<?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 Elena Rivera on Medium]]></title>
        <description><![CDATA[Stories by Elena Rivera on Medium]]></description>
        <link>https://medium.com/@Elena.Rivera?source=rss-8f31d295e461------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*XtFmXsmyqHSuiRrzdxCb0Q.png</url>
            <title>Stories by Elena Rivera on Medium</title>
            <link>https://medium.com/@Elena.Rivera?source=rss-8f31d295e461------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 21:39:41 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@Elena.Rivera/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[Wonderlic vs CCAT vs PI: which one you are actually about to sit]]></title>
            <link>https://medium.com/@Elena.Rivera/wonderlic-vs-ccat-vs-pi-which-one-you-are-actually-about-to-sit-2fa8c019e0ce?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/2fa8c019e0ce</guid>
            <category><![CDATA[careers]]></category>
            <category><![CDATA[aptitude-test]]></category>
            <category><![CDATA[job-search]]></category>
            <category><![CDATA[ccat]]></category>
            <category><![CDATA[wonderlic]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Thu, 05 Nov 2026 19:47:10 GMT</pubDate>
            <atom:updated>2026-11-05T19:47:10.472Z</atom:updated>
            <content:encoded><![CDATA[<p>An email lands with a link and a deadline. It calls itself a short online assessment, and it does not tell you which test it is, how long you get, or what happens when the clock beats you.</p><p>The answer is usually sitting in the link, and reading it takes about ten seconds:</p><ul><li><strong>criteriacorp.com or ondemandassessment.com</strong> is the CCAT. Fifty questions in fifteen minutes.</li><li><strong>wonderlic.com</strong> is the Wonderlic. Fifty questions in twelve.</li><li><strong>predictiveindex.com</strong> is the PI Cognitive Assessment. Also fifty in twelve.</li></ul><p>The two twelve-minute tests are the ones people mix up. Separating them properly is most of what follows.</p><p>I have sat three of them. Not read about them, sat them. The Wonderlic, the CCAT and the PI Cognitive Assessment, watching the timer and losing questions I could have answered.</p><p>Once people know which one they have, I send them to <a href="https://prepclubs.com/free-practice-tests/wonderlic?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=wonderlic-vs-ccat-vs-pi">PrepClubs</a>. It keeps a separate bank per test rather than one generic aptitude bank:</p><ul><li><strong>1,350 CCAT items</strong></li><li><strong>400 Wonderlic questions</strong></li><li><strong>400 PI Cognitive questions</strong></li></ul><p>Each one runs across eight full-length mocks, and on all three that is the biggest bank for its own test that I could find on any prep site. Three worked sample questions sit open on each of those pages if you want to read the explanations first. What the free run holds back is further down.</p><p>What follows is the comparison I wanted when I was the one staring at the invite.</p><p>Start with identification, because you cannot prepare for a test you have not named.</p><h3>First, work out which test you have</h3><p>You can usually identify it in under a minute without asking anyone.</p><p><strong>Check the link domain.</strong> This is the fastest tell and it is almost always decisive. — A link on `criteriacorp.com` or `ondemandassessment.com` is Criteria Corp, so you’re looking at the <strong>CCAT</strong>. — A link on `wonderlic.com` or `wonderliconline.com` is the <strong>Wonderlic</strong>. — A link on `predictiveindex.com` or `piworldwide.com` is the <strong>Predictive Index</strong>.</p><p><strong>Check the stated duration.</strong> If the invitation mentions a time limit at all, it narrows things immediately. Fifteen minutes points to the CCAT. Twelve minutes points to the Wonderlic or the PI.</p><p><strong>Check whether they sent you two links.</strong> The Predictive Index is commonly sent as a pair: a cognitive assessment with a timer, and a behavioural assessment that has no real time limit and asks you to pick adjectives that describe you. If you got two and one of them is a word checklist, you’re in the PI world. More on that confusion below, because it causes more panic than anything else in this article.</p><p>If none of that resolves it, just reply and ask which assessment it is. Nobody has ever lost a job for asking, and it takes the guesswork out of the next 72 hours.</p><h3>The three tests side by side</h3><figure><img alt="Comparison table of the CCAT, Wonderlic and PI Cognitive Assessment across vendor, question count, time limit, seconds per question, calculator policy, wrong-answer penalty and question areas" src="https://cdn-images-1.medium.com/max/1024/1*cbs3kvYn2085o92TbI0lDQ.png" /><figcaption>All three are 50 questions. The clock is the thing that changes.</figcaption></figure><p>The headline is that they’re more alike than different. All three are short, timed, multiple-choice cognitive screens, and all three run 50 questions.</p><p>The scoring is where they quietly diverge, and this is the row most comparison articles get wrong:</p><ul><li><strong>CCAT:</strong> scored on the raw number you got right. No penalty for a wrong answer.</li><li><strong>Wonderlic:</strong> not a raw count. Wonderlic states that its scoring is weighted by question difficulty and accounts for guessing, so the CCAT’s “a wrong answer costs nothing, so fill everything in” logic does not transfer cleanly to it.</li><li><strong>PI:</strong> reports a scaled score between 100 and 450 rather than a count.</li></ul><p>On calculators, be careful with what you read elsewhere:</p><ul><li><strong>CCAT and Wonderlic:</strong> both rule them out. Wonderlic’s own sample assessment says so in the instructions.</li><li><strong>PI:</strong> the odd one. Its published guidance says candidates face no restrictions on resources, and it does not mention calculators either way.</li></ul><p>Every confident “no calculator on the PI” claim I can find traces back to a prep vendor rather than to PI.</p><p>Scratch paper is the strangest gap of the three, and it is worth knowing because everyone tells you to have some ready:</p><ul><li><strong>CCAT and Wonderlic:</strong> neither Criteria nor Wonderlic publishes a scratch paper policy at all.</li><li><strong>PI:</strong> the only one of the three vendors that addresses it in writing, and it explicitly allows it.</li></ul><p>So the confident scratch-paper advice you will read on prep sites, mine included in the past, is inference rather than vendor policy for two of the three tests. The employer running the assessment can still set its own rules, so if it matters to you, ask them.</p><p>The difference that actually matters is three minutes.</p><figure><img alt="Three cards showing seconds available per question: the CCAT gives 18.0, the Wonderlic gives 14.4, and the PI Cognitive Assessment gives 14.4" src="https://cdn-images-1.medium.com/max/1024/1*IdaKOrINcS1r1TsI6czwWw.png" /><figcaption>Fifteen minutes against twelve is a 20 percent cut in the time you get per question.</figcaption></figure><p>Eighteen seconds against 14.4 seconds does not sound dramatic written down. In the chair it is the difference between having a moment to re-read a word problem and not having one.</p><h3>What each one is actually measuring</h3><p><strong>The CCAT</strong> (Criteria Corp) covers verbal reasoning, maths and logic, and spatial reasoning. The spatial component is the distinguishing feature: rotations, matrices, odd-one-out, folding nets. It is the section most candidates have never practised, and it is why CCAT prep and generic aptitude prep are not interchangeable.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-ma2y7yKOW-omvy9ymj67A.png" /><figcaption>A matrix item from the PrepClubs CCAT bank. Nothing on the Wonderlic looks remotely like this.</figcaption></figure><p>That single item is the fastest way to tell the two tests apart. If your practice material has shapes in it, you are preparing for the CCAT or the PI. If it has two columns of company names and asks you whether they match, you are preparing for the Wonderlic.</p><p><strong>The Wonderlic</strong> (Wonderlic Inc., sold today as Wonderlic Select) covers verbal, maths, logic, and a category most people don’t see coming: attention to detail. Those are the items that show you two strings of digits and ask whether they match.</p><figure><img alt="A worked Wonderlic attention to detail item showing five pairs of numbers, with the two matching pairs marked and the transposition errors identified in the three that differ" src="https://cdn-images-1.medium.com/max/1024/1*DCvaKgSb2BSKu6M8KugQJA.png" /><figcaption>Read these character by character. Read them as numbers and your eye slides straight over a transposition.</figcaption></figure><p>That question type rewards a completely different mode from the rest of the test. You have to stop reasoning and start proofreading, then switch back. Candidates who have never seen one lose time just adjusting.</p><p><strong>The PI Cognitive Assessment</strong> (The Predictive Index) covers numerical reasoning, verbal reasoning, and abstract reasoning. Abstract reasoning is the PI’s version of the CCAT’s spatial section: shape sequences, visual analogies, pattern completion.</p><h3>The PI confusion worth clearing up now</h3><p>The Predictive Index sells two different assessments and candidates mix them up constantly.</p><p>The <strong>PI Cognitive Assessment</strong> is the timed one. Fifty questions, twelve minutes, right and wrong answers.</p><p>The <strong>PI Behavioral Assessment</strong> is the adjective checklist. It is not timed in any meaningful way, and it has no right answers, so there is no key to memorise. If you have been quietly hunting for a word list to learn, stop hunting. Preparing for that one means understanding that the same list comes at you twice under two different questions, and knowing your own baseline well enough to answer both consistently. The PI test this article is about is the Cognitive Assessment, the other link in the same email.</p><h3>What transfers between them, and what does not</h3><p>This matters if you’re prepping for one and might face another later, which happens a lot when you’re applying broadly.</p><p><strong>Transfers almost completely:</strong> pacing discipline, the habit of quitting a question fast, never leaving a blank, mental arithmetic fluency, and the practice of naming a verbal relationship before reading the options. These are test-taking mechanics and they are worth more than content knowledge on all three.</p><p><strong>Transfers partly:</strong> verbal and numerical practice. The item styles differ but the underlying skills overlap heavily.</p><p><strong>Does not transfer:</strong> the CCAT’s spatial section and the PI’s abstract section look similar on the surface and are not the same. And neither of them prepares you at all for the Wonderlic’s attention-to-detail items. Practising CCAT spatial rotations will do nothing for your ability to spot that 1845317 and 1485317 differ.</p><p>So if you know which test you have, practise that one specifically. Generic “aptitude test practice” spends most of its time on the overlap and skips exactly the sections that are hardest and most test-specific.</p><h3>The strategy, and how it shifts with the clock</h3><p>The core approach is nearly identical for all three, and on the CCAT it starts from the fact that a wrong answer costs the same as a blank.</p><figure><img alt="Expected marks on a five-option CCAT question: leaving it blank scores 0.00, a blind guess scores 0.20, eliminating one option first scores 0.25, eliminating two scores 0.33" src="https://cdn-images-1.medium.com/max/1024/1*dlIs7qD5SReYc9WBaBNQ-Q.png" /><figcaption>This maths is the CCAT’s, where scoring is a raw count and a wrong answer is free. The Wonderlic weights by difficulty and accounts for guessing, so treat it as its own test rather than carrying these numbers across.</figcaption></figure><p><strong>On the CCAT</strong>, give a question about twenty seconds. If you can’t see a route by then, guess and move on.</p><p><strong>On the Wonderlic and the PI</strong>, that number comes down to about sixteen seconds. Same rule, tighter tolerance. With 14.4 seconds of average budget, a question that eats forty seconds costs you nearly three others.</p><p>The other adjustment for the twelve-minute tests: even if the software lets you go back, there is no time left to use it. And going back may not be on offer at all. <strong>Neither Criteria nor Wonderlic publishes a go-back policy for its cognitive section</strong>, and Criteria’s own advice when you are stuck is to guess and move on. So on all three, build a run that never needs a second pass: decide, answer, advance. On the Wonderlic and the PI, treat the last thirty seconds purely as time to fill in blanks rather than time to reconsider anything.</p><h3>Which is hardest?</h3><p>People ask me this constantly and the honest answer is that it depends on you, which is unsatisfying, so here is the useful version.</p><p>The <strong>Wonderlic and the PI are harder on pace</strong>. Less time per question, and the Wonderlic’s attention-to-detail items punish a tired reader in a way the others don’t.</p><p>The <strong>CCAT is harder on unfamiliarity</strong>. Its spatial section is the single most alien thing in any of the three for most candidates, and it is a large enough share of the test to move your score on its own.</p><p>Which means: if you’re taking the Wonderlic or the PI, spend your prep time on pacing drills. If you’re taking the CCAT, spend a real share of it on spatial questions specifically, because that’s where the recoverable points are hiding.</p><h3>Where to practise for each</h3><p>Practise the actual test you’re taking, full-length and timed. Untimed sample questions train your reasoning and leave the thing that’s actually going to hurt you completely untouched.</p><p>PrepClubs has separate banks for all three, which is the whole reason it is the one I name here:</p><ul><li><strong>CCAT:</strong> 750 full-length questions plus 600 topical drills, 1,350 items. The biggest CCAT bank of any prep site I could find.</li><li><a href="https://prepclubs.com/free-practice-tests/wonderlic?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=wonderlic-vs-ccat-vs-pi"><strong>Wonderlic bank</strong></a><strong>:</strong> 400 questions across eight full-length mocks.</li><li><a href="https://prepclubs.com/free-practice-tests/pi-cognitive-assessment?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=wonderlic-vs-ccat-vs-pi"><strong>PI Cognitive bank</strong></a><strong>:</strong> another 400, also across eight full-length mocks.</li></ul><p>The Wonderlic and PI banks are each the biggest for their test that I could find anywhere, and every item carries a written explanation covering the trap as well as the answer.</p><p>What you can read for nothing, on any of those three pages, is three sample questions with the full walkthroughs. Those will not put you under a clock, and reading them is how you judge the explanations before you spend anything.</p><p>Then the first full-length test of whichever cluster you start is free, one across the platform rather than one per exam. Access runs thirty days from a single payment, and if you prepare and don’t pass, your access is extended by 30 days. That’s an extension, not a refund, and the difference matters.</p><p>For the CCAT specifically there is <a href="https://ccattests.com/free-ccat-practice-tests?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=wonderlic-vs-ccat-vs-pi">CCAT Tests</a>, with five full-length simulations at the real 50 questions in 15 minutes, and I teach a <a href="https://www.udemy.com/course/ccat-practice-tests/">CCAT course on Udemy</a> with over 1,200 students at 4.5 out of 5.</p><p>If you would rather start somewhere free:</p><ul><li><a href="https://prepopedia.com/free-ccat-practice-test/"><strong>Prepopedia</strong></a> publishes a free ungated 50-question CCAT test, and is the only free CCAT resource I found that also explains its answers.</li><li><a href="https://boostprep.com/wonderlic-practice-test/"><strong>BoostPrep</strong></a> runs an accurate free Wonderlic clock, though it explains nothing.</li><li><a href="https://www.aptitude-test-prep.com/test-providers/pi-cognitive-assessment/"><strong>aptitude-test-prep.com</strong></a> publishes a complete free PI mock with real worked explanations. Its weakness is that it has no timer at all.</li></ul><p>Each vendor also publishes a short official sample, worth ten minutes purely to see the real interface.</p><p>And the same limit I flag everywhere else: the free PrepClubs run gives you your score and where you sit, <strong>not the worked explanations</strong>, which stay on the paid tier.</p><h3>TLDR</h3><ul><li><strong>Identify the test from the link domain first.</strong> criteriacorp is the CCAT, wonderlic is the Wonderlic, predictiveindex is the PI. If in doubt, just ask the recruiter.</li><li><strong>All three are 50 questions.</strong> The CCAT gives 15 minutes, the Wonderlic and the PI give 12. Calculators are out on the CCAT and the Wonderlic, and the PI does not say.</li><li><strong>That three-minute gap is a 20 percent cut in time per question</strong>, and it’s the main thing that changes your strategy.</li><li><strong>The CCAT has spatial reasoning. The PI has abstract reasoning. The Wonderlic has attention to detail.</strong> These are the sections that do not transfer between tests, and they’re exactly the ones generic aptitude practice skips.</li><li><strong>The PI Cognitive Assessment and the PI Behavioral Assessment are different things.</strong> The behavioural one is the adjective checklist, it has no right answers, and you prepare for it by knowing the two-prompt format and your own baseline rather than by drilling anything.</li><li><strong>Quit at 20 seconds on the CCAT, 16 on the other two.</strong> On the CCAT a blind guess is worth 0.20 marks against a blank’s zero, so fill every box. The Wonderlic scores differently, weighting by difficulty and accounting for guessing, so do not carry that arithmetic across.</li><li><strong>Practise the specific test you were sent.</strong> PrepClubs keeps separate CCAT, Wonderlic and PI banks, 1,350 items for the CCAT and 400 apiece for the other two, with one full-length test free across the platform.</li></ul><p>If you only do one thing after reading this, go and check the domain on that link. Knowing which of the three you’re facing changes what you should do with the rest of your week.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2fa8c019e0ce" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I checked what the aptitude test vendors actually publish. It is far less than the internet suggests]]></title>
