respond, are also expected to maintain a polished, public-facing document showcasing their best work — work that, in many cases, they can’t fully share because of confidentiality agreements. They’re spending weeks of unpaid labor on an artifact that may get seven seconds of attention from a recruiter and zero seconds of attention from the interview panel.</p><p>This isn’t a minor inconvenience<strong>. This is a systemic labor extraction from the most vulnerable participants in the hiring process. </strong>And the justification — “we need to see your work” — falls apart the moment you realize that the people who could most meaningfully evaluate that work, the interview panelists, aren’t looking at it.</p><p>If the portfolio mattered — really mattered, in the way that a work sample or a structured interview matters — then it would be part of the loop. It would be treated like evidence, not decoration.</p><p>And here’s the thing<strong>: we already have a format that works.</strong> It’s already in the loop. It’s been there for years. Companies just decided it wasn’t enough.</p><h3>The Thing We Already Do</h3><p>UX researchers already present case studies in interview loops. This is not a proposal. This is standard practice.</p><p>You get 30, maybe 45 minutes to walk through one to three case studies of your choosing. You show how you framed the problem, how you designed the research, how you synthesized findings, how you communicated impact. The panel asks questions. They probe the messy parts — the tradeoffs, the failures, the moments where the research said one thing and the stakeholders wanted another. You narrate your thinking in real time, and they evaluate how you think, not just what you’ve polished for public display.</p><p><strong>This works. </strong>It generates the signal that matters. It lets the panel evaluate a candidate’s actual reasoning and communication, not their ability to build a website. It is, by any reasonable standard, the most informative part of the research interview loop.</p><p>And yet — somehow — companies have decided this isn’t sufficient. The new trend is to layer a portfolio requirement <em>on top of</em> the case study presentation.</p><p>Not <em>instead</em> of it. On <em>top</em> of it.</p><p>So now you’re building a portfolio that nobody looks at, <em>and</em> you’re still presenting case studies in the loop, which is where the real evaluation happens. The portfolio isn’t replacing anything. <em>It’s just additional unpaid labor that sits between you and the interview you were going to do anyway.</em></p><p>I’m not saying portfolios are inherently useless. For visual designers, where the medium is the message, a portfolio makes sense. The work is visual. The evaluation is visual. The format fits the function. But for UX researchers, whose work lives in synthesis, in frameworks, in the messy translation between data and decisions — <strong>a static portfolio is the wrong container for the right content.</strong> It’s like evaluating a musician by looking at their sheet music, never asking them to play, and then asking them to play anyway because that’s how it’s always been done. Except now you’re also making them print the sheet music on nice paper first.</p><h3>What I’d Actually Like to See</h3><p>I don’t think portfolios should be abolished. I think if they’re going to exist, they should do something.</p><p>Here’s a proposal that’s so simple it almost feels like a trick:<strong> let the interview panel pick.</strong></p><p>The candidate submits a portfolio. The panel actually reviews it — and not in a “glance at it in the elevator on the way to the interview room” way, but substantively. Then the panel selects the top three case studies they’re most interested in hearing about, and tells the candidate which ones they’d like them to present.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zxEdkUYjABLZa3VEoEMqQQ.png" /><figcaption>The portfolio becomes a menu: you write it, they order from it.</figcaption></figure><p>That’s it. That’s the whole idea.</p><p><strong>I’ve done this exactly once</strong>. And just recently. I was in a loop where I asked the panel to select which case studies they wanted me to present rather than choosing my own. The conversation that followed was different — more focused, more engaged, more like a real exchange between researchers than a performance. I’m currently in the final stages of that process. One data point isn’t a study, but it’s more evidence than the traditional portfolio model has ever produced.</p><p>But think about <em>what it does.</em></p><p>The panel gets to hear about the work that’s most relevant to their team, their problems, their open questions. No more sitting through a case study about e-commerce checkout optimization when what you really wanted to know about was the candidate’s regulatory compliance research. You pick what matters to you.</p><p>And the candidate? The candidate <em>gets signal. </em>They get to see what the panel cares about before they walk into the room. If the panel picks the study about longitudinal behavior change over the study about rapid usability testing, that tells the candidate something about the team’s priorities, their research maturity, what kind of work they’ll be doing if they get the job.<strong> That’s information that benefits both sides. </strong>The candidate can tailor their narrative to what the panel actually wants to hear, which means the panel gets a better presentation, which means the evaluation is more useful.<strong> Everyone wins.</strong></p><p>This also solves the engagement problem. The portfolio isn’t a decoration anymore. It’s not a toll booth. It becomes the thing that shapes the conversation — a menu that the candidate writes and the panel orders from. It has a reason to exist beyond “we’ve always asked for one.” It becomes a genuine artifact of collaboration between the candidate and the hiring team, before a single interview has started.