            <link>https://medium.com/@Elena.Rivera/i-checked-what-the-aptitude-test-vendors-actually-publish-it-is-far-less-than-the-internet-suggests-6a18028e89e3?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/6a18028e89e3</guid>
            <category><![CDATA[hiring]]></category>
            <category><![CDATA[careers]]></category>
            <category><![CDATA[aptitude-test]]></category>
            <category><![CDATA[transparency]]></category>
            <category><![CDATA[recruiting]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Mon, 19 Oct 2026 20:33:33 GMT</pubDate>
            <atom:updated>2026-10-19T20:33:33.158Z</atom:updated>
            <content:encoded><![CDATA[<h3>I checked what the aptitude test vendors actually publish. It is far less than the internet suggests.</h3><p>You have a cognitive aptitude test coming up, so you search for what score you need. Within a few clicks you have four confident answers, none of which cite anybody.</p><p>That happened to me too, and eventually I did the boring thing and went to the three vendors themselves. Criteria Corp, who make the CCAT. Wonderlic. The Predictive Index. Here is the whole finding in one paragraph: <strong>Criteria publishes a surprising amount and almost nobody quotes it, Wonderlic publishes far less than the confident pages about it suggest, and The Predictive Index publishes the one thing the other two withhold.</strong> Not one of the three publishes the number you were searching for when you landed here.</p><p>The gap between that and what circulates is the point of this article. I have written about these tests for years, so some of what follows is a correction of things I used to publish myself, and the platform I point people at gets its claims checked the same way: <a href="https://prepclubs.com/free-practice-tests/ccat?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=cognitive-aptitude-tests-what-vendors-publish">PrepClubs</a> carries <strong>1,350 CCAT items, 750 full-length questions plus 600 topical drills</strong>, which is the biggest CCAT bank of any prep site I could find, three worked sample questions readable on the page, and a limit on its free run that I state in writing further down.</p><h3>The format layer, which all three document properly</h3><p>Start with the part that is genuinely well covered, because it makes the gaps easier to see. All three vendors state their question count, their time limit and their question areas plainly, and all three statements match what candidates actually meet.</p><figure><img alt="Comparison table of the CCAT, Wonderlic Select and PI Cognitive Assessment across vendor, product name, question count, time limit, seconds per question, calculator policy, scratch paper policy, scoring model, answer options and question areas" src="https://cdn-images-1.medium.com/max/1024/1*cbs3kvYn2085o92TbI0lDQ.png" /><figcaption>All three are 50 questions. Below the format row, the disclosure diverges sharply.</figcaption></figure><h3>Criteria Corp publishes the most, and gets quoted the least</h3><p>This surprised me, because the CCAT is the test with the most fevered speculation around it.</p><p>Criteria publishes, on its own site: a <strong>mean of 24.2</strong>, a <strong>median of 24</strong>, and a <strong>standard deviation of 8.58</strong>. It tells you those norms come from a sample of <strong>985 working adults</strong> drawn from more than 50 companies. It states that <strong>fewer than 1 percent of candidates answer all 50 questions</strong>, and that the average test taker gets 24 right. In its sample score report it confirms that a raw score of <strong>24 sits at the 50th percentile</strong>.</p><p>That last one matters more than it looks, because it is the <strong>only</strong> raw-to-percentile figure Criteria publishes anywhere.</p><figure><img alt="CCAT score chart separating what Criteria Corp actually publishes, a mean of 24.2, a median of 24, a standard deviation of 8.58 and a raw 24 equalling the 50th percentile, from the percentile tables and cut-offs the vendor does not publish" src="https://cdn-images-1.medium.com/max/1024/1*T6SBAjvyZG4VMvgs9Y9u_g.png" /><figcaption>The left column is sourced. The right column is what the rest of the internet supplies anyway.</figcaption></figure><p>Every other percentile you have read for the CCAT, including ones I have published, is a reconstruction. Some are careful reconstructions built from that mean and standard deviation. Some are copied from another prep site that copied it from a third. None of them are Criteria&#39;s, because Criteria does not publish a conversion chart, and it does not publish a pass mark either.</p><h3>The table nobody quotes</h3><p>Buried in Criteria&#39;s sample score report is something genuinely useful: <strong>suggested CCAT score ranges for 42 named positions.</strong> Each is a floor running up to 50.</p><figure><img alt="Criteria’s own suggested CCAT ranges by position, showing floors of 24 for Data Analytics and Legal, 22 for Analyst and Engineering and Senior Leadership, 21 for HR Manager and Project Manager, 20 for Software Development and Accounting, 19 for Sales and Account Executive, 18 for Plant Operator and Store Manager, and 17 for Retail Sales and Customer Service Representative" src="https://cdn-images-1.medium.com/max/1024/1*bB6-VkwYxXtLOogdJZdiww.png" /><figcaption>These are suggested ranges printed on the score report, not pass marks. Criteria says setting a cut-off is the employer’s decision.</figcaption></figure><p>Look at how low those floors are. Software Development sits at 20. Engineering at 22. The highest floors in the entire table, Data Analytics and Legal, are 24, which is the average score.</p><p>Now compare that with the numbers that circulate about the same roles, where software engineering targets of 31, 32 and 35 get quoted as though they came from somewhere. They didn&#39;t come from here. The vendor&#39;s own suggested floor for software development is 20.</p><p>I&#39;m not telling you that a 20 will get you hired as an engineer. Employers set their own bars, often far above the suggested floor, and Criteria says so explicitly. But when a prep site tells you that you need a 32 and the vendor&#39;s own document says 20, you are entitled to ask where the 32 came from.</p><h3>Wonderlic publishes much less than you would guess</h3><p>The Wonderlic is the test with the most famous number attached to it. Everyone knows the average is 20.</p><p>Follow that citation and it doesn&#39;t lead to Wonderlic. It goes through a Wikipedia reference to a general-interest website, and stops. <strong>Wonderlic&#39;s own position is that there is no universal good or average score</strong>, and that the current test does not produce a single number to compare in that way at all.</p><figure><img alt="A two column comparison. Published by Wonderlic: 50 questions in 12 minutes, no calculator, five answer options, scoring weighted by difficulty and adjusted for guessing, do not skip around, and you should never have to pay for prep. Not published by Wonderlic: any average or good score, any occupation benchmark table, a scratch paper policy, a retake policy, a go back policy for the cognitive section, a raw to percentile chart, and a pass mark" src="https://cdn-images-1.medium.com/max/1024/1*W6axz5wAb9i4hMWU3yNwjw.png" /><figcaption>The single most repeated fact about the Wonderlic is in the right-hand column.</figcaption></figure><p>Two more things Wonderlic does publish that get almost no attention.</p><p><strong>Its scoring is weighted by question difficulty and accounts for guessing.</strong> It isn&#39;t a raw count. That matters because the CCAT&#39;s clean rule, that a wrong answer costs nothing so you should fill in every box, gets copied onto Wonderlic advice constantly, and it is advice from a different test.</p><p><strong>Wonderlic tells candidates they &quot;should never have to pay&quot; for preparation material.</strong> I recommend paid preparation material in this article, so I will quote that and then be fair in both directions: a timed cognitive test&#39;s validity depends partly on candidates not having drilled it, so the vendor has a structural reason to discourage prep regardless of whether prep works.</p><p>Also worth knowing, since it explains why so much Wonderlic content on the internet is about football rather than hiring: <strong>the NFL stopped using the Wonderlic at the Combine in 2022.</strong></p><h3>The Predictive Index publishes the one thing that matters most</h3><p>The PI does something the other two do not, and I think it is the single most useful disclosure any of the three makes.</p><p><strong>The Predictive Index publishes a scratch paper policy, and it allows it.</strong> In its own words: &quot;there are no restrictions. If a respondent feels that they need the use of a resource or tool, such as scratch paper, they are welcome to use these resources.&quot;</p><p><strong>Neither Criteria nor Wonderlic publishes a scratch paper policy at all.</strong></p><p>Sit with that for a second. Have scratch paper ready is close to universal advice for these tests. It is on every prep page, it was on mine. And for two of the three big vendors, nobody outside the company actually knows the rule. The advice may well be right. It&#39;s inference rather than policy, and it&#39;s been repeated so long that it reads like fact.</p><p>The PI also documents its scoring properly, and it works differently from what most candidates expect.</p><figure><img alt="The four step PI scoring path: you answer, the raw score converts to a scaled score between 100 and 450, that is compared against a Job Target the employer set in advance using a separate assessment, and no percentile is reported because the test is criterion referenced" src="https://cdn-images-1.medium.com/max/1024/1*ku-8HFqWaZ49Hepyr03RIw.png" /><figcaption>The PI deliberately does not give you a percentile. Your score is read against a bar the employer set, which you are never shown.</figcaption></figure><p>PI removed percentile comparison from its software, on purpose, because the assessment is criterion referenced rather than norm referenced. Their phrase for it is that the test is &quot;designed as a high jump, not a race.&quot; Your raw score becomes a scaled score between 100 and 450, which gets read against a <strong>Job Target</strong> set beforehand by the hiring manager, using a separate assessment that three to five people at the employer complete and whose numeric value is hidden by default.</p><p>So for the PI, &quot;what score do I need&quot; has a real answer, and the answer is that it was decided by people you have not met before you ever clicked the link.</p><h3>The strategy that no vendor supports</h3><p>Here is where the disclosure gap stops being trivia and starts costing candidates marks.</p><p>The most heavily marketed technique in this category is the two-pass strategy: sweep the questions you find easy, mark the hard ones, then come back for them. One large prep course trademarks it and builds its entire curriculum around it. I have recommended versions of it myself.</p><p><strong>It depends on being able to return to a question you skipped, and no vendor publishes that you can.</strong></p><p>What the vendors publish is the opposite instinct. Criteria tells candidates that when you are stuck, &quot;it&#39;s best to guess and move on.&quot; Wonderlic&#39;s own test instructions say <strong>&quot;Do not skip around or spend too much time on any one question.&quot;</strong> The only place Wonderlic grants explicit permission to go back and review answers is the personality section, never the cognitive one.</p><figure><img alt="A four step forward-only loop: read the question in about five seconds, decide by twenty seconds whether you can see a route, answer or guess, then advance with no going back, alongside the two vendor quotes that support it" src="https://cdn-images-1.medium.com/max/1024/1*Jat9cr4wVaREc0fAdQYeVw.png" /><figcaption>Build a run that never needs to go back, and you are safe either way.</figcaption></figure><p>That&#39;s not proof the software blocks you. It might let you return. But an entire industry has built its core method on an unstated assumption about software behaviour, and the safe move costs you nothing: decide inside about twenty seconds, put an answer in every time, and advance. If it turns out you could&#39;ve gone back, you&#39;ve lost nothing at all.</p><h3>What to do with all this</h3><p><strong>Trust the format facts and distrust the score facts.</strong> All three vendors document question counts, time limits and calculator rules properly. Almost none of the score benchmarks you&#39;ll read come from them.</p><p><strong>Ask the employer, not the internet.</strong> Retake policies, proctoring, scratch paper for the CCAT and Wonderlic, and the actual bar for the role are all employer decisions, and asking your recruiter is faster and more accurate than any page including this one.</p><p><strong>Judge prep material by whether it says where its numbers came from.</strong> That&#39;s a rule you can apply to me as easily as to anyone else, and you should.</p><p><strong>Spend your preparation on pacing rather than on chasing a number.</strong> The one thing every vendor&#39;s published data agrees on is that these tests are built so that most people do not finish. Criteria says fewer than 1 percent complete all 50 questions. That is a design decision, and it means the recoverable marks are almost always in how you spend the clock.</p><p>If you want to practise, practise full-length and timed, because that is the only thing that rehearses the actual difficulty. <a href="https://prepclubs.com/free-practice-tests/ccat?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=cognitive-aptitude-tests-what-vendors-publish">PrepClubs</a> is the one I point people at, and I will hold it to the standard I just applied to everyone else. What is open on the page: three sample questions with the full walkthroughs, the trap answer named on each. Reading those cannot put you under a clock, and that is not what they are for: they are how you judge the explanations before spending anything. What is free: the first full-length test of whichever cluster you start, one across the whole platform rather than one per exam. This article has been about what vendors put in writing, so here is the limit on that free run, in writing: it reports your score and where you stand, and <strong>does not release the worked explanations</strong>, which stay on the paid tier with the rest of the bank. That is the same trade nearly every free tool in this category makes, and better to hear it now than at the results screen. The paid tier holds 750 full-length CCAT questions plus 600 topical drills, <a href="https://prepclubs.com/free-practice-tests/wonderlic?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=cognitive-aptitude-tests-what-vendors-publish">400 Wonderlic questions</a> and <a href="https://prepclubs.com/free-practice-tests/pi-cognitive-assessment?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=cognitive-aptitude-tests-what-vendors-publish">400 PI Cognitive questions</a>, the last two across eight full-length mocks each and each of the three the biggest bank for its own test that I could find on any prep site, every item with a written explanation covering the trap as well as the answer. Access runs thirty days from a single payment, and if you prepare and do not pass, the Pass Guarantee extends your access by 30 days rather than returning your money.</p><p>For free options that cost nothing and are genuinely good: <a href="https://prepopedia.com/free-ccat-practice-test/"><strong>Prepopedia</strong></a> for the CCAT, and <a href="https://www.aptitude-test-prep.com/test-providers/pi-cognitive-assessment/"><strong>aptitude-test-prep.com</strong></a> for the PI, which publishes a complete 50-question mock with real worked explanations and no email gate. Each vendor also gives away a small official sample, and all three are worth ten minutes purely to see the real interface. <a href="https://mjunaidkhalid.com/ccat-ultimate-guide-with-practice-tests-free-resources-2025/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=cognitive-aptitude-tests-what-vendors-publish">This CCAT format and question-type guide</a> covers that test in the same detail.</p><h3>TLDR</h3><ul><li><strong>Criteria publishes the most and gets quoted the least.</strong> Mean 24.2, median 24, standard deviation 8.58, from 985 working adults across 50+ companies. Fewer than 1 percent finish all 50 questions.</li><li><strong>A raw 24 is the 50th percentile, and it is the only percentile Criteria publishes.</strong> Every other CCAT percentile you have read is a reconstruction.</li><li><strong>Criteria publishes suggested score ranges for 42 positions</strong>, with floors from 17 to 24. Software Development&#39;s floor is 20, which is far below the numbers prep sites quote for the same role.</li><li><strong>&quot;The average Wonderlic score is 20&quot; is not a Wonderlic figure.</strong> It traces through Wikipedia to a general-interest site. Wonderlic says there is no universal good or average score.</li><li><strong>Wonderlic scoring is weighted by difficulty and accounts for guessing</strong>, so the CCAT rule that a wrong answer is free does not transfer to it.</li><li><strong>The Predictive Index is the only one of the three that publishes a scratch paper policy</strong>, and it allows it. For the CCAT and the Wonderlic, that universal piece of advice is inference rather than policy.</li><li><strong>The PI gives you no percentile on purpose.</strong> It is criterion referenced: a scaled score from 100 to 450, read against a Job Target the employer set and never shows you.</li><li><strong>The two-pass strategy has no vendor support.</strong> Criteria says guess and move on. Wonderlic says do not skip around. Build a forward-only run instead and you are covered either way.</li><li><strong>The NFL stopped using the Wonderlic in 2022.</strong></li></ul><p>The most useful thing I did while writing this was read three vendor websites properly, which took an afternoon and contradicted about half of what I thought I knew. Before you trust the next score benchmark you read, including any of mine, check whether the page tells you where the number came from.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6a18028e89e3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What is a good Wonderlic score? The honest answer nobody gives you]]></title>
            <link>https://medium.com/@Elena.Rivera/what-is-a-good-wonderlic-score-the-honest-answer-nobody-gives-you-55150121ce8a?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/55150121ce8a</guid>
            <category><![CDATA[job-search]]></category>
            <category><![CDATA[aptitude-test]]></category>
            <category><![CDATA[hiring]]></category>
            <category><![CDATA[careers]]></category>
            <category><![CDATA[wonderlic]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Sun, 11 Oct 2026 20:05:53 GMT</pubDate>
            <atom:updated>2026-10-11T20:05:53.288Z</atom:updated>
            <content:encoded><![CDATA[<p>The timer stops, the screen clears, and you are left holding a two digit number with nothing to measure it against. So you do the obvious thing and search for what a good Wonderlic score is, and every page you open hands you a different answer with the same flat confidence.</p><p>The honest answer comes in three parts, and only the third is any real use to you. <strong>The famous average is not a Wonderlic figure.</strong> The vendor publishes no benchmark you could measure yourself against, and says in its own words that there is no universal good or average score. And the comparison that actually decides your outcome is one employer’s bar for one role, which nobody publishes, so the best use of the hours you have left is not hunting for a table.</p><p>I’ve sat in that chair. I took the Wonderlic as part of a hiring process, and the moment the clock ran out I was doing exactly what you are probably doing now, trying to reverse-engineer a raw number into a verdict about a job I wanted. So I searched, and what came back was a stack of pages quoting the same average, ranking occupations by required score, and not one of them citing the company that makes the test. Where I send people to spend those hours instead is <a href="https://prepclubs.com/free-practice-tests/wonderlic?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-wonderlic-score">PrepClubs</a>, whose Wonderlic bank runs <strong>400 questions across eight full-length mocks</strong>, the biggest Wonderlic question bank of any prep site I could find, the first of them free, with three worked sample questions sitting open on the page. I am about to hold a lot of pages to a sourcing standard, so hold that recommendation to it too, including the part near the end where I say plainly what its free run withholds.</p><h3>The average everyone quotes, and where it actually came from</h3><p>Run the search and you’ll meet one number over and over, stated flatly, usually in the opening line, usually with nothing attached to it. It’s worth following that citation trail all the way down, because the bottom is not where you would expect.</p><p>The claim that “the average Wonderlic score is 20” traces back through a Wikipedia citation to a general-interest website. Not a technical manual. Not a norm study. Not any document the test publisher has put its name to. From there it was copied into prep pages, then into listicles, then into the confident first sentence of nearly every article competing for the search you just ran.</p><p>The company contradicts it directly. Wonderlic says there is no universal good or average score, and that the current test does not yield a single numerical score you could line yourself up against other people with in the first place.</p><p>The occupation tables are the same story one level worse. You have almost certainly seen one: job titles down the left, a required score beside each, laid out as though a hiring committee ratified it. The vendor publishes no such table, nobody who reprints it says where it came from, and I could not find an origin either.</p><p>Here’s the uncomfortable part. Most of those pages aren’t lying on purpose. They copy each other, and nobody in the chain ever went back to the source to check.</p><h3>What Wonderlic actually publishes about its own test</h3><figure><img alt="Two columns comparing what Wonderlic publishes about its test, including 50 questions in 12 minutes, no calculator, five answer options and scoring weighted by difficulty, against what it does not publish, including any average score, any occupation benchmark table, a scratch paper policy, a retake policy, a go-back policy and a pass mark" src="https://cdn-images-1.medium.com/max/1024/1*W6axz5wAb9i4hMWU3yNwjw.png" /><figcaption>The green column is everything the vendor stands behind. Almost every number you have read about this test lives in the red one.</figcaption></figure><p>Start with what is solid, because there’s more of it than the internet suggests. The product is sold today as <strong>Wonderlic Select</strong>, formerly WonScore. It is <strong>50 questions in 12 minutes</strong>, which works out at <strong>14.4 seconds per question</strong>. No calculator. Five answer options. It is usually not proctored, and the vendor instructs candidates plainly: <strong>“Do not skip around or spend too much time on any one question.”</strong> Permission to revisit is granted only in the personality section, never in the cognitive one.</p><p>One published detail matters more than the rest, and it is almost never mentioned. <strong>Wonderlic scoring is weighted by question difficulty and accounts for guessing.</strong> Your result is not a raw count of correct answers, and that quietly kills a piece of advice you will find all over this niche: the CCAT logic that a wrong answer costs nothing, so fill in every box on the way past. That reasoning is sound on the CCAT. It does not transfer here.</p><p>Now the list that makes this page different. Wonderlic publishes no average score, no occupation benchmark table, no raw-to-percentile chart, no pass mark, no scratch paper policy and no retake policy. Six things you were probably hoping to look up, and the company that owns the instrument declines to state any of them.</p><p>Worth putting next to that: Wonderlic tells candidates <strong>“you should never have to pay”</strong> for prep material, and the NFL stopped using the test at the Combine in 2022. The first of those is a fair quote to hold against any page recommending paid prep, this one included. Better you see it here than find it later and wonder why it was left out.</p><h3>So what is your score actually being compared against?</h3><p>Your result goes to one employer, who reads it against the bar they’ve set for one specific role in one specific hiring round. Wonderlic does not publish that bar, and neither does the employer. It moves with the shape of the applicant pool that week, the seniority of the job, and how much weight that team puts on a cognitive screen at all.</p><p>So even a genuine national average would not tell you what you want to know. Sitting mid-distribution and clearing a bar are different questions, and only the second decides anything. You are not really asking whether your score is good. You are asking whether it was good enough for one person you have never met.</p><p>Nobody online can answer that. Anyone who says they can is reading from the table that came from the website that ended up in the Wikipedia citation.</p><h3>What real disclosure looks like, for contrast</h3><p>“No data” sounds like a shrug until you see the alternative sitting right next door.</p><p>Criteria Corp sells the CCAT, a direct competitor in the same pre-employment cognitive category. Here’s what it puts in public, on its product page and its sample score report.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T6SBAjvyZG4VMvgs9Y9u_g.png" /><figcaption>This is the floor of what a cognitive test vendor can tell you about your own result, and it is a floor the Wonderlic sits below.</figcaption></figure><p>A <strong>mean of 24.2</strong> and a <strong>median of 24</strong>. A <strong>standard deviation of 8.58</strong>. The norm sample behind those figures, which is <strong>985 working adults from over 50 companies</strong>. The statement that <strong>a raw 24 is the 50th percentile</strong>, which gives you one hard anchor to locate yourself against. The observation that <strong>fewer than 1% of candidates answer all 50 questions</strong>, which turns an unfinished test from a personal failure into the expected outcome. And <strong>suggested score ranges for 42 positions</strong>, from a floor of 17 for a retail sales role up to 24 for data analytics and legal work.</p><p>None of that makes the CCAT a better test. Test quality and disclosure quality are separate things. What the comparison shows is that publishing this material is entirely possible, that a competitor in the same market does it, and that the reason you cannot locate your Wonderlic result is a decision rather than a technical limit.</p><p>Even Criteria stops short of a full percentile chart and says outright that setting a cut-off is the employer’s decision. So the ceiling on disclosure is low across this whole category. The floor is not where the Wonderlic is standing.</p><h3>A worked question, because the score is built out of these</h3><p>The competing pages give you a number to feel bad about. Here’s something you can use instead: a real item of the kind that decides your result, with the trap shown rather than hidden. This one comes from the PrepClubs Wonderlic bank.</p><blockquote><strong><em>Which is smallest?</em></strong><em> 0.333, 0.3, 0.33, 0.303, 0.311</em></blockquote><p>Most people read the five decimals, feel that the shortest one is somehow the smallest, hesitate, then start comparing digits in an order they are inventing as they go. On a 14.4-second budget, that hesitation is the whole cost.</p><p>The method removes it. <strong>Pad every number with zeros until they are all the same length</strong>: 0.330, 0.303, 0.333, 0.300, 0.311. Now compare them as whole numbers, 330 against 303 against 333 against 300 against 311. The smallest is 300, so the answer is <strong>0.3</strong>.</p><p>The trap answer is <strong>0.303</strong>, and it is well built. The two zeros catch the eye and make the number look tiny, and there is a quiet instinct that the shortest option cannot be the smallest one. Both pull the same way, which is what makes the distractor work.</p><p>Notice what that took. Not arithmetic ability, one fixed procedure applied without deliberating. A second item from the same bank: a circle with a <strong>diameter of 14 cm</strong> and pi given as <strong>22/7</strong> has a circumference of <strong>44 cm</strong>, because you cancel the 7 before multiplying. The trap is 154, which is the area in square centimetres. It’s a perfectly correct calculation answering a question nobody asked.</p><p>That is what the test measures in practice. Whether you have a routine, or whether you are improvising under a clock.</p><p>One more shape worth recognising, because it turns up on the Wonderlic, the CCAT and the PI alike.</p><figure><img alt="A live 50 question practice screen with 11 minutes 34 seconds left, stating that Mia has watched 7 movies which is 2 more than Ivan, and that Ivan has watched more than Taylor, then asking whether the statement Taylor has watched fewer than 5 movies is true, false or uncertain" src="https://cdn-images-1.medium.com/max/1024/1*FRAXmAKQ1pf9hyLRiCmajg.png" /><figcaption>From the PrepClubs bank, shown in the CCAT interface, though the item type appears on all three tests.</figcaption></figure><p>Mia at 7 puts Ivan at <strong>5</strong>. Ivan beats Taylor, so Taylor sits at <strong>4 or fewer</strong>, which makes “fewer than 5” <strong>True</strong>. The trap is Uncertain, and it catches people the moment a chain has three names in it, because uncertainty starts to feel like the careful answer. Worth knowing that the careful-sounding option is a distractor the item writers plant on purpose.</p><h3>What to actually do with the number you have</h3><p>Four things, none of which need a benchmark to be useful.</p><p><strong>Stop treating an unfinished test as a bad test.</strong> Wonderlic does not publish how many questions candidates typically reach. Criteria does publish it for the CCAT, where fewer than 1% of people answer all 50, and the CCAT is the more generous of the two at 18.0 seconds a question against 14.4. Treat that as directional rather than as a Wonderlic statistic, but it does suggest that finishing was never the design intent.</p><figure><img alt="Three cards comparing pace: the CCAT gives 18.0 seconds per question across 50 questions in 15 minutes, the Wonderlic gives 14.4 seconds across 50 questions in 12 minutes, and the PI Cognitive Assessment also gives 14.4 seconds" src="https://cdn-images-1.medium.com/max/1024/1*IdaKOrINcS1r1TsI6czwWw.png" /><figcaption>Fourteen point four seconds covers reading the question, reading five options, deciding, and clicking. Not fourteen seconds of thinking.</figcaption></figure><p><strong>If you’re asked to retake it, ask the employer what the vendor will not tell you.</strong> There’s no published retake policy, so whether the second score replaces the first, and how long you have to sit it, are facts only the person running your process holds. Ask them directly. It’s a normal thing to ask, and the answer changes how you spend the next few days.</p><p><strong>Do not build a plan around returning to skipped questions.</strong> The vendor tells you not to skip around and grants revisit rights only in the personality section. Answer in order, keep the queue moving, and never sit staring at one item while the questions behind it go unread.</p><p><strong>Then spend the rest of your time on pacing, because that’s the only variable you control.</strong> Three days of practice will not meaningfully raise your processing speed, and no page should tell you otherwise. What it reliably does is stop you leaking points you could already earn: no freezing on an unfamiliar item, no inventing a method mid-question, no arriving at question 30 with four minutes left.</p><p><strong>“You end up recommending a paid product, so why trust your version?”</strong> Fair. Everything above is either a direct quote from a test vendor or an admission that a number you wanted does not exist, and the parts that make the product I recommend look worse are still on the page.</p><h3>Where I would practise for this</h3><p>You want full-length, timed, fifty-question runs, because the clock is the real difficulty and a page of sample questions cannot reproduce it.</p><p><a href="https://prepclubs.com/free-practice-tests/wonderlic?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-wonderlic-score">PrepClubs</a> is built for the short window this article is aimed at. <strong>Three sample Wonderlic questions sit open on the page with the full walkthroughs</strong>, which is how you judge the explanations before spending anything rather than a substitute for a timed run. Then <strong>the first full-length test is free</strong>, one across the platform rather than one per exam, and the whole bank sits on the paid tier: <strong>400 questions across eight full-length mocks</strong>, the biggest Wonderlic bank of any prep site I could find, every item carrying a written explanation of the trap as well as the answer, in the style of the ones worked above.</p><p>And the limitation, since this article is about people quoting numbers they cannot support. <strong>The free run gives you your score and where you sit against other people who took the same test. It does not give you the worked explanations, which stay on the paid tier.