</p><p>If you’re not willing to do this — if the portfolio is just a screening artifact that gets skimmed in the first five minutes of the hiring process and then never referenced again — then stop requiring it. Let candidates submit a resume and a cover letter. Let them describe their work in an initial conversation with the recruiter. Stop outsourcing your screening labor to people who are already stretched thin.</p><p>But if you’re going to make me build it, don’t just glance at it and then ask me to present whatever I want anyway. <em>Look</em> at it. Pick the parts that interest you. Tell me what you want to hear about. Let me show up prepared to have the conversation you actually want to have.</p><p><strong>That’s not a radical idea. That’s just good research practice. And if a UX research team can’t see the value in letting the data shape the conversation — well, that might tell you something about the team.</strong></p><p><em>Liz Dugan is a tenured, UX researcher with a master’s in cognitive psychology and a decade of experience leading research in high-stakes, regulated environments. They have built, rebuilt, and agonized over more portfolios than they care to count, and have never once had an interviewer reference something specific from their portfolio before asking them to present. They remain irritated about this. Find their other writing on Medium.</em></p><h4><strong>Sources:</strong></h4><ol><li><strong>Schmidt &amp; Hunter (1998)<br></strong> Schmidt, F. L., &amp; Hunter, J. E. (1998). The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings. <em>Psychological Bulletin</em>, 124(2), 262–274.</li><li><strong>Sackett et al. (2022)<br></strong> Sackett, P. R., Zhang, C., Berry, C. M., &amp; Lievens, F. (2022). Revisiting meta-analytic estimates of validity in personnel selection: Addressing systematic overcorrection for restriction of range. <em>Journal of Applied Psychology</em>, 107(11), 2040–2068.</li><li><strong>TheLadders (2018)<br></strong> Ladders, Inc. (2018). <em>Eye-Tracking Study: 7.4 Seconds</em>. TheLadders.com. (Industry white paper, not peer-reviewed.)</li><li><strong>Layoff numbers (262,000 in 2023; 153,000 in 2024)<br></strong> Layoffs.fyi aggregate tracker. 262,735 in 2023; 152,104 in 2024.</li><li><strong>UX job openings dropped 71%<br></strong> Brookshier, K. (2023). UX Design and UX Research Job Listings Plunged in 2023. <em>Indeed Design</em>.</li><li><strong>Fewer than half found work within three months<br></strong> Abad, C. (2024). State of the design job market 2024. <em>ADPList Substack</em>. (Analysis of 640,000+ designer job changes; 49.5% landed within three months in 2024, down from 67.9% in 2019.)</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e456b9a5c9d1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/the-portfolio-paradox-e456b9a5c9d1">The Portfolio Paradox</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I’m not mad; I’m autistic]]></title>
            <link>https://medium.com/design-bootcamp/im-not-mad-i-m-autistic-ec68fc3856e5?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/ec68fc3856e5</guid>
            <category><![CDATA[work]]></category>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[autism]]></category>
            <category><![CDATA[ux]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Mon, 29 Jun 2026 15:24:32 GMT</pubDate>
            <atom:updated>2026-07-08T21:44:26.506Z</atom:updated>
            <content:encoded><![CDATA[<p><em>My tone for anger, frustration, concentration, and directness is the same. That’s not a bug. But it sure gets treated like one.</em></p><p>Here is a thing that happens to me — not every day, not in every meeting, but with<em> enough regularity</em> that I’ve stopped being surprised by it and started writing about it instead:</p><p>I’m in a meeting. I’m focused. I’m engaged. I’m thinking carefully about the problem and saying exactly what I think about it — clearly, directly, without padding it with softeners or the vocal equivalent of a smiley-face sticker. I’m not upset. I’m not frustrated. I’m doing the thing that every corporate values poster says they want: <em>being honest and direct.</em></p><p>And then someone asks me if I’m okay.</p><p>Or worse — and this is the <em>fun</em> version — someone doesn’t ask. They just <em>decide</em>. They pull a colleague aside afterward and say something like, “<em>Liz seemed really angry in that meeting.</em>” Or they adjust their behavior around me — suddenly careful, suddenly diplomatic in a way that wasn’t there five minutes ago. The temperature in the room shifted, and I’m the last person to know why. Which, as a UX researcher, is an irony I appreciate more than I enjoy.</p><p>Here’s the thing, though: most of the time, this doesn’t happen. Most of the time, my colleagues would tell you I’m a great teammate. Easygoing. Fun to collaborate with. Clear communicator. The kind of researcher clients genuinely enjoy working with. I know this because they’ve said it — in performance reviews, in recommendations, in the kind of unsolicited feedback that makes you think maybe you’re pulling it off.</p><p><strong>And I am pulling it off.</strong> That’s the problem. What they’re experiencing isn’t my natural communication style. It’s a performance I’ve been building, refining, and running since long before I knew why I needed it. It’s a mask so well-constructed that most people don’t know it’s there. And the moments when it cracks — when I’m too burned out to find it, too depleted to run it, too deep in a high-stakes moment to afford the cognitive overhead — those are the moments that generate the stories. The “<em>Liz seemed angry</em>” stories. And because the mask works so well the other 90% of the time, those moments land harder. They become the anomaly that people remember, not the pattern they take for granted.</p><h4><strong>I’m autistic.</strong></h4><p>And this is one of the most <em>exhausting</em> parts of it — not the sensory stuff, not the executive function stuff, not trying to figure out if that person was joking or if I need to update my resume. The exhausting part is that the thing I do best — making other people comfortable — <strong>is the thing that costs me the most</strong>. And nobody sees the invoice.</p><h3>The Tone Problem Is a Translation Problem</h3><p>There’s a concept in autism research called the “double empathy problem,” coined by Damian Milton in a 2012 paper in <em>Disability &amp; Society</em>. The traditional framing of autism — the one most people absorb by cultural osmosis — says autistic people have a social communication deficit. We struggle to read neurotypical social cues, and that’s our problem to solve. Our homework. Our deficit. How <em>generous</em> of everyone to name it for us.</p><p>Milton flipped that: <strong>the communication breakdown between autistic and neurotypical people goes both ways</strong>. It’s not that autistic people can’t communicate. It’s that two different communication systems are trying to interface, and neither one is inherently broken. They’re just <em>different</em>. Think Bluetooth and Wi-Fi. Both work fine. They just don’t talk to each other, and nobody blames the Bluetooth.</p><p>Crompton, Ropar, Evans-Williams, Flynn, and Fletcher-Watson tested this empirically in 2020 (<em>Autism</em>). They used diffusion chains — basically a scientific telephone game — and found that autistic-to-autistic information transfer was just as effective as neurotypical-to-neurotypical. Same accuracy. Same rapport. But in mixed-neurotype chains, information degraded and rapport dropped. <strong>The communication gap isn’t in either group; it’s in the space between them. </strong>The gap has two parents, and only one of them is being sent to therapy.</p><p>This is exactly what happens with my tone — when people actually hear it.</p><p>Among most, other autistic people, my directness reads as… directness. Nobody flinches. Nobody checks on me afterward. Nobody schedules a follow-up meeting to discuss my affect. The signal arrives intact. But when that same signal crosses into a neurotypical context without the mask buffering it, it gets mistranslated. Direct becomes blunt. Focus becomes tense. Neutral becomes cold. And cold, for an AFAB person, becomes <em>angry</em>.</p><figure><img alt="A tattooed figure in glasses stands split down the middle — same face, same expression, two worlds. On the left, red chaos: jagged edges, exclamation marks, scribbled thought bubbles, silhouettes already whispering. A room that’s decided they’re angry. On the right, blue calm: flowing lines, an observing eye, quiet geometric order, the same silhouettes at ease. Concentric rings pulse outward on both sides — same signal, different receivers. Neither side changed. Only the interpretation did." src="https://cdn-images-1.medium.com/max/1024/1*g1fJZU6nFVWRYZr7izqyJA.jpeg" /><figcaption>Same face, two readings. The left is what the room decides I’m feeling. The right is what’s actually happening. The expression never changed — just the interpreter.</figcaption></figure><h3>My Face and My Feelings Are Not on the Same Channel</h3><p>Part of what makes this so persistent — and part of why the mask exists in the first place — is that my external expression doesn’t reliably map onto my internal state. <strong>My face and my feelings are on two different Wi-Fi networks</strong>, and they have never successfully synced.</p><p>This isn’t <em>unusual</em> for autistic people. Roughly half of autistic adults experience alexithymia — difficulty identifying and describing their own emotions. (If you’ve ever been asked “<em>how does that make you feel</em>?” and your honest answer is “I’ll get back to you in four to six business days,” congratulations, you might be in this club.)</p><p>Cook, Brewer, Shah, and Bird published a pivotal study in <em>Psychological Science</em> (2013) showing that the emotion-recognition difficulties previously attributed to autism were actually <em>predicted</em> by alexithymia, not autism severity.</p><p><strong>Read that again: </strong>it’s not that autistic people can’t recognize emotions.</p><p>It’s that some of us process our own emotional signals <em>differently</em>. Bird and Cook followed up the same year in <em>Translational Psychiatry</em>, proposing that the mismatch between what autistic people feel and what their faces and voices express is driven by alexithymia. A person can feel deeply engaged, deeply calm, or deeply happy while displaying an expression that others read as blank, flat, or hostile.</p><p><strong>This lands differently depending on what the observer expects to see. </strong>And what they expect to see is shaped by a lifetime of reading neurotypical faces and tones. When someone’s expression doesn’t match the neurotypical template for “interested” or “friendly” or “fine,” the brain doesn’t default to “maybe they express it differently.” It defaults to “<em>something is wrong with this person.</em>”</p><p>That’s not <em>malice</em> on the observer’s part. But <em>it’s not neutral</em>, either. It’s a bias that runs so deep most people don’t even know it’s running.</p><p><strong>And it’s the reason the mask has to exist at all.</strong></p><h3>The Thin-Slice Problem (or: You Had Me at Hello, and You Were Wrong)</h3><p>Here’s where the science gets <em>uncomfortable</em> — or, if you’re autistic and reading this, where the science gets <em>vindicating</em>.</p><p>Sasson, Faso, Nugent, Lovell, Kennedy, and Grossman published a study in <em>Scientific Reports</em> (2017) that I think about more often than is probably healthy. They showed neurotypical observers brief video clips of autistic and non-autistic adults — thin slices, just a few seconds — and asked for first impressions. <strong>Neurotypical observers rated autistic people lower on likability, warmth, and desire to interact.