</strong> Access runs thirty days from a single payment, and if you prepare with it and do not pass, the Pass Guarantee extends your access by 30 days rather than returning your money.</p><p>If your test turns out to be the CCAT instead, which happens more often than you’d think when the invite email names no test at all, <a href="https://ccattests.com/free-ccat-practice-tests?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-wonderlic-score">CCAT Tests</a> has five full-length simulations at the real 50 questions in 15 minutes, and <a href="https://mjunaidkhalid.com/ccat-ultimate-guide-with-practice-tests-free-resources-2025/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-wonderlic-score">this walkthrough of the CCAT question types</a> covers each one.</p><p>If you’d rather use something free and official, Wonderlic publishes a short official sample assessment. Five cognitive questions and no answer key, so it teaches you nothing about your mistakes, but it’s worth two minutes to see the real question formats.</p><h3>Where to actually take one</h3><p>One more note on where to find a timed run, since that is the whole recommendation.</p><p>The first full-length test on PrepClubs is free, so a timed run tonight costs you nothing, and here is the limit on it, said plainly in an article about unsupported numbers. <strong>The free run gives you a score and a benchmark. It does not release the worked explanations</strong>, which stay on the paid tier. Given that this page has spent its length arguing the benchmark is the least useful thing about your score, that is a fairly honest description of what you get for nothing.</p><p>If you want free and nothing paid at all, <a href="https://boostprep.com/wonderlic-practice-test/">BoostPrep</a> runs an accurate 12-minute clock over 50 questions with no account needed. Its limitation is the same shape, only worse: I counted 101 explanation containers on its free tests and 100 render empty, so it will hand you a number and nothing to do with it.</p><h3>TLDR</h3><ul><li><strong>The average Wonderlic score you keep reading is not a Wonderlic figure.</strong> It traces through a Wikipedia citation to a general-interest website, and has been copied ever since.</li><li><strong>Wonderlic says there is no universal good or average score,</strong> and that the current test does not produce a single number for that kind of comparison.</li><li><strong>The vendor publishes no occupation table, no percentile chart, no pass mark, no retake policy and no scratch paper policy.</strong> Any page handing you one made it up or inherited it.</li><li><strong>What it does publish:</strong> Wonderlic Select, 50 questions in 12 minutes, 14.4 seconds each, no calculator, five options, and scoring weighted by difficulty that accounts for guessing. CCAT guessing advice does not carry over here.</li><li><strong>Criteria Corp publishes a mean of 24.2, a median of 24, an SD of 8.58, a 985-adult norm sample and suggested ranges for 42 positions.</strong> That is what disclosure looks like when a vendor chooses it.</li><li><strong>Your real comparison is one employer’s bar for one role,</strong> which nobody publishes, so no table online can answer your question. Practise the pacing instead. PrepClubs runs 400 Wonderlic questions across eight full-length mocks, the biggest Wonderlic bank I could find anywhere, the first one free.</li></ul><p>If you are reading this the night before, close the tab on the score tables. Take one timed run, find out which questions ate your clock, and fix that.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=55150121ce8a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[More than 100 people have asked me how to pass the CCAT. Here is everything I tell them.]]></title>
            <link>https://medium.com/@Elena.Rivera/more-than-100-people-have-asked-me-how-to-pass-the-ccat-here-is-everything-i-tell-them-16e35e978ab8?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/16e35e978ab8</guid>
            <category><![CDATA[cognitive-testing]]></category>
            <category><![CDATA[aptitude-test]]></category>
            <category><![CDATA[ccat]]></category>
            <category><![CDATA[careers]]></category>
            <category><![CDATA[job-search]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Mon, 14 Sep 2026 16:01:02 GMT</pubDate>
            <atom:updated>2026-09-14T16:01:02.011Z</atom:updated>
            <content:encoded><![CDATA[<p>Over the last four years, more than a hundred people have messaged me on LinkedIn with some version of the same question. They have a <a href="https://www.criteriacorp.com/assess/cognitive-aptitude/criteria-cognitive-aptitude-test-ccat">Criteria Cognitive Aptitude Test</a> scheduled, usually in the next three days, usually standing between them and a job they actually want.</p><p>I’ve written back to almost all of them. The reply has changed a lot. The test itself has not: the Criteria Cognitive Aptitude Test is fifty questions in fifteen minutes, no calculator, five options a question, verbal reasoning and maths and spatial reasoning mixed together, and no penalty for a wrong answer. What follows is the whole of what I now tell people, in order: why pacing costs more marks than ability does, the twenty-second rule that fixes it, why you should never leave a blank, what actually helps on each of the three question types, what a good score looks like and what employers really ask for, and where to sit a timed full-length test tonight.</p><p>The first version of my answer was about getting better at the questions. That version wasn’t much use, or at least it was aimed at the wrong thing. The single biggest score change available to most people in a 72-hour window has almost nothing to do with how good they are at analogies or number series. It is <strong>practice </strong>and <strong>speed. </strong>All questions are “easy” enough that if you had no time limit, you could easily get 45+ out of 50. The <em>time restriction </em>is what makes it hard; hence speed and practice is needed. I’ve shared resources of where you can practice towards the end of the article as well.</p><p>One thing up front, because it is load-bearing. I am not a psychometrician and I have no line into Criteria Corp. I sat this test, badly at first, and what follows is what I worked out in the chair rather than off a spec sheet. Where I point people who want timed reps is <a href="https://prepclubs.com/free-practice-tests/ccat?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">PrepClubs</a>, and its terms get the same scrutiny at the end as everybody else’s.</p><h3>What the CCAT actually is</h3><p>The CCAT is <a href="https://www.criteriacorp.com/assess/cognitive-aptitude/criteria-cognitive-aptitude-test-ccat">Criteria Corp’s cognitive aptitude test</a>. Fifty questions, fifteen minutes, no calculator. It mixes verbal reasoning, maths and logic, and spatial reasoning, and it is scored on the number of correct answers with <strong>no penalty for a wrong one</strong>.</p><p>That last detail matters more than anything else in this article, and I will come back to it.</p><p>Fifteen minutes for fifty questions works out at eighteen seconds each. Not eighteen seconds of thinking. Eighteen seconds of reading the question, reading five options, deciding, and clicking.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yxn38xX5jvHDxW24bXb4SQ.png" /></figure><p>Most people do not finish. That’s by design, and it’s the first thing to make peace with, because not finishing is not what costs you marks. Blanks are.</p><h3>The arithmetic nobody does</h3><p>Here is the ninety seconds I promised.</p><p>You have 900 seconds. Suppose you hit a genuinely hard spatial question at number 7 and decide to crack it. You give it sixty seconds. It felt like twenty, because hard questions distort time in exactly one direction.</p><p>Sixty seconds instead of eighteen is forty-two seconds of overspend. Forty-two seconds is two and a third questions you will now never see. Do that four times across the test and you have spent 168 extra seconds, which is nine questions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*17iF-82AJkeQd7AqbeDD0Q.png" /></figure><p>Now price those nine questions. Five options each, no penalty for a wrong answer, so a blind guess is worth about 0.2 marks. Nine unreached questions is roughly two marks of pure score, gone, in exchange for four hard questions you probably got wrong anyway.</p><p>Two marks does not sound like much until you know where the bands sit.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Cph9qicWFw8Q0K6l_4WP8Q.png" /></figure><p>The gap between “average” and “good” is about six marks. The gap between “good” and “competitive” is about eight. If pacing errors are quietly costing you two to four marks, that is a third of the distance to the next band, and you can recover all of it without getting one bit smarter at the actual questions.</p><p>That is why my answer changed.</p><h3>The 20-second rule</h3><p>The rule is this: <strong>if you cannot see a route to the answer within about twenty seconds, guess and move on.</strong> Not the answer, the route. You don’t need to know the answer is 96. You need to know the way in is 16 times 6.</p><p>Twenty seconds is slightly over pace on purpose. It gives you a little room on the questions that deserve it while still capping your worst case.</p><p>The hard part is not knowing the rule. Everyone who reads a CCAT guide knows the rule. The hard part is that abandoning a question you’ve already sunk thirty seconds into feels like waste, so you throw good time after bad. That instinct is the thing to practise against, and it is the reason timed practice beats untimed practice by such a wide margin. Untimed practice trains your reasoning. Timed practice trains your quitting.</p><h3>Skip the two-pass plan, and here is why</h3><p>Nearly every CCAT guide online, including guides I used to trust, tells you to run two passes: sweep the easy questions first, then come back for the hard ones. I have stopped giving that advice, because I went looking for the thing it depends on and could not find it.</p><p>A review pass only works if the software lets you go back. <strong>Criteria does not publish a policy saying it does.</strong> What Criteria does tell candidates is the opposite instinct: when a question is not coming, “<a href="https://www.criteriacorp.com/candidates/ccat-prep">it’s best to guess and move on.</a>” Wonderlic, on its own test, is blunter still, <a href="https://wonderlic.com/blog/candidates/wonderlic-sample-questions/">instructing candidates not to skip around at all</a>. Neither vendor commits in writing to letting you return to a question you left.</p><p>So build a run that never needs to go back.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*v4Gt5iUvrugB1_RKgBFw9g.png" /></figure><p>The rule that replaces the second pass is simple and it is worth more than the plan it replaces: <strong>never move past a question without putting an answer in it.</strong> Not a flag, not a mental note, an actual selected option. Guess it if you have to. A blank and a wrong answer score exactly the same on the CCAT, so a blank is strictly worse, and a blank you were planning to return to is worse still if the return never comes.</p><p>If you take one thing from this article, take that one, because it costs nothing.</p><h3>The three question types, and what actually helps</h3><p>Now the part about getting better at the questions, which compounds with the pacing rather than competing with it.</p><h3>Verbal</h3><p>Analogies are the highest-volume verbal item and the easiest to speed up. The move is to <strong>name the relationship in plain words before you look at a single option</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CTtIbkK0XeNPBZsaWrb6KA.png" /></figure><p>Run the method on that one. ILLEGIBLE means something cannot be read, and reading is the kind of beholding the pair points at, so the relationship is “cannot be verbed”. Test each option against that single sentence and only <strong>INCOMPREHENSIBLE is to UNDERSTAND</strong> survives, because incomprehensible means cannot be understood. INDEFATIGABLE is to ROUSE reads learned and costs you nothing but time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nmKtINelHtAbRUOvCdiPpQ.png" /></figure><p>If you read all five options first and wait to feel which one sounds right, you will burn twenty-five seconds and still be choosing between two. Naming the relationship converts a fuzzy comparison into a yes-or-no test you can run five times quickly.</p><h3>Maths and logic</h3><p>There’s no calculator, which sounds worse than it is. The questions are built to be done without one, so if you find yourself setting up long division, you have usually missed the intended route.</p><p>Learn to build any percentage out of 10 percent, 5 percent and 1 percent, and a whole category of questions collapses.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9JyNQtyQVhcx88rt6U-P1A.png" /></figure><p>Notice the last step there. On a multiple choice test you can check a candidate answer by multiplying instead of solving by dividing, and multiplying is much faster and much safer under pressure. That habit alone is worth a few seconds on every arithmetic word problem.</p><h3>Spatial</h3><p>This is the section people find most alien, and it is also the one where practice pays back fastest, because the rule set is small and it repeats.</p><p>The core habit: <strong>track one attribute at a time</strong>. Do not try to perceive the whole pattern at once.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ug5_777twCAecZM6eQuNnA.png" /></figure><p>Most candidates lose time on matrices by staring at the grid waiting for the pattern to appear. It doesn’t appear. You have to interrogate it one axis at a time, and once you do, these become some of the fastest questions on the test rather than the slowest.</p><h3>What prep does not do</h3><p>Here’s the part the prep landing pages leave out.</p><p>Three days of practice will not meaningfully change your underlying processing speed. If you are hoping to go from struggling to comfortable in a weekend, that’s not something I’ve seen, and I am not going to tell you otherwise. What practice reliably does is stop you leaking the points you were already capable of earning: it removes the pacing errors, it kills the freeze on unfamiliar spatial items, it makes you stop leaving blanks, and it means you spend your fifteen minutes on the questions you can actually do.</p><p>In my experience that is worth several marks to most people, and several marks is often exactly the difference between the band you are in and the band the employer asked for.</p><p><strong>“Can I just use ChatGPT to prep?”</strong> People ask me this constantly now. It’s fine for explaining a concept you got wrong. It’s close to useless for the thing that actually matters here, because the whole difficulty of the CCAT is a fifteen-minute clock, and a chatbot can’t put you under one. You need a timer and a full-length test, not an explainer.</p><h3>Two questions I get every time, answered short</h3><ol><li><strong>“Is the CCAT an IQ test?”</strong> <br>No. It is a pre-employment aptitude test from Criteria Corp, built to predict performance in a role. It correlates with general cognitive ability without being a clinical instrument, and the practical difference is the one that matters to you this week: the CCAT is speed-gated, usually unproctored, and scored against other applicants. That combination responds to preparation. Treating it as a fixed IQ number is the single most expensive belief you can hold right now, because it is the reason people walk in cold. If you want the longer version of that argument, the most thorough treatment of it I came across is <a href="https://mjunaidkhalid.com/2024/12/23/is-the-ccat-an-iq-test/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">this piece on whether the CCAT is an IQ test</a>.</li></ol><p><strong>“What score do I actually need?”</strong></p><p>Here is the honest version. <a href="https://www.criteriacorp.com/files/Criteria-InfoBrief-CCAT.pdf">Criteria publishes a mean of 24.2 and a median of 24, and confirms that a raw 24 is the 50th percentile</a>. That is the only raw-to-percentile figure the vendor publishes. Criteria does not publish a conversion chart and does not publish a pass mark, so every percentile table you will find online is a reconstruction. Employers set their own bar, and they do not publish it either. Crossover, the employer most people are prepping for when they land on a page like this, keeps its number inside the hiring portal rather than anywhere on its website, which is exactly why you cannot find it. So here it is second-hand from someone who was inside. Junaid Khalid spent about four years at Crossover and puts the bar at <strong>35 out of 50 for most entry-level roles</strong>, <strong>40 and above once you are looking at the six-figure ones</strong>, and <strong>43 and above as excellent</strong>. He also notes that the badge is valid for six months, so one sitting covers a run of applications rather than one. That is one person’s account rather than a published figure, and I would rather say so than dress it up with a source it does not have. Treat Criteria’s 24 as the one solid published number and everything else as directional. His fuller breakdowns of <a href="https://mjunaidkhalid.com/2024/09/08/what-is-a-good-ccat-score-understanding-your-results/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">what a good CCAT score looks like</a> and <a href="https://mjunaidkhalid.com/2026/06/10/ccat-score-percentiles-explained-what-your-score-means-and-what-recruiters-see/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">how the percentiles are usually reported</a> are the most complete reconstructions I found anywhere, and he says himself that reconstructions are all they are. Treat every number in them as directional.