</strong> After a few seconds. Before the autistic people had done anything wrong, including, critically, being wrong.</p><p><strong>But here’s the part that should be projected onto the wall of every HR department on the planet: </strong>when observers read text-only transcripts of the same people, those biases disappeared. The <em>content</em> of what autistic people said was fine. The judgments were based entirely on delivery — prosody, affect, body language. <em>Style</em>, not <em>substance</em>.</p><p>Think about what that means for someone like me — someone who has learned to mask well enough that colleagues call them “easygoing” and “fun to collaborate with.” Every interaction is a live performance where I’m translating my natural delivery into something that won’t trigger those thin-slice judgments.</p><p><strong>And it works.</strong> Most of the time, it works <em>beautifully</em>.</p><p>But it works because I’m spending cognitive resources on the translation that I could be spending on the actual work. And the moment the resources run out, the unmasked delivery comes through — and the thin-slice machinery kicks in. Not because I changed. Because the performance <strong>stopped</strong>.</p><h3>Now Add Gender, and Watch It Get Worse</h3><p>Everything I’ve described so far applies to autistic people broadly. But for autistic AFAB people — people assigned female at birth — there’s an additional layer that makes the whole thing <em>significantly, and almost impressively, harder</em>.</p><p>Gendered social expectations demand that AFAB people perform warmth. Smile. Soften. Hedge. Say “<em>I think maybe</em>” when you mean “<em>this is wrong.</em>” Laugh at the thing that isn’t funny. Apologize before you disagree. The emotional labor of femininity isn’t optional — it’s the price of admission to being perceived as a normal, likable, non-threatening person. And the toll booth <em>never</em> closes.</p><p>Autistic AFAB people often can’t — or won’t — pay that price. Not because we’re rude. Because our brains are wired for direct, literal communication. We say what we mean. We (<em>mostly</em>) skip the softeners. We don’t perform the warmth preamble before the honest content. And in a world that expects AFAB people to be warm first and competent second, directness doesn’t read as confidence. It reads as <em>aggression</em>.</p><p>Which is a <em>wild</em> conclusion to draw about someone who is <em>literally</em> just answering the question they were asked.</p><p>Bargiela, Steward, and Mandy (2016, <em>Journal of Autism and Developmental Disorders</em>) studied women diagnosed with autism late in life and documented how the “female autism phenotype” — including learned social mimicry — led professionals to miss their autism entirely. These women had spent years performing normalcy, exhausting themselves to meet gendered social expectations that were never designed for their neurology. <strong>Their autism wasn’t absent. It was invisible</strong>, because the diagnostic criteria were built around how autistic boys present.</p><p><strong>The rubric was never designed to see us, and then everyone acts surprised when it doesn’t.</strong></p><p>Lai, Lombardo, Auyeung, Chakrabarti, and Baron-Cohen (2015, <em>Journal of the American Academy of Child and Adolescent Psychiatry</em>) established that autistic girls tend to internalize their differences while boys externalize them. Boys’ autism is visible — it disrupts the classroom, it triggers referrals. Girls’ autism is <em>quiet</em> — it disrupts the girl, and nobody notices until the cost becomes unsustainable. By which point, of course, the damage has been filed under “personality.”</p><p>Hull and colleagues (2017, <em>Journal of Autism and Developmental Disorders</em>) found that autistic women reported higher rates of masking and social camouflaging than men, and that the long-term consequences included exhaustion, identity confusion, and mental health deterioration. This is the double bind, and it’s airtight:<strong> mask and lose yourself, or don’t mask and be read as aggressive, cold, or “too much.”</strong></p><p>Pick your damage.</p><h3>What It’s Like From the Inside</h3><p>I want to be clear about something: most days, the mask holds. Most days, I am the colleague people describe in their LinkedIn recommendations — collaborative, organized, someone who brings good energy and makes complex work feel manageable. <em>That’s not fake</em>. Those qualities are <em>real</em>. But the delivery system is engineered, and the engineering has a power supply.</p><p>Here’s what people don’t see: I’m a UX researcher, and the most demanding masking I do is for my participants. When I’m running a research session, I’m building rapport, reading the room, creating the psychological safety that makes people honest enough to provide the most useful data they can — while walking a tightrope over participant straining or stressing. That takes <em>even more effort </em>than a typical masking act. That performance is non-negotiable — the <em>quality</em> of the research depends on it.</p><p><strong>My participants get my best mask, every time,</strong> because their comfort is what produces the data my team needs to make good decisions. Which means I have a budget, and my participants get first claim. By the time I’m in the debrief, the team sync, the stakeholder readout — the tank is lower. My colleagues get a version of me that’s running on whatever’s left. Not because I don’t care about them, but because<strong> I already spent that energy where my work demanded it most.</strong></p><p>It’s my users or it’s my colleagues. I choose my users every time.</p><p>And most of the time, there’s <em>enough</em> left. Most of the time, the residual is enough to keep the performance running at a level where nobody notices the seams. <strong>But masking is cumulative</strong>. It doesn’t reset overnight the way people who don’t do it assume it does. A hard week of research sessions, back-to-back stakeholder meetings, a few too many days of performing approachability with nothing replenishing the supply — and the balance <em>tips</em>.