</p><h3>If your test is in 72 hours</h3><p>The plan I send people, in order:</p><ol><li><strong>Take one full-length, timed test today.</strong> Fifteen minutes, no calculator, no pausing, scratch paper only. You can do it for free here: <a href="https://prepclubs.com/free-practice-tests?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">Aptitude test practice</a>. The score matters far less than the map of where your time went.</li><li><strong>Read every question you got wrong and sort the reasons into four buckets:</strong> misread it, knew it but was too slow, did not know the concept, never reached it. Those four have completely different fixes, and lumping them together is why people plateau.</li><li><strong>Drill only your worst category.</strong> If bucket three is mostly spatial, do spatial until the rule set stops surprising you.</li><li><strong>Take a second full-length test the day before.</strong> This one is about pacing, not learning. Practise quitting at twenty seconds.</li><li><strong>On the day: scratch paper ready, tabs closed, and never leave a blank.</strong></li></ol><h3>Where to actually practise</h3><p>You need full-length, timed, fifty-question tests with worked explanations. Sample questions scattered through a blog post won’t do it, because they don’t put you under the clock, and the clock is the test.</p><p><a href="https://prepclubs.com/free-practice-tests/ccat?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">PrepClubs</a> is built for exactly the 72-hour window this article is about.</p><p>The honest terms: <strong>you get one free full-length CCAT Test, your pick of cluster, and you do need to sign up.</strong> They also happen to have the <em>biggest practice tests and practice questions bank </em>across all platforms. But that is of course behind a paywall.</p><p>One more thing before you click: <strong>the free run gives you your score and where you sit, not the worked explanations.</strong> Those sit behind the paywall, along with the rest of the bank.</p><p>That is the same trade almost every free tool in this category makes, and you may as well know it going in. The paid bank has 750 full-length CCAT questions plus 600 topical drills, each with a written explanation covering the trap as well as the answer. It also tells you which segments/question types you got wrong the most (e.g. Mathematical reasoning) and also tells you which areas you spent most time in i.e. what question types take you the longest to answer.</p><p>If you prepare with it and do not pass, the Pass Guarantee extends your access by 30 days. That is an extension of access rather than a refund, and it is worth knowing which one you are buying before you buy it.</p><p><a href="https://ccattests.com/free-ccat-practice-tests?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">CCAT Tests</a> is the CCAT-only option, with five full-length simulations at the real fifty questions in fifteen minutes, plus topical drills and analytics on which question types are costing you. And if you want the long-form version of everything above, this <a href="https://mjunaidkhalid.com/ccat-ultimate-guide-with-practice-tests-free-resources-2025/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">CCAT ultimate guide</a> goes deeper on each question type than anything else I read.</p><p>If you would rather not pay for anything, free ungated full-length tests do exist. The two I checked were Prepopedia, which publishes a full 50-question test with no signup and explains its answers, and PrepTerminal, which advertises a free ungated full-length simulation with solution review at the end. Criteria Corp also publishes <a href="https://www.criteriacorp.com/candidates/practice-tests">three official sample questions,</a> which is too few to practise on but worth two minutes to see the real interface.</p><h3>TLDR</h3><ul><li><strong>If you do one thing tonight, take a full-length timed test. </strong><a href="https://prepclubs.com/free-practice-tests/ccat?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">PrepClubs</a> is where I would start, because it has the biggest CCAT bank of any platform I looked at and the free run puts you through a real 50-in-15 sitting and tells you where you rank. <a href="https://ccattests.com/free-ccat-practice-tests?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=100-people-asked-me-how-to-pass-the-ccat">CCAT Tests</a> is the CCAT-only alternative if five simulations is all you want.</li><li><strong>The CCAT is 50 questions in 15 minutes.</strong> Eighteen seconds each, no calculator, and no penalty for a wrong answer.</li><li><strong>The biggest available score gain is pacing, not intelligence.</strong> Four questions at sixty seconds each costs you nine questions at the end of the test.</li><li><strong>Use the 20-second rule.</strong> If you cannot see the route in twenty seconds, guess and move on. Practise quitting, not just solving.</li><li><strong>Never leave a blank.</strong> A guess is worth about 0.2 marks and a blank is worth zero. There is no downside.</li><li><strong>Name the relationship before reading the options</strong> on analogies. Build percentages from 10, 5 and 1. Read matrices one axis at a time.</li><li><strong>Prep will not raise your ceiling much in three days.</strong> It stops you leaking the points you can already earn, which is usually several marks, which is usually the band you needed.</li><li><strong>Take one timed full-length test today, not more sample questions.</strong></li></ul><p>The hundred-and-somethingth person to message me will get a link to this page instead of a long reply. If that is you: take a timed one today, and quit at twenty seconds.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=16e35e978ab8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Wonderlic test: 50 questions in 12 minutes, explained properly]]></title>
            <link>https://medium.com/@Elena.Rivera/the-wonderlic-test-50-questions-in-12-minutes-explained-properly-5985a35c7696?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/5985a35c7696</guid>
            <category><![CDATA[cognitive-aptitude]]></category>
            <category><![CDATA[careers]]></category>
            <category><![CDATA[wonderlic]]></category>
            <category><![CDATA[wonderlic-test]]></category>
            <category><![CDATA[aptitude-test]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Thu, 10 Sep 2026 23:06:01 GMT</pubDate>
            <atom:updated>2026-09-10T23:06:01.544Z</atom:updated>
            <content:encoded><![CDATA[<p>Fourteen point four seconds a question, five answer options, no calculator. Here is what Wonderlic publishes about its own test, and the six things it does not. I have also shared resources you can use to prepare or practice the test towards the second half of the article.</p><p>An employer sends you a link and calls it a short assessment. You look it up, and within about four clicks you have read that the average score is 20, that some quarterback once scored 48, and that you should definitely buy somebody’s course.</p><p>I sat the Wonderlic as part of a hiring process. When the timer stopped I did exactly what you are probably doing now: stared at a number with no context and started searching, and got a different answer on every page. The test is sold today as Wonderlic Select, and the cognitive part of it is fifty questions in twelve minutes with no calculator and five options a question, which works out at 14.4 seconds each and roughly twenty percent less time per question than the CCAT gives you.</p><p>So here is the version I wanted, in order: what the test actually is and what it is called now, the four question types including the attention-to-detail items nobody expects, why the famous average score of 20 is not a Wonderlic figure, how the scoring actually works and why the CCAT rule about free wrong answers does not carry across, the six things the vendor declines to publish at all, a forward-only run you can hold to under the clock, and where to sit a timed full-length test tonight. Where I point people who want timed reps is <a href="https://prepclubs.com/free-practice-tests/wonderlic?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=wonderlic-test-explained">PrepClubs</a>, and its terms get the same scrutiny at the end as everybody else’s. I’ve tried to include the parts most pages leave out.</p><h3>What the Wonderlic is now</h3><p>First, the name, because it has changed and most pages have not noticed. The product is sold today as <strong>Wonderlic Select</strong>, and before that it was WonScore. If you are reading a page that calls it the “Wonderlic Personnel Test,” you’re reading something written a while ago and not updated since.</p><p>The cognitive part is <strong>50 questions in 12 minutes</strong>. No calculator. Five answer options per question. It covers verbal reasoning, maths, logic, and a fourth category most candidates never see coming, which I will get to.</p><p>Twelve minutes for fifty questions is <strong>14.4 seconds each</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yxn38xX5jvHDxW24bXb4SQ.png" /><figcaption>The Wonderlic gives you 20 percent less time per question than the CCAT, which is the test people usually compare it to.</figcaption></figure><p>Fourteen seconds has to cover reading the question, reading five options, deciding, and clicking. Very little of it is thinking. On the questions you find easy you’ll beat it comfortably. On the rest you won’t come close, and the entire test is really about how quickly you can tell those two groups apart.</p><h3>What Wonderlic publishes, and what it does not</h3><p>This is the part that changes how you should read every other page on this subject, so I want to put it early rather than bury it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*komCnrFTsiYbyNLm9vm0QQ.png" /></figure><p>Two things in there deserve their own paragraphs.</p><h3>The average score of 20 is not a Wonderlic number</h3><p>You will read it everywhere. It is on the first page of results, it is <a href="https://en.wikipedia.org/wiki/Wonderlic_test">in the Wikipedia article</a>, and it gets repeated as though it came off a technical manual.</p><p>Follow the citation and it does not lead to Wonderlic. It leads through a Wikipedia reference to a general-interest website, and that’s where it stops. Wonderlic’s own position is that there is <strong>no universal good or average score</strong>, and that the current test does not produce a single numerical score to compare in that way at all.</p><p>That’s genuinely useful to know, because if you finished your test and are hunting for the number that tells you whether you passed, you’re hunting for something that doesn’t exist. What exists is an employer, a role, and a bar that employer set for that role, and none of the three are published.</p><p>Compare that with Criteria Corp, who publish the CCAT. Criteria publish a mean of 24.2, a median of 24, a standard deviation of 8.58, the fact that their norms rest on a sample of 985 working adults across more than 50 companies, the fact that a raw 24 sits at the 50th percentile, that fewer than 1 percent of candidates answer all 50 questions, and suggested score ranges for 42 different positions. That is what disclosure looks like. Wonderlic publishes none of the equivalent.</p><h3>The scoring is not a simple count</h3><p>This one trips up people who prepared for a different test first. The CCAT is scored on the raw number you got right, with no penalty for a wrong answer, which is why every CCAT guide tells you to fill in every box.</p><p><strong>The Wonderlic does not work that way.</strong> <br>What Wonderlic publishes is that the test is <a href="https://wonderlic.com/blog/candidates/wonderlic-select-test-frequently-asked-questions-from-candidates/">scored by a proprietary algorithm</a> that weighs all of your responses against the demands of the specific job. That is not a raw count, and it is not something anyone outside the company can reverse-engineer. So the clean CCAT logic, that a wrong answer is free and therefore you should answer everything blind at the end, is not something you can assume carries across, and anyone telling you it does is describing the wrong test. Wonderlic does not publish enough detail for me or anyone else to give you exact guessing arithmetic here, and I would rather say that than invent a rule.</p><p>What still holds: answer the questions you reach, deliberately, and do not leave a question you have actually read sitting empty while you go looking for an easier one.</p><h3>The four question types</h3><h3>Verbal</h3><p>Mostly vocabulary and word relationships, and it moves fast. A representative item from the PrepClubs question bank:</p><blockquote><strong><em>SAGACITY most nearly means:</em></strong><em> Weakness | Anger | Speed | Fear | Wisdom</em></blockquote><p>The answer is wisdom. The way to get there in four seconds rather than fifteen is to notice that the clue is sitting at the front of the word: <strong>sage</strong> means a wise person, and you already know that one. Weakness is the trap, and it catches people who guess from the soft sound of the word rather than from its root.</p><p>That is the whole verbal strategy in one item. When a long word blocks you, look for a shorter word hiding inside it before you start weighing the options against each other. If nothing is hiding in there, guess from whichever option is in the right topic area and move.</p><p>The other verbal shape is sentence completion, where the sentence hands you the answer if you read the whole of it before looking down.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Zq0owJBQcErUvtVKPDEyCA.png" /></figure><p>The sentence pairs the blank with “dedication”, and both words describe an athlete coming back from a serious injury and inspiring her teammates. Two words in the same register, both positive, both about effort sustained over time. <strong>Determination</strong> is the only option that fits the slot. Insubordination and Elation point the wrong way entirely, and Perfection sits in the right register with the wrong meaning, which is what makes it the distractor worth two seconds of hesitation.</p><h3>Maths</h3><p>No calculator, which sounds worse than it is, because the arithmetic is built to be done without one.</p><blockquote><strong><em>A number is tripled and then 4 is added. The result is 25. What is the original number?</em></strong><em> 6 | 7 | 5 | 9 | 8</em></blockquote><p>Work backwards, undoing the last step first. The result is 25 and 4 was added last, so before that it was 21. That 21 came from tripling, so divide by 3 and you have 7. Check it forwards: 7 tripled is 21, plus 4 is 25.</p><p>Notice that the wrong answers are not random. Option 5 catches somebody who takes the 4 off correctly and then divides 21 by 4 instead of by 3. That is what a well-built distractor looks like, and it is why “the answer that feels about right” is a dangerous instinct on a timed test.</p><h3>Logic</h3><p>Short deductive items, usually two premises and a conclusion to evaluate. These reward reading precisely rather than reasoning cleverly, and the common failure is importing knowledge from the real world into a question that only wants you to work inside the premises given.</p><h3>Attention to detail</h3><p>Here is the category nobody expects, and the one that separates the Wonderlic from the CCAT more than anything else.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-28Zq0_ufyQI8IdqsSDTag.png" /></figure><p>The answer is 2. The reason this is hard under a clock is that it asks you to stop reasoning and start proofreading, then switch back for the next question. The errors planted in these items are specifically the ones a normal reader glides over: two adjacent characters swapped, or one character replaced.</p><p>The technique is to compare position by position and refuse to read the strings as whole numbers. If you read 1845317 and 1485317 as numbers, your brain helpfully tells you they are the same. If you read them as character sequences, the swap at positions two and three is obvious.</p><p><strong>It is also the clearest signal of which of the two tests you are actually looking at.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yH6nMcYn9YBF1ZT_epU86A.png" /></figure><p>Confusing the two is common enough that it is worth checking what you have been sent before you spend a weekend on the wrong thing.</p><h3>Strategy: plan a run that never goes backwards</h3><p>Most Wonderlic advice online, including advice I used to repeat, tells you to sweep the easy questions first and come back for the hard ones. I have stopped saying that, because I went looking for the thing that plan depends on and could not find it.</p><p>A review pass only works if the software lets you return to a skipped question. Wonderlic does not publish that it does. What Wonderlic does publish, in its own test instructions, is the opposite instruction: <strong>“</strong><a href="https://wonderlic.com/wp-content/uploads/2024/07/07.11.24-Select-Sample-Assessment-Downloadable.pdf"><strong>Do not skip around or spend too much time on any one question</strong></a><strong>.”</strong> The only place the test grants explicit permission to go back and review your answers is the personality section, never the cognitive one.