</p><p><strong>When I hit a certain level of burnout, I can’t even find my mask.</strong></p><p>It’s not a choice to drop it. It’s not defiance or carelessness. The mask is simply <em>not</em> accessible. The cognitive resources it requires to locate, activate, and maintain are the same resources that burnout depletes first. And what’s left underneath — the unmasked version — <em>is still me</em>. Still competent, still engaged, still saying things that are accurate and useful. But without the warmth preamble. Without the softeners. Without the melody that tells neurotypical listeners “<em>this person is safe</em>.”</p><p>That’s when I’ll be in a meeting, fully engaged, saying exactly what I think with exactly the precision I think it deserves, and afterward I’ll learn — sometimes directly, sometimes through the grapevine, sometimes through the extremely fun experience of hearing it years later — that someone experienced me as angry. Not disagreeing.</p><p>Not direct. <strong>Angry</strong>.</p><p>And the thing that makes it so disorienting is that <em>I can’t tell when it’s happening</em>. I don’t experience the shift. From the inside, I’m the same person I was last week, when everyone was fine with me. The room changed around me, and I only find out later, when the damage had already been done. It’s like finding out you’ve been on mute for the whole call, except the opposite — I was on broadcast, and apparently the frequency sounded hostile.</p><p>And when I do find out — sometimes months later, sometimes years — I <em>carry</em> it. I ruminate. I go back and apologize to people who have long since moved on from a moment I am still replaying. Not because I think I did something wrong, but because the idea that someone experienced me as harmful, when I was just <em>being</em>, is the kind of thing <strong>my brain will not let go of</strong>. The cruelest part of the mismatch isn’t the misread itself.</p><p><strong>It’s that they forgot about it and I never will.</strong></p><p>I officiated roller derby — and I was good at it. Clear calls, consistent performance, direct feedback to my crew during and after bouts. The kind of feedback designed to make everyone sharper. The kind of feedback that, in any other context, or even gender, would be called “leadership.” Near the end, before my medically-enforced departure, I was told I wouldn’t advance to certain higher levels of officiating leadership. The reason? My feedback was “severe.”</p><p>Not inaccurate. Not unhelpful. Not wrong. <strong>Severe</strong>.</p><p>I wasn’t yelling. I wasn’t demeaning anyone. I was giving precise, direct, unpadded feedback — the same way I do everything when the mask isn’t running — and because my tone didn’t come gift-wrapped in enough warmth for the people evaluating me, the content got reclassified. What I said didn’t matter. <strong>How it sounded, did.</strong> And “severe” is just the roller derby world’s version of the same word I’ve heard in every other context: <em>too much.</em></p><p>And then there are the moments where even the possibility of choosing disappears.</p><p>When something goes sideways in a tight window — a session is about to derail, a deadline compresses, a decision needs to happen in the next thirty seconds — the mask is the first thing that gets cut. Not strategically. Automatically.</p><p><strong>Masking takes processing time.</strong></p><p>The cognitive overhead of padding, softening, performing approachability — that’s bandwidth, and when the bandwidth gets allocated to actually solving the problem, the performance drops. What comes out is pure signal: clear, direct, efficient, and completely undecorated. Which is exactly when people are most primed to read directness as aggression — because everyone’s stressed, and the person who isn’t performing stress-appropriate warmth becomes the problem in the room. <strong>The irony is that those unmasked moments are usually when I’m at my most competent. </strong>But competence without warmth, in a body people expect warmth from, reads as threat.</p><p>Neither option is free.</p><p>Mask, and people think you’re fine — great, even — but the cost is invisible and cumulative and eventually unsustainable. Run out of mask, and people think something is wrong with you, when the only thing that changed is that they’re seeing you without the filter. <strong>Both cost something.</strong> And I didn’t build this system.</p><h3>What I’d Actually Like</h3><p>I don’t want special treatment. I want the same thing every direct communicator wants, but that autistic AFAB people almost never get:<strong> the benefit of the doubt.</strong></p><p>When my tone doesn’t match your expectations, consider the possibility that your expectations are the variable — not my emotional state. When I say something directly, try hearing the words before you interpret the delivery. When my face doesn’t perform the emotion you’d expect, consider that my face might not be a reliable broadcast system for what I’m actually feeling, and that’s okay.</p><p>It’s not broken. It’s just not your channel.</p><p>And if you’ve worked with me for months and your experience has been that I’m collaborative, reliable, and easy to work with — and then one week I seem “off” — maybe the question isn’t “<em>what’s wrong with Liz.</em>” Maybe the question is “<em>what has Liz been sustaining this whole time, and what does it mean that it’s showing?</em>”</p><p>The research is clear: the communication gap between autistic and neurotypical people is mutual, not one-sided. My directness is not aggression. My flat affect is not hostility. My focus face and my angry face might look the same to you, but they are not the same experience for me. And I promise you — if I were actually mad, you would not need to wonder about it.</p><p><strong>I’m not mad. I’m autistic.</strong></p><p>And I’m tired of being translated by people who don’t speak my language and then being told the translation is my fault.</p><p><em>Liz Dugan is an autistic UX researcher and roller derby official with a master’s in cognitive psychology. They’ve spent a decade leading research in high-stakes environments, and their colleagues would tell you they’re collaborative, organized, and fun to work with — which is both true and the result of a performance most people don’t know they’re watching. Their feedback has been called “severe,” their directness has been called “intense,” and their face has been called “a lot.” They remain, at press time, </em><strong><em>not mad</em></strong><em>. Find their other writing on Medium.</em></p><h3>Sources</h3><ol><li>Milton, D.E.M. (2012). On the ontological status of autism: the ‘double empathy problem.’ <em>Disability &amp; Society</em>, 27(6), 883–887.</li><li>Crompton, C.J., Ropar, D., Evans-Williams, C.V.M., Flynn, E.G., &amp; Fletcher-Watson, S. (2020). Autistic peer-to-peer information transfer is highly effective. <em>Autism</em>, 24(7), 1704–1712.</li><li>Cook, R., Brewer, R., Shah, P., &amp; Bird, G. (2013). Alexithymia, not autism, predicts poor recognition of emotional facial expressions. <em>Psychological Science</em>, 24(5), 723–732.</li><li>Bird, G. &amp; Cook, R. (2013). Mixed emotions: the contribution of alexithymia to the emotional symptoms of autism. <em>Translational Psychiatry</em>, 3, e285.</li><li>Sasson, N.J., Faso, D.J., Nugent, J., Lovell, S., Kennedy, D.P., &amp; Grossman, R.B. (2017). Neurotypical peers are less willing to interact with those with autism based on thin slice judgments. <em>Scientific Reports</em>, 7, 40700.</li><li>Bargiela, S., Steward, R., &amp; Mandy, W. (2016). The experiences of late-diagnosed women with autism spectrum conditions: An investigation of the female autism phenotype. <em>Journal of Autism and Developmental Disorders</em>, 46(10), 3281–3294.</li><li>Lai, M.C., Lombardo, M.V., Auyeung, B., Chakrabarti, B., &amp; Baron-Cohen, S. (2015). Sex/gender differences and autism: Setting the scene for future research. <em>Journal of the American Academy of Child and Adolescent Psychiatry</em>, 54(1), 11–24.</li><li>Hull, L., Petrides, K.V., Allison, C., Smith, P., Baron-Cohen, S., Lai, M.C., &amp; Mandy, W. (2017). “Putting on my best normal”: Social camouflaging in adults with autism spectrum conditions. <em>Journal of Autism and Developmental Disorders</em>, 47(8), 2519–2534.</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ec68fc3856e5" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/im-not-mad-i-m-autistic-ec68fc3856e5">I’m not mad; I’m autistic</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I can’t do a lot. But that can be everything to my kid]]></title>
            <link>https://eadugan28.medium.com/i-cant-do-a-lot-but-that-can-be-everything-to-my-kid-e828c9ded66d?source=rss-4fa13b03a8f6------2</link>
            <guid isPermaLink="false">https://medium.com/p/e828c9ded66d</guid>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[parenting]]></category>
            <category><![CDATA[chronic-illness]]></category>
            <dc:creator><![CDATA[Liz Dugan]]></dc:creator>
            <pubDate>Wed, 24 Jun 2026 15:01:03 GMT</pubDate>
            <atom:updated>2026-06-24T15:01:03.329Z</atom:updated>
            <content:encoded><![CDATA[<p><em>On the math of a small, finite body and an enormous love, and why “just good enough” is, scientifically, the whole game.</em></p><p>I need you to know that I wrote this for myself; but maybe, one day, my kiddo will read it and see just how much I love him.</p><p>But first, <strong>let me tell you what I can’t do.</strong></p><p>I can’t be the parent who’s vertical from dawn to bedtime. I can’t promise my body a week in advance. I can’t be the one who never has to cancel, never has to lie down in the middle of the afternoon, never has to do the awful arithmetic of <em>if I spend it here, what won’t I have left for bedtime.</em> There’s a whole catalogue of parenting I will simply never qualify for, and I’ve made a concerted effort to stop applying for those jobs. The rejection letters were killing me.</p><p>So here is the sentence my brain hands me, helpfully, on the bad days, in the smug little voice of a critic who has never once parented anyone: <em>You can’t do a lot.</em></p><p>And the sentence is true. That’s the irritating part. <strong>I <em>can’t</em> do a lot.</strong></p><h4>But.</h4><h3>What I can do.</h3><p>I can be the <strong>maker.</strong> The one who sits at the kitchen table with the glue that doesn’t dry fast enough and the googly eyes that go everywhere and helps a small person bring a thing out of their head and into the world. That doesn’t take a marathon body. It takes a chair, a flat surface, and the willingness to take seriously a vision that involves forty pipe cleaners and a deeply non-negotiable amount of glitter.</p><p>And sometimes being the maker is quieter and stranger than that. Sometimes it’s me, at the table, making a leather wallet from scratch — cutting it, stitching it, the whole thing — and then <em>painting</em> it to match one of my kid’s drawings at their request. Not a tidied-up version. The real one. Down to every precise yet <em>messy</em> stroke, the place where the line wobbled, the color that went slightly outside where it meant to. Because the wobble is the point. The wobble is <em>him.</em> I am, essentially, a forger of my own child’s art, and I will reproduce his imperfect masterpiece stitch by stitch, stroke by stroke, so that a thing he made with three markers and total conviction gets to live on something built to last. I even sacrifice skin to put his art prominently on my body for all to see. That’s not a marathon. That’s a chair, good light, and a refusal to let his vision be temporary.</p><p>I can be the <strong>baker.