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*v4Gt5iUvrugB1_RKgBFw9g.png" /></figure><p>So the practical run is:</p><ol><li><strong>Read the question and decide fast whether you can see a route to the answer.</strong> Not the answer itself, the route. On the Wonderlic that decision needs to land inside about 16 seconds, tighter than the 20 you would allow yourself on the CCAT.</li><li><strong>Put an answer in.</strong> Every time, before you move on.</li><li><strong>Advance, and do not plan to come back.</strong></li></ol><p>The hard part isn’t knowing this. It’s that walking away from a question you’ve already spent twenty seconds on feels like waste, so you throw good time after bad. That instinct is what timed practice actually trains, and it is why untimed practice is close to useless here. Untimed practice builds confidence that evaporates the moment the clock starts.</p><p>The moment that stuck with me from my own sitting was hitting a word problem I knew I could solve, working out that it would cost me around 40 seconds, and making myself skip it. Forty seconds is three or four easier questions. You are not really being asked whether you can solve this one. You are being asked whether you can tell, in two seconds, which ones to abandon.</p><h3>The questions everyone asks</h3><p><strong>Can I use a calculator?</strong> No. <a href="https://wonderlic.com/wp-content/uploads/2024/07/07.11.24-Select-Sample-Assessment-Downloadable.pdf">Wonderlic’s own sample assessment</a> says so in the instructions.</p><p><strong>Is scratch paper allowed?</strong> Wonderlic does not publish a policy on it, and neither does Criteria for the CCAT. Interestingly, the Predictive Index is the only one of the three big vendors that addresses scratch paper in writing, and it allows it. Anyone telling you confidently what the Wonderlic permits is guessing.</p><p><strong>Is it proctored?</strong> Usually not.</p><p><strong>Can I retake it?</strong> Wonderlic does not publish a retake policy. That is the employer’s call.</p><p><strong>Didn’t the NFL use this?</strong> It did, for decades, at the Combine. <a href="https://www.si.com/nfl/2022/01/05/nfl-eliminate-wonderlic-test-combine"><strong>The NFL stopped using the Wonderlic in 2022</strong></a><strong>.</strong> If you found this page through a quarterback’s score, that is where a lot of the internet’s Wonderlic content comes from, and it is why so much of it is about football rather than about hiring.</p><h3>The honest limits of preparing</h3><p>An article that ends by recommending paid prep should be straight about what prep can and cannot do, rather than quietly leave that out.</p><p>It is also worth noting that vendors have a structural reason to discourage prep. A timed cognitive test’s validity depends partly on candidates not having drilled it, so discouraging preparation serves the instrument regardless of whether preparation works.</p><p>Here is where I land. A few days of practice will not meaningfully change how fast you process information. What it reliably does is stop you leaking the marks you were already capable of earning: it removes the pacing errors, it means the attention-to-detail items do not surprise you, and it means you spend twelve minutes on the questions you can actually do rather than on three you cannot. That is a smaller claim than most prep sites make. It is the one I can stand behind.</p><h3>What to do if your test is soon</h3><ol><li><strong>Take one full-length, timed test today.</strong> Twelve minutes, no calculator, no pausing. The score matters far less than the map of where your time went.</li><li><strong>Sort your wrong answers into four buckets:</strong> misread it, knew it but was too slow, did not know the concept, never reached it. Those four have completely different fixes, and lumping them together is why people plateau.</li><li><strong>Drill your worst category only.</strong></li><li><strong>Take a second timed test the day before</strong>, purely to practise deciding fast.</li></ol><p>For the practice itself, be picky about one thing: it has to be full-length and timed. Untimed sample questions are the most common thing I see sold as Wonderlic practice, and they train the one part of this test that was never the problem.</p><p>If you would rather not pay, free ungated full-length simulations do exist, and the two I checked were <strong>BoostPrep</strong> and <strong>wonderlictestpractice.com</strong>. Both put you on a real 12-minute clock without a signup, which is the part that matters most. Go in knowing the trade: neither one explains a single answer. Every explanation slot on both sites renders empty, so you will finish with a number and no idea which habits produced it. The only free Wonderlic option I found that actually explains its reasoning is testprepreview, and it gives you five questions before an email wall, counts its timer upwards rather than down, and is testing the Wonderlic Basic Skills Test, which is a different instrument from the one you were probably sent.</p><p><a href="https://prepclubs.com/free-practice-tests/wonderlic?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=wonderlic-test-explained">PrepClubs</a> is the one I point people at, and the honest terms are these: <strong>you get one free full-length test, your pick of cluster, and you do need to sign up which is also free.</strong> One free test across the whole platform, not one per exam, and better to know that here than at the signup screen. I have spent this article on what Wonderlic declines to publish, so here is what PrepClubs declines to give away: <strong>the free PrepClubs run reports your score and where you sit, and does not release the worked explanations.</strong> They are behind the paid tier, which is the same trade BoostPrep and wonderlictestpractice make, except that they never charge you. What the paid tier buys is a 400-question Wonderlic bank where every item has a written explanation covering the trap as well as the answer. If you prepare with it and do not pass, the Pass Guarantee extends your access by 30 days. That is an extension of access rather than a refund, and you should know which one you are buying. There is also a longer <a href="https://mjunaidkhalid.com/ccat-ultimate-guide-with-practice-tests-free-resources-2025/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=wonderlic-test-explained">guide to cognitive aptitude tests</a> worth a look, and <a href="https://ccattests.com/free-ccat-practice-tests?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=wonderlic-test-explained">five free full-length CCAT simulations</a> if it turns out the CCAT is the test you were actually sent.</p><h3>TLDR</h3><ul><li><strong>If you do one thing tonight, take a full-length timed test. <br></strong><a href="https://prepclubs.com/free-practice-tests/wonderlic?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=wonderlic-50-questions-12-minutes">PrepClubs</a> is where I would start, because its Wonderlic bank runs to 400 questions with a written explanation on every one, and the free run puts you on a real 12-minute clock and tells you where you sit. BoostPrep and wonderlictestpractice are the alternatives, but neither explains an answer.</li><li><strong>The Wonderlic is 50 questions in 12 minutes.</strong> <br>That is 14.4 seconds each, about 20 percent tighter than the CCAT. It is <a href="https://wonderlic.com/select-hiring-assessment/cognitive/">sold today as Wonderlic Select</a>.</li><li><strong>No calculator.</strong> <br>Five answer options. Verbal, maths, logic, and attention to detail.</li><li>The famous “average score of 20” is not a Wonderlic figure. It traces through a Wikipedia citation to a general-interest website. Wonderlic says <a href="https://wonderlic.com/blog/candidates/wonderlic-select-test-frequently-asked-questions-from-candidates/">there is no universal good or average score</a>.</li><li><strong>Scoring is a proprietary algorithm weighed against the demands of the job, not a raw count, so do not assume the CCAT rule that a wrong answer is free carries across.</strong></li><li><strong>Wonderlic publishes no average, no occupation table, no percentile chart, no pass mark, no scratch paper policy and no retake policy.</strong> Treat any page that states these confidently with suspicion.</li><li><strong>Plan a forward-only run.</strong> Wonderlic’s own instructions say not to skip around, and no go-back policy is published for the cognitive section. Decide inside about 16 seconds, answer, advance.</li><li><strong>Attention-to-detail items need proofreading, not reasoning.</strong> Compare character by character or your eye will glide over a transposition.</li><li><strong>The NFL stopped using the Wonderlic in 2022.</strong></li><li><strong>Practise full-length and timed.</strong> Untimed sample questions train the one part of this test that was never the problem.</li></ul><p>If you do only one thing tonight, set a timer for twelve minutes and take a full test. Not to see the score. To find out where your minutes go.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5985a35c7696" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[My Short Slack Messages Kept Coming Across as Rude.]]></title>
            <link>https://medium.com/@Elena.Rivera/my-short-slack-messages-kept-coming-across-as-rude-5c1f744572f3?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/5c1f744572f3</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[remote-work]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[writing]]></category>
            <category><![CDATA[communication]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Wed, 09 Sep 2026 14:44:41 GMT</pubDate>
            <atom:updated>2026-09-09T14:44:41.446Z</atom:updated>
            <content:encoded><![CDATA[<h3>My Short Slack Messages Kept Coming Across as Rude. AI Dictation Fixed It Without Making Me Write More</h3><figure><img alt="A typographic title card reading “I wasn’t annoyed. I was busy.” in large type on a plain background." src="https://cdn-images-1.medium.com/max/1024/1*v_362sze38qY45t8OGe4Hg.jpeg" /></figure><p>Three weeks ago, at the end of a one to one, somebody I work with told me he could usually tell what mood I was in from my Slack messages.</p><p>I asked what mood I had been in that week.</p><p>“<strong><em>Annoyed</em></strong>”, he said. “<strong><em>You’ve been annoyed since last week.</em></strong>”</p><p>I had not been annoyed at any point that week. I had been buried in a launch, answering everything as quickly as possible.</p><p>What had been quietly holding the line before that week, without my ever noticing, was AI dictation.</p><p>I picked it up in August for a reason that had nothing to do with tone.</p><p>Launch week was the week I stopped using it, because I spent five days in back to back calls in rooms with other people in them, and typing was the only option I had left.</p><h3>Busy And Annoyed Look The Same In Four Words</h3><p>The reason this happens is mechanical rather than emotional, which is why being a nice person does not protect you from it.</p><p><strong>Typing is slow.</strong> When you are behind, <strong>you compress</strong>. And the first thing compression removes is the part of the message that was carrying the tone.</p><p>Watch what actually disappears when you are in a hurry:</p><ul><li><strong>The greeting.</strong> Two words, cut first.</li><li><strong>The reason.</strong> Why you cannot do the thing, or why you need it by Monday.</li><li><strong>The timing.</strong> When you will come back to them.</li><li><strong>The acknowledgement. </strong>The half sentence where you register that their thing also matters.</li></ul><p>None of that is decoration. That is the entire social payload of a work message, and it is also the most expensive part to type, so it goes first every single time.</p><p>What survives compression is the <strong>instruction</strong>. <strong>“Can you send the file.” “Not this week.” “Fine.”</strong></p><blockquote>A short message is<strong> not a rude message</strong>. It is a message somebody wrote <strong>while</strong> <strong>doing something else.</strong> From the outside those two are <strong>indistinguishable</strong>.</blockquote><p>The other half of the problem is that <strong>nobody </strong>tells you.</p><p>My colleague only said it because we had reached the end of a one to one and run out of agenda.</p><p>Somebody had actually tried once, weeks earlier, and I walked straight past it.</p><p>A designer replied to a two word message of mine with <em>“sorry, have I done something?”</em></p><p>I said no, no, all good, and we both moved on. I did not think about it again until I was sitting there counting.</p><p><strong>Nobody books a meeting to inform a person that their punctuation has been frightening the team.</strong></p><h3>This Is Measurable, And It Is Bad</h3><p>I went looking for research on this afterwards, mostly hoping to find out it was not just me.</p><p>In 2005, Justin Kruger, Nicholas Epley, Jason Parker and Zhi-Wen Ng published <a href="https://web-docs.stern.nyu.edu/pa/kruger_email_ego.pdf">a study on how well tone survives email </a>in the Journal of Personality and Social Psychology.</p><p>In one of their experiments, people who <strong>listened </strong>to a statement <strong>decoded its intended tone</strong> nearly three quarters of the time.</p><p>People who read the same statement<strong> as an email</strong> were at chance. Fifty percent. A coin flip.</p><p>The senders, meanwhile, were <strong>confident </strong>they were being understood.</p><p>It gets more specific than that. A team at Binghamton University led by Danielle Gunraj, published in Computers in Human Behavior.</p><p>They gave 126 students short exchanges ending in <strong>one word replies</strong>, half with a period and half without.</p><p>The ones ending in a period were rated as <strong>less sincere</strong>.</p><p>So “<em>Okay.</em>” is a worse message than “<em>Okay</em>”, and the entire difference is a single dot you did not spend a moment thinking about.</p><blockquote><strong>Lesson:</strong> You are not being judged on your intent. You are being judged on the shortest, cheapest, most compressed version of yourself, which is the version you produce when you are busiest.</blockquote><h3>What I Found When I Scrolled Back Through A Month</h3><p>After that one to one I went back through a month of my own sent messages and counted, by hand, because I did not trust my memory of my own tone.</p><p>Separating the two piles was easier than it sounds.</p><p>Everything I had spoken has<strong> a recording sitting in my dictation app’s history</strong> with a timestamp on it, so I matched those against my sent column and sorted the month that way.</p><p>Two groups came out of it. The ones I had typed ran to a median of eleven words. The ones I had spoken ran to thirty eight.</p><figure><img alt="A two bar chart comparing the median length of typed work messages at eleven words against spoken messages at thirty eight words." src="https://cdn-images-1.medium.com/max/1024/1*fZBn7nIfWJiqb03UFzbhyA.jpeg" /><figcaption>Same person, same week, same colleagues. Three and a half times the words.</figcaption></figure><p>The part that surprised me was that the spoken ones <strong>were not more polite.</strong></p><p>I checked<strong> for that specifically</strong>, because I assumed I had been unconsciously softening things when I talked.</p><p>I was not saying please more often. I was not adding exclamation marks or warm sign offs.</p><p>I was giving reasons.</p><figure><img alt="A two panel comparison showing a typed message reading “Can’t this week.” beside a spoken version giving the same answer with a reason and a date." src="https://cdn-images-1.medium.com/max/1024/1*wMNFMV09R6hHaBgk3uUDxg.jpeg" /><figcaption>Same answer, same person, eleven seconds apart in effort.</figcaption></figure><p>A reason is the <strong>cheapest warmth there is</strong>. It costs <strong>one </strong>clause.</p><p>It is also the exact clause you delete when you are typing at <strong>speed</strong>, because at speed a clause <strong>feels like a luxury</strong>.</p><p>Out loud, it is not a luxury. It is just how sentences come out.</p><h3>The Thing I Had Been Doing For A Completely Different Reason</h3><p>I had started speaking my messages a few weeks before any of this, for reasons that had nothing to do with tone.</p><p>I wrote about that at the time: I had a folder full of <a href="https://medium.com/@Elena.Rivera/forty-one-of-my-voice-notes-last-month-were-emails-i-never-sent-d5178fafbeb0">voice notes that were really unsent emails</a>, and the fix was to stop recording notes to myself and start speaking things directly into the field they were going to.</p><p>That was a<strong> speed decision</strong>. I wanted the minutes back.</p><p>The tone thing was a <strong>side effect I did not notice at all</strong>, and only diagnosed weeks later, after somebody told me I had been annoyed since Tuesday.</p><p>We tried a few tools across the team before landing anywhere.</p><p>The one that stuck for us was <a href="https://contextli.com/use-cases">Contextli, </a>mostly because of one design decision: you pick the format before you speak rather than cleaning up afterwards.</p><p>The first time it properly landed for me I was in a Jira comment box, saying one long shapeless sentence about why a ticket was stuck.</p><p>A tidy three line comment appeared in the field with my cursor still sitting at the end of it.</p><p>I had <strong>not switched windows or opened anything.</strong> That was the entire pitch, as far as I was concerned.</p><p>One thing worth being clear about, because it confused me at first: the mode <strong>changes the register</strong>, not what <strong>you actually said</strong>.