</strong> And not the box-mix kind, though we’ve teamed up against a box and our own optimism plenty of times. I bake <em>breads.</em> And my kid — my kid <em>loves</em> the bread, but what he loves most, what he will stop dead in the middle of anything to come do, is <strong>punching the dough.</strong> First proof. That moment when it’s risen and soft and <em>smug</em> and just sitting there asking for it. Wherever he is, whatever he’s doing, the second it’s time, he materializes in the kitchen to drive his fist into it with the focused glee of a tiny demolition crew. It’s his job now. It’s a ritual we both honor. And here’s the quiet luxury of it: rising dough doesn’t care that I had to sit down while making it. The bread keeps its own schedule, gentle and patient, and somehow it has made room in itself for exactly the kind of parent I am — one who can’t rush, and doesn’t have to, because the dough won’t either. <strong>The memory doesn’t ask how many spoons it cost me.</strong> The memory smells like warm flour and sounds like a kid yelling “is it ready to punch yet.”</p><p>I can be the <strong>fixer.</strong> The toy with the snapped wheel, the seam that came apart, the thing that means more than its three dollars of plastic because it’s <em>theirs.</em> I can sit with the little surgical kit of tape and glue and stubbornness and bring the beloved thing back to life. Hell, I’d craft a tiny forge if that was what was necessary. There’s a specific way a kid looks at you when you’ve repaired something they thought was gone forever, and <strong>I would crawl over broken glass for that look</strong>, except I don’t have to, because fixing things is, blessedly, a low-impact activity (for the most part).</p><p>I can be the <strong>creator.</strong> The inventor of the bit, the keeper of the running joke, the one who builds the small private world that only the two of us live in — the voices we use to greet each other, the made-up rules, the stories that pick up exactly where we left them. That kind of building doesn’t require a single thing my illness has taken. It just requires <em>me,</em> which it turns out I still have plenty of.</p><p>I can be the other half of the <strong>Boogie-Dugies.</strong> Lights off, blanket up to the chin, a bowl between us of the official sanctioned mixture — popcorn, rainbow Goldfish, and every flavor of Pez, a combination no nutritionist would sign off on and no force on earth could make us change. And the show is always a ghost-hunting one, because we are not there to <em>believe.</em> We are there to <em>debunk.</em> We brace for the jump-scare and gasp anyway, and then immediately turn into the world’s smallest, most skeptical forensics team — <em>that’s a draft, that’s the camera autofocusing, that “voice” is the radiator, absolutely not, next.</em> I can’t always run around outside, but <strong>I can one hundred percent be the person you choose to be deliciously terrified next to </strong>and then dismantle the evidence with. That’s a role with a horizontal option. And I excel at it.</p><figure><img alt="Illustrated portrait of a parent with silver-gray bob and glasses sitting on a couch in dim blue light, a small blonde curly-haired child nestled on their lap. The child is beaming, clutching popcorn. The parent’s expression is soft, tired, present — arm wrapped around the kid, blanket draped over them both. The glow of an unseen screen lights their faces. The room is dark and quiet. Nothing is happening. Everything is happening." src="https://cdn-images-1.medium.com/max/1024/1*OrnZ84buXmMRKNdOStEKaA.jpeg" /><figcaption>Most of my best parenting happens in the margins. Not the big moments — the small ones. The ones where I’m too tired to perform productivity but still here, still holding him, still showing up in the only way I can right now. This is what “I can’t do a lot” actually looks like. It looks like enough.</figcaption></figure><p>I can be the <strong>Markiplier “play with” partner.</strong> This one I started — I introduced him to <em>my</em> Markiplier videos, the ones I already loved, because he was curious one day, and the plan backfired in the most wonderful possible way:<strong> I now have a <em>permanent</em> watch partner</strong>, and the truth is I can’t watch them without him anymore. It would feel wrong. He’s part of the experience now — knowing why this part is funny and that part is scary, the inside references, the shared flinch. Being a ‘<em>with’</em> person costs almost nothing physically and is worth almost everything. I can do ‘<em>with’</em> all day long.</p><p>And I can be the <strong>waterpark connoisseur</strong> — a love I also handed down, deliberately, and have zero regrets about. I introduced my son to waterparks and now we have <em>opinions,</em> and we are <em>not</em> apologetic about them. I know which park, which slide, which time of day, where the good shade is, which snack stand is a trap and which is a triumph. I may have to plan the energy for it like a heist; I may have to nap before and crater after. But there is always a lazy river when I need to recover, <strong>and even my breaks are part of the game</strong>: he spends them trying to <em>escape</em> me, the two of us on donuts, maneuvering to dodge the waterfalls and the cannon spray. And here’s my favorite part — sometimes I steer <em>into</em> a trap on purpose. I take the full blast. <strong>Because watching me squirm and sputter is one of his most cherished pastimes on this earth,</strong> and I will gladly get hosed in the face to be the reason he’s laughing that hard. The planning is the love. The map in my head is the love. The deliberate faceful of water is the love. I built all of it <em>for</em> him.</p><h3>The thing about a little.</h3><p><strong>Look at that list again</strong>. None of it is <em>a lot,</em> by the world’s exhausting definition of a lot. Nobody’s giving out medals for sitting at a table with glue. There’s no parenting trophy for knowing the waterpark.</p><p>But to my kid, that list isn’t <em>a ‘little.’</em> To my kid, that list is the <em>whole shape of being loved.</em> And here’s the thing I have spent years and a great deal of difficult, interior renovation learning to believe: <strong>those two truths are allowed to coexist.</strong> I can do a little, <em>and</em> what I do can be everything. The size of the act and the size of what it means to the person receiving it are measured on completely different scales.</p><p>I am not the sum of what my body can’t do. <strong>I’m the sum of the table, the oven, the dough, the painted wallet, the tape, the bit, the blanket, the controller, and the map.</strong> That’s not a consolation prize. That’s a <em>childhood.</em></p><h3>The part where the science backs me up.</h3><p>My therapist is the one who first handed me <strong>the idea of the “good enough” parent.</strong> They said it gently, the way you offer someone a chair they didn’t know they were allowed to sit in. And because I am who I am, I couldn’t just accept the comfort — I had to go check whether it actually held weight. It does. It holds an <em>enormous</em> amount.</p><p>The whole idea didn’t come from a motivational mug. It came from the British pediatrician and psychoanalyst Donald Winnicott, who spent the 1940s and 50s watching real mothers with real babies and arrived at a frankly radical idea for his time: the “good enough mother,” the “ordinary devoted mother,” who loves their child <em>and inevitably gets it wrong,</em> is not a failed version of some perfect parent. <strong>They’re the actual ideal.</strong> Winnicott argued that a parent’s small, <em>ordinary</em> failures — the moment you don’t arrive instantly, the need you don’t perfectly anticipate — are not damage. They’re <em>how a child learns the world is real and survivable.</em> The perfect, frictionless parent would actually rob the kid of something. The concept was explicitly designed to <em>defend</em> ordinary parents against the cruelty of the idealized one.</p><p>Decades later, the psychologist Bruno Bettelheim took Winnicott’s phrase and made it the title of an entire book — <em>A Good Enough Parent</em> (1987) — whose entire argument is that the <em>pursuit</em> of perfect parenting is itself harmful, and that <strong>the ordinary, fallible, loving parent is exactly what a child needs.</strong></p><p>And then, because feelings are nice but <em>data</em> is better, a team led by the psychologist Susan Woodhouse went and <em>measured</em> it. In a 2020 study published in <em>Child Development</em> — one of the most-cited articles that journal ran that year — Woodhouse and her colleagues followed 83 mothers and their babies, deliberately choosing infants who were temperamentally irritable and families under socioeconomic strain. They were looking for what it actually takes to produce a <em>securely attached</em> child. The answer? Caregivers needed to “get it right” only about <strong>50 percent of the time</strong> to reliably produce secure attachment. Half. Not all of it. Not most of it. <em>Half,</em> delivered with the willingness to repair the other half.</p><p>Let that number sit on the bad days. <strong>Fifty percent.</strong> I have spent years feeling like a C-minus parent or worse because I can’t run the full course, and the most rigorous research we have on the subject says the passing grade was a coin flip the whole time — and that what closes the gap isn’t stamina, it’s <em>showing up to repair, and meaning it.</em> That, for the most part, I can do lying down.</p><h3>So here’s where I land.</h3><p>I can’t do a lot. I want to be honest about that, the same way I’m honest about everything, because pretending otherwise would make a liar of me and my kid deserves a parent who tells the truth.</p><p>I can’t do a lot. <strong>But what I can do, I do all the way. And to my kid, what I can do is everything.</strong></p><p>The maker, the baker, the fixer, the creator, the Boogie-Dugie, the play-with partner, the keeper of the waterpark maps. That’s not the parenting I’d have ordered from a catalogue of unlimited bodies. But it’s mine, it’s real, and I am actively working to believe that that is enough — and now <strong>I’ve got the peer-reviewed receipts to prove that “enough” was never the consolation prize.</strong></p><p>It was the whole game.</p><p><em>I am a non-binary mom, AuDHD, and chronically ill. I write essays about the small, devastating, faintly ridiculous moments that shape how we see ourselves — in the hope someone else finds themselves in them, too. Mostly, though, I’m just here for the snacks.</em></p><p><strong>Sources</strong></p><p><em>Donald W. Winnicott — originator of the “good enough mother” and “ordinary devoted mother” concepts, developed across his work in the 1940s–50s (notably “Mind and its Relation to the Psyche-Soma,” 1949, and “Transitional Objects and Transitional Phenomena,” 1953). The concept was explicitly framed to defend ordinary parents against idealization and to argue that a parent’s inevitable small failures are developmentally necessary, not harmful.</em></p><p><em>Bruno Bettelheim (1987), “A Good Enough Parent: A Book on Child-Rearing,” Alfred A. Knopf — extends Winnicott’s idea into a book-length argument that striving for perfect parenting is itself counterproductive.</em></p><p><em>Susan S. Woodhouse, Julie R. Scott, Allison D. Hepworth &amp; Jude Cassidy (2020), “Secure Base Provision: A New Approach to Examining Links Between Maternal Caregiving and Infant Attachment,” Child Development, 91(1), e249–e265. In a sample of 83 low-SES mothers and temperamentally irritable infants, caregivers who responded appropriately to attachment needs roughly 50% of the time reliably produced secure attachment — emphasizing that consistent repair, not constant perfection, is what matters.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e828c9ded66d" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>