</p><p>Choosing Slack makes it sound like Slack.</p><figure><img alt="The Contextli context selection showing different writing modes, with the Slack mode selected." src="https://cdn-images-1.medium.com/max/1024/1*Un561G8kaiISGwBGTmpWzA.png" /><figcaption>The whole trick is choosing this before you open your mouth, not after.</figcaption></figure><p>One thing I should say plainly, because I have read enough breathless tool posts to know exactly what they leave out.</p><p><strong>I will not do this with other people in the room.</strong></p><p>Their FAQ says you can speak quietly and a close range microphone will pick you up, and I believe it, <strong>but I still will not do it</strong>.</p><p>Saying a client’s budget out loud<strong> three feet from somebody else’s desk </strong>is not a microphone problem.</p><p>If you work in a shared space, be honest with yourself about how much of this you would really use.</p><p>I have a door.<strong> That is a privilege</strong>, and it is doing a lot of quiet work in this article.</p><p>Other things exist and are good. Wispr Flow is the one most people I know use.</p><p>If you have never tried any of this at all,<strong> macOS and Windows both ship dictation for free</strong>. Starting there costs you nothing but the ten minutes it takes to find the setting.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FZaKcIYMjNYNf4lEuC7%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2Fsnl-saturday-night-live-season-45-ZaKcIYMjNYNf4lEuC7&amp;image=https%3A%2F%2Fmedia2.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExZW9lc2VlZWp1YTF0Nm5qOGd3NXg2M3RzcnA3dmcxd2oxZjBleW5laCZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FZaKcIYMjNYNf4lEuC7%2Fgiphy_s.gif&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="252" frameborder="0" scrolling="no"><a href="https://medium.com/media/5d5946efd8a1f018bc7a2d0d19d2c7b0/href">https://medium.com/media/5d5946efd8a1f018bc7a2d0d19d2c7b0/href</a></iframe><h3>What Speaking Actually Adds, And It Is Not Manners</h3><p>The mechanism is embarrassingly simple once you see it.</p><p>Words are <strong>expensive </strong>when you <strong>type them </strong>and nearly free when you <strong>say them</strong>.</p><p>So when you <strong>type</strong>, <strong>you spend the budget on the instruction</strong> and cut the rest.</p><p>Nobody in a corridor says “<em>not this week</em>” and walks off. You would say “<em>not this week, I’m underwater with the launch, try me Monday</em>” without deciding to, because those nine extra words cost you nothing.</p><blockquote>The reason you sound warmer out loud is not that you are a nicer person out loud. It is that the warmth was never the expensive part in speech.</blockquote><p>That is the whole finding, and <strong>it took a 2005 study and a colleague willing to say an awkward thing</strong> for me to reach it.</p><h3>Where This Does Not Work</h3><p>Two limits that I have come across, since I have already made this sound like a solved problem.</p><ol><li><strong>It does not fix a message that is actually annoyed.</strong> If you are irritated, speaking it will produce a longer, more articulate, more clearly irritated message.</li><li><strong>Some messages should be short.</strong> “<em>Yes</em>” is a complete and warm answer to <em>“shall I book it?”</em>. Padding a yes is its own kind of noise.</li></ol><p>Three weeks later, at the end of a different one to one, the same colleague told me the launch must have wrapped up.</p><p>I asked him how he could possibly know that.</p><p>He said my messages had gone back to being a normal length.</p><h3>If You Cannot Dictate, Do These Three Things Instead</h3><p>Most of the benefit here is not the technology. It is knowing what compression removes, which you can just put back by hand.</p><ol><li><strong>Add the reason.</strong> One clause, attached to the answer. “<em>Can’t today, I’m in launch reviews until four.</em>” That single clause does more work than any amount of exclamation marks.</li><li><strong>Say when.</strong> A specific time is the <strong>warmest thing </strong>you can put in a short message, because it tells the other person they have not been dropped.</li><li><strong>Drop the full stop on one word replies.</strong> I know how that sounds. The Binghamton study above is the reason, and it costs you nothing to test it for a week.</li></ol><p>None of that requires a tool, a subscription, or a room with a door.</p><p>It does require <strong>accepting that the version of you other people read all day </strong>is<strong> not the considered version</strong>. It is the version that was in a <strong>hurry</strong>.</p><p>Mine had apparently been annoyed since a week, and at least two people had been politely working around it rather than asking.</p><p>What has somebody assumed about you from a message you barely remember sending?</p><p><strong>I found the research after the fact, which is the wrong order and I know it. Those two studies are the only reason I believe this rather than simply liking it, and both are worth ten minutes even if you never dictate a word.</strong></p><p><strong>Tell me the shortest message you have ever sent that got read completely wrong.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5c1f744572f3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What is a good CCAT score? Bands, percentiles, and what Crossover actually wants]]></title>
            <link>https://medium.com/@Elena.Rivera/what-is-a-good-ccat-score-bands-percentiles-and-what-crossover-actually-wants-df8c6f1b3489?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/df8c6f1b3489</guid>
            <category><![CDATA[hiring]]></category>
            <category><![CDATA[aptitude-test]]></category>
            <category><![CDATA[ccat]]></category>
            <category><![CDATA[cognitive-aptitude-test]]></category>
            <category><![CDATA[crossover]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Sun, 06 Sep 2026 20:22:28 GMT</pubDate>
            <atom:updated>2026-09-06T20:22:28.485Z</atom:updated>
            <content:encoded><![CDATA[<p><a href="https://www.criteriacorp.com/assess/cognitive-aptitude/criteria-cognitive-aptitude-test-ccat">Criteria Corp</a> publishes exactly one percentile. Everything else you have read is a reconstruction, including the number <a href="https://www.crossover.com/">Crossover</a> actually uses, which is documented only by the people who have been through the funnel.</p><p>The test ends the moment the clock does, and you’re left holding a raw number out of fifty with no idea whether it just cost you the job. So you search for what a good CCAT score is, and the first five results hand you five different percentile tables, each stated with the same flat confidence.</p><p>I’ve been in that chair myself, twice, and neither sitting told me what my raw number was actually worth.</p><p>The most extensively documented run at this test that I could find belongs to <a href="https://mjunaidkhalid.com/ccat-ultimate-guide-with-practice-tests-free-resources-2025/">Junaid Khalid</a>, who has published his own <a href="https://mjunaidkhalid.com/2024/09/08/what-is-a-good-ccat-score-understanding-your-results/">progression</a> in public: a first practice score of 27 off a paid PDF, then 41, then 45, then 48 out of 50 on a real sitting for an employer. He has spent the four years since fielding the same question from strangers on LinkedIn, and he has written the answer up at length on <a href="https://mjunaidkhalid.com/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-ccat-score">his own sit</a>e. Most of the first-hand material below is his, or comes from other people who have been through the same funnel, cross-checked against what Criteria Corp actually publishes. The checking is mine.</p><p>One request before the numbers. Hold me to the sourcing standard I am about to hold everybody else to, and when I point you at <a href="https://prepclubs.com/">PrepClubs</a> near the end, hold that to it as well. One disclosure while we are here, since this whole piece is about where numbers come from: Junaid Khalid, whose documented work most of the first-hand material rests on, also runs PrepClubs. Weigh both accordingly.</p><h3>References</h3><ul><li><a href="https://www.criteriacorp.com/assess/cognitive-aptitude/criteria-cognitive-aptitude-test-ccat">Criteria Corp, CCAT product page:</a> the mean of 24.2, the median of 24, the standard deviation of 8.58 and the 985-person norm sample.</li><li><a href="https://www.criteriacorp.com/files/Criteria-Sample-CCAT.pdf">Criteria’s sample CCAT score report: </a>the one published raw-to-percentile point, a raw 24 at the 50th.</li><li><a href="https://www.criteriacorp.com/files/Criteria-ScoreReportGuide-CCAT.pdf">Criteria’s score report guide:</a> the suggested score ranges for 42 positions.</li><li><a href="https://www.criteriacorp.com/candidates/ccat-prep">Criteria’s candidate prep page</a>: fewer than 1% of candidates answer all 50, and the advice to guess rather than leave a blank.</li><li><a href="https://mjunaidkhalid.com/ccat-ultimate-guide-with-practice-tests-free-resources-2025/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-ccat-score">Junaid Khalid, CCAT ultimate guide</a>: the most extensively documented account of the test and how to prepare for it.</li><li><a href="https://mjunaidkhalid.com/2024/09/08/what-is-a-good-ccat-score-understanding-your-results/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-ccat-score">Junaid Khalid, what is a good CCAT scor</a>e: his own run from 27 to 48, and Crossover’s six-month re-sit window.</li><li><a href="https://mjunaidkhalid.com/2026/06/10/ccat-score-percentiles-explained-what-your-score-means-and-what-recruiters-see/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-ccat-score">Junaid Khalid, CCAT score percentiles explained:</a> his percentile table, labelled an estimate rather than a measurement.</li><li><a href="https://mjunaidkhalid.com/2024/06/25/crossovers-hiring-process-what-to-expect-with-the-ccat-assessment/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-ccat-score">Junaid Khalid, Crossover’s hiring process</a>: where the CCAT sits in that funnel.</li></ul><p>The honest version has four parts. Criteria Corp publishes exactly one raw-to-percentile point, and every chart you have seen extends it by guesswork. It publishes something far more useful that almost nobody quotes. Employers set the real bar, and they do not publish it either. And there is a number one of the biggest CCAT employers in the world actually uses, which comes from neither of them but from the people who have been through that funnel.</p><h3>The single percentile Criteria publishes</h3><p>Here’s the whole of it. <strong>A raw 24 is the 50th percentile.</strong> That figure comes off Criteria’s own sample score report, and it’s the only point on the raw-to-percentile curve that the company that builds the test has ever put in public.</p><p>Around it sits a small amount of genuinely solid distribution data. The <strong>mean is 24.2</strong>, the <strong>median is 24</strong>, and the <strong>standard deviation is 8.58</strong>. Those come from Criteria’s product page. The norm group behind them is <strong>985 working adults</strong>, aged eighteen and up, drawn from <strong>over 50 companies</strong>. The test itself has been administered several million times since, but the norms still rest on those 985 people, which is worth knowing before you treat a percentile as though it were a measurement of the entire working world.</p><p>One more published number changes how the whole thing should feel. <strong>Fewer than 1% of candidates answer all 50 questions.</strong> Criteria states that on its candidate-facing page. Not finishing is the designed outcome, not evidence that you choked.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Cph9qicWFw8Q0K6l_4WP8Q.png" /></figure><p>Now the list that makes this page different from the ones above it in the results. Criteria publishes <strong>no full percentile conversion chart</strong>, <strong>no pass mark</strong>, <strong>no cut-off</strong>, no scratch paper policy and no policy on returning to a skipped question. That’s five things you probably came here to look up, and the vendor declines to state any of them.</p><p>So when a page tells you that 32 is the 64th percentile, or that 42 puts you in the top few percent, ask yourself where that came from. I went looking for the origin of several of those tables and could not find one. They get copied sideways between prep sites, gaining confidence at every hop and never gaining a source.</p><h3>The reconstruction, done in front of you</h3><p>You can get further than “nobody knows”, as long as you say out loud what you are assuming. Here is the arithmetic, with the assumption on the table.</p><p>Take the two published figures: a <strong>mean of 24.2</strong> and a <strong>standard deviation of 8.58</strong>. A raw 35 sits 10.8 marks above the mean, and 10.8 divided by 8.58 is <strong>1.26 standard deviations</strong>. If the distribution were roughly normal, 1.26 standard deviations above the mean lands somewhere close to the <strong>90th percentile</strong>.</p><p>Criteria never says the distribution is normal. That is my assumption, not theirs, and a 50-question test with a hard time gate tends to bunch up in ways a clean bell curve does not. So take that as a rough locator and nothing sturdier. <strong>Treat it as directional.</strong></p><p>What the arithmetic does prove is more useful than the number it produces. Every percentile table you have read was built exactly this way, out of two published statistics and an unstated assumption about distribution shape, by somebody who then printed the output without showing the working. I would rather show you the working and let you decide how much weight it carries.</p><h3>The score data Criteria does publish, which nobody quotes</h3><blockquote>A note on the gap: while Criteria’s floors suggest that 20 to 35 is a good score, for most employers it is not. Everyone who has documented the Crossover funnel puts the floor for its beginner-level jobs at 35/50, and Junaid Khalid’s guides put the jobs paying $100K+/year at 40 or above. Meaning you NEED to prepare properly for the test and not aim for the median.</blockquote><p>Criteria publishes <strong>suggested score ranges for 42 different positions.</strong> Every range is a floor that runs up to 50. The floors go like this:</p><ul><li><strong>24:</strong> Data Analytics and Statistics, Legal</li><li><strong>22:</strong> Analyst, Engineering, Finance Manager, Medicine and Healthcare, Science and Research, Senior and Executive Leadership</li><li><strong>21:</strong> HR Manager, IT and Computing Systems, Marketing and Product, Nursing, Project Manager</li><li><strong>20:</strong> Accounting, Aviation, Business Operations, Education, Financial Services, IT Manager, IT Support, Management, Sales Manager, Software Development</li><li><strong>19:</strong> Art and Design, Customer Service Manager, Emergency and Protective, Medical Assistance, Office and Administrative Support, Paralegal, Sales and Account Executive</li><li><strong>18:</strong> Plant Operator, Store Manager</li><li><strong>17:</strong> Customer Service Representative, Retail Sales, Sales Representative, Service Technician</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KFkWPzWDwXGlU_NLkTiEGg.png" /></figure><p>Look at where the top of that range sits. <strong>The highest suggested floor Criteria publishes is 24</strong>, for data analytics and legal work, and 24 is the median score. The lowest is 17. On the vendor’s own reckoning, an average score clears the suggested floor for every one of the 42 positions it lists.</p><p>Two cautions before you relax. These are <strong>suggested ranges on a report, not pass marks</strong>, and Criteria says plainly that setting a cut-off is the employer’s decision. And employers routinely set a bar well above the suggestion, which is the entire reason the rest of this article exists.</p><p>Still, it’s worth sitting with. The one organisation with the full dataset suggests floors between 17 and 24, and the pages telling you that anything under 35 is a failure are not citing it, because it does not support them.</p><h3>What Crossover actually wants, according to the people who have been through it</h3><p>Crossover is the employer a large share of the people reading this are prepping for. It hires remotely at volume, it puts the CCAT early in the funnel, and it is the single most common answer when somebody asks what score they need. Junaid Khalid gets the same question from the other side of the inbox: by his own count, more than a hundred people have asked him on LinkedIn how to pass it.</p><p>Here is the awkward part, and it’s the reason you haven’t found a straight answer. <strong>Crossover does not publish its cut-off.</strong> The number lives inside the hiring portal, behind a login, where candidates and staff see it in the context of a specific role. No amount of searching from the outside will surface it, which is precisely why the vacuum has filled up with invented figures.</p><p>What we have instead is the consistent account of people who have been through that funnel, the most extensively documented of them being Junaid Khalid’s. <strong>For most entry-level roles it is 35 out of 50.</strong> <strong>For the higher-paying roles, the ones at $100,000 and above, the usual bar is 40 and above.</strong> And <strong>43 and above is excellent</strong>, which in practice is the point where the number stops being a thing anyone brings up again.</p><p>There is one more piece from the same set, and it is the one that should change how you plan the next few months rather than the next few days. Crossover lets candidates re-sit roughly every six months. Your score follows you across applications inside that window, which means a weak sitting is not something you fix next weekend, and it is the strongest argument I know for treating your first attempt as the real one rather than burning it to see what the test looks like.</p><p>You cannot verify either of those from a URL, and I’m not going to dress them up with a citation they do not have. That is first-hand knowledge from people who have been through Crossover’s process, Junaid Khalid’s documented account chief among them, offered as exactly that. It’s also more than any page on this search result is willing to give you, and I would rather tell you where a number came from than launder it through a fake source.</p><p>What I will not do is extend it. I have seen tables that split Crossover by function, giving software engineers one number and growth marketing another, and tables that hand out cut-offs for McKinsey and Goldman Sachs and half the Fortune 500. Those are invented. Most of those firms do not even use the <a href="https://ccattests.com/">CCAT</a>.</p><h3>Two numbers, two different meanings</h3><p>Notice what just happened. The vendor suggests floors between 17 and 24. One large employer uses 35, and 40 once the salary crosses six figures. Both are true at the same time, and the gap between them is the thing nobody explains.</p><p>The suggested range answers “is this person capable of the work.” The employer’s bar answers a different question entirely: “how many applicants do I have, and where do I cut to get to a shortlist I can interview this week.” <strong>That second number is about supply.</strong> It moves with the applicant pool, the seniority of the role, and how much that team trusts a cognitive screen in the first place.</p><p>Which is why “what is a good CCAT score” has no single answer, and why the pages promising one are the pages you should trust least. Your score goes to one employer, who reads it against a bar set for one role in one hiring round. Junaid Khalid keeps a longer breakdown of <a href="https://mjunaidkhalid.com/2024/09/08/what-is-a-good-ccat-score-understanding-your-results/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-ccat-score">what a good CCAT score means in practice</a> and of <a href="https://mjunaidkhalid.com/2026/06/10/ccat-score-percentiles-explained-what-your-score-means-and-what-recruiters-see/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-ccat-score">how percentiles get reported</a> on his own site. Treat every number in both as directional, mine included.</p><h3>A worked question, because the score is built out of these</h3><p>Percentile hunting is a way of avoiding the only thing that moves your number. Your raw score is fifty individual decisions made at <strong>18.0 seconds each</strong>, and the difference between 27 and 35 is usually a handful of questions you could already answer, lost to method rather than to ability.</p><p>Here’s one from the PrepClubs <a href="https://prepclubs.com/free-practice-tests/ccat">CCAT Questions bank</a>, exactly as it appears on the clock.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EYkEZ4GDX4ba3p1vZ3R_7Q.png" /><figcaption>CCAT Example Question (Mathematical reasoning)</figcaption></figure><p><strong><em>12 is 40% of what number?</em></strong><em> 20, 24, 25, 30, 32</em></p><p>The slow route is to set up 12 divided by 0.4, which means dividing by a decimal under pressure, which is where people stall or fumble a place value. On an 18-second budget, that hesitation is the whole cost of the question.</p><p>The fast route is the one method worth learning by name. <strong>Add a zero to the first number, remove a zero from the second.</strong> 12 becomes 120, 40 becomes 4, and 120 divided by 4 is <strong>30</strong>. Two seconds, no decimals, no scratch paper, which matters because Criteria publishes no scratch paper policy at all.</p><p>The trap answer here is <strong>32</strong>, sitting at the bottom of the list where a rushed eye lands, and it is well built. It’s close to 30, it’s the largest option, and if you’ve half-formed the idea that you should be multiplying rather than dividing then it feels roughly right. <strong>Close enough to feel safe.</strong></p><p>Now run the arithmetic that actually matters. If you carry a method like that into eight questions instead of improvising on each one, and each one saves you fifteen seconds, you have bought back two full minutes. Two minutes at 18 seconds a question is another six or seven questions reached. That is the mechanism behind most of the gap between a 27 and a 35, and none of it is raw intelligence.</p><p><strong>“But you cannot study for an aptitude test.”</strong> You can’t raise your processing speed much in three days, and no page should tell you otherwise. What preparation reliably does is stop you leaking marks you could already earn.</p><h3>What to actually do with the number you have</h3><p>Four things, none of which require a percentile chart.</p><ol><li><strong>Find out which questions ate your clock, not how many you got wrong.</strong> A raw score tells you the total and hides the cause. The useful question is whether you lost time on the spatial items, on a single number series you refused to abandon, or on decimal arithmetic you were doing the long way.</li><li><strong>Set your target off the role, not off the internet.</strong> <br>If you are applying to Crossover for an entry-level position, 35 is the number to plan around, 40 once the salary crosses six figures, and 43 if you want the score to stop being a topic. All three come from people who have been through that funnel rather than from any page Crossover publishes, so weigh them accordingly. If you are applying somewhere else entirely, ask the recruiter what the bar is. It’s a normal question and the answer is sometimes given.</li><li><strong>Stop treating an unfinished test as a failed one.</strong> <br>Fewer than 1% of candidates reach question 50. If you got to 38 and answered them well, you’re ahead of most of the room.</li><li><strong>Fill in every blank before the clock stops.</strong> <br>The CCAT carries <strong>no penalty for a wrong answer</strong> and has <strong>five options</strong>, so a blind guess is worth <strong>0.20</strong> of a mark and a blank is worth zero. Criteria says it directly: when you are stuck, it is best to guess and move on. Ten seconds of clicking at the end is worth about two marks, which in this range is the difference between one band and the next.</li></ol><h3>Where I would practise, and what each option withholds</h3><p>Whatever you use, it has to be <strong>full-length and timed</strong>, because the clock is the entire difficulty and a page of sample questions cannot reproduce it.</p><p>If you want free and nothing paid, <a href="https://prepclubs.com/free-practice-tests">PrepClubs</a>, <a href="https://ccattests.com/">CCAT Tests</a> and <a href="https://prepopedia.com/free-ccat-practice-test/">Prepopedia</a>, are the strongest free CCAT option I found.</p><p>Fifty questions, fifteen minutes, no signup, and genuine question images. Its one structural catch is that the explanations live in a separate untimed mode rather than inside the timed run, so you get the clock or the teaching, and you switch modes to get both. <a href="https://www.prepterminal.com/ccat-practice-test">PrepTerminal</a> advertises a free ungated fifty-question simulation with full solution review at the end, which if it delivers is the most generous free tier in the category. I verified the offer on the page rather than the delivered product, so click it and judge for yourself. Ignore the improvement rates and pass rates printed around it, which I could not source and two of which contradict what Criteria publishes.</p><p><a href="https://prepclubs.com/free-practice-tests/ccat?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-ccat-score">PrepClubs</a> is built for the seventy-two hour window this article is aimed at. The honest terms, stated here rather than at the signup screen: <strong>you get one free full-length test, your pick of cluster, and signup is required.</strong> One per person across the whole platform.</p><p>And the limitation, since this whole page is about people quoting numbers they cannot support. <strong>The free run gives you your score and where you stand. It does not release the worked explanations, which sit behind the paywall.</strong> That’s the same trade Prepopedia makes with its two modes and the same one BoostPrep makes by shipping a free test with empty explanation boxes. Full access to a cluster is <strong>$49</strong>, which opens the 750-question full-length CCAT bank plus 600 topical questions, every one carrying a written explanation of the trap as well as the answer. Prepare with it and miss the bar anyway and the Pass Guarantee extends your access by 30 days rather than refunding you.</p><p>If you want more free reps first, <a href="https://ccattests.com/free-ccat-practice-tests?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=what-is-a-good-ccat-score">CCAT Tests</a> runs five full-length simulations at the real fifty questions in fifteen minutes with explanations attached, and Junaid Khalid’s <a href="https://www.udemy.com/course/ccat-practice-tests/">CCAT course on Udemy</a> covers each question type in the order the test throws them at you.</p><h3>TLDR</h3><ul><li><strong>Criteria publishes exactly one raw-to-percentile point: a raw 24 is the 50th percentile.</strong> Every fuller chart online is a reconstruction, usually an unlabelled one.</li><li><strong>What it does publish:</strong> a mean of 24.2, a median of 24, a standard deviation of 8.58, a norm sample of 985 working adults from over 50 companies, and the fact that fewer than 1% of candidates finish all 50 questions.</li><li><strong>It publishes suggested floors for 42 positions, from 17 to 24.</strong> The highest suggested floor is the median score. Those are suggestions, and Criteria says the cut-off is the employer’s call.</li><li><strong>Crossover does not publish its bar; it lives inside the hiring portal.</strong> From people who have been through it, Junaid Khalid’s documented work most of all: <strong>35 for most entry-level roles, 40 and above for the six-figure ones, and 43 and above is excellent.</strong> First-hand accounts, not a citation.</li><li><strong>Crossover re-sits run about six months apart. Same source. </strong>A bad sitting is not something you fix next weekend, so treat the first attempt as the real one.</li><li><strong>Ignore the named-employer tables.</strong> McKinsey, Goldman and the rest are invented, and most of those firms do not use this test.</li><li><strong>Guess every blank.</strong> No penalty, five options, so a guess is worth 0.20 and a blank is worth nothing.</li><li><strong>The gap between 27 and 35 is method, not ability.</strong> <br><a href="https://prepclubs.com/free-practice-tests">PrepClubs</a> gives you one free full-length test, signup required, and the explanations sit behind a paywall.</li></ul><p>If you are reading this the night before, close the percentile tables. Take one timed run, find the two question types that ate your clock, and fix those.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=df8c6f1b3489" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The best Wonderlic practice tests (Ranked)]]></title>
            <link>https://medium.com/@Elena.Rivera/the-best-wonderlic-practice-tests-ranked-986c6e4bff19?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/986c6e4bff19</guid>
            <category><![CDATA[aptitude]]></category>
            <category><![CDATA[cognitive-aptitude-test]]></category>
            <category><![CDATA[aptitude-test]]></category>
            <category><![CDATA[wonderlic]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Sun, 06 Sep 2026 01:37:02 GMT</pubDate>
            <atom:updated>2026-09-06T01:37:02.631Z</atom:updated>
            <content:encoded><![CDATA[<p>TLDR; <a href="https://prepclubs.com/free-practice-tests/wonderlic">PrepClubs Wonderlic Free Practice Test </a>comes out near the top of this ranking for me for two reasons: it was <strong>timed</strong> and the questions similarity was the highest with the actual exam questions I faced.</p><p>But in the article I have listed 5+ others that I used personally, or reviewed because someone asked me to. So feel free to skip to that part.</p><p>I sat the <a href="https://wonderlic.com/select-hiring-assessment/cognitive/">Wonderlic cognitive ability test</a> myself, and before as well as afterwards I went through the <a href="https://prepclubs.com/free-practice-tests/wonderlic">free Wonderlic practice tests</a> one at a time to work out which of them would actually have helped. <br>(Basically a lot of people around me messaged me over time asking <em>how did you prepare :) ).</em></p><p>I’ll go ahead and list all the sources I personally tried + the ones that I’ve heard about from others, etc.</p><p>I didn’t rank these off their sales pages. I opened each one, started the test, and read the configuration the quiz was actually running: how many questions it served, what the time limit was set to, and whether the promised explanation had anything inside it.</p><h3>The pattern that runs through almost every free Wonderlic CAT test</h3><p>Here’s the finding, and I have not seen another page state it.</p><blockquote><strong>The clock and the explanations almost never ship together for free.</strong></blockquote><p>One site gives you a flawless simulation, 50 questions and a countdown that behaves like the real thing, and then every one of its 50 explanation slots renders empty. You get a number and nothing else.</p><p>Another writes genuinely good worked explanations, has no countdown at all, and is quietly testing you on the Wonderlic <strong>Basic Skills Test</strong>, a different instrument from the <strong>50 question cognitive screen </strong>you were sent. Or the questions are just <em>not the kind </em>that appear in the actual test. You get the picture.</p><h3>Key features of Wonderlic Cognitive Ability Test</h3><p>The Wonderlic is 50 questions in 12 minutes. That is <strong>14.4 seconds a question</strong>, covering reading the item, reading five options, deciding and clicking. It’s the tightest clock of the three big cognitive screens.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yxn38xX5jvHDxW24bXb4SQ.png" /></figure><p>This is why I recommend that you <strong>must </strong>ensure you do <strong>timed </strong>practice tests. Because an untimed practice test measures something the exam never asks. Answering a Wonderlic question is close to unrelated to answering one in fourteen seconds with a number counting down in the corner.</p><p>I know that because I got it wrong in the chair. I hit a word problem I could definitely solve, and I could feel myself starting to solve it. Then I did the arithmetic instead: roughly 40 seconds against a budget of 14.4, so about 25 seconds of overspend, which is nearly two other questions I’d never reach. I skipped it. That decision was not a reasoning skill, and no untimed practice test would have taught it to me.</p><p>One warning, because prep sites copy this mistake off each other.</p><p><strong>Don’t carry the CCAT guessing logic onto the Wonderlic.</strong> The CCAT scores raw correct answers, so a blind guess there is free expected value. Wonderlic scoring is weighted by difficulty and already accounts for guessing, so that maths does not transfer.</p><h3>The six things I checked before ranking anything</h3><p>I want these on the page before the ranking, so you can see where the entry I recommend loses.</p><ol><li><strong>Does it serve all 50 questions</strong>, or a five question taster.</li><li><strong>Is there a real 12 minute countdown that runs</strong>, verified across separate loads.</li><li><strong>Is there a worked explanation for every item</strong>, and does it name the trap answer.</li><li><strong>Is it even the right exam.</strong> The Wonderlic Basic Skills Test is a separate product.</li><li><strong>What the free tier contains</strong> before you hand over an email.</li><li><strong>The real price</strong>, taken from the checkout rather than the sales page.</li></ol><p>Prices below are what I saw on 1 September 2026. Treat them as directional.</p><h3>Top Wonderlic Cognitive Ability Practice Tests</h3><h3>1. PrepClubs, the one I recommend, so read this entry hardest</h3><p><strong>Rating: </strong>9.5/10</p><p><strong>Pros:</strong></p><ul><li>On <a href="https://prepclubs.com/">PrepClubs</a>, all tests are timed.</li><li>You’re allowed to pause the test (since you’re “prepping”).</li><li>Has AMAZING analytics i.e. it told me <em>which question types </em>I got wrong the most (verbal reasoning in my case) and also told me which question types I spent “most time on” i.e. I needed to work on THAT too since time is a big thing in this type of test.</li><li>I think they have the biggest question bank with 8 Full Fledged Practice Tests and 400+ practice questions.</li><li>Questions closely match the difficulty &amp; pattern of the actual test.</li></ul><p><strong>Cons:</strong></p><ul><li><strong>The free tier is one free full-length test, no credit card required. </strong>For full practice tests set, you need to pay.</li><li>The free run reports your score and where you sit against the benchmark. It does not release the worked explanations, which sit behind the paywall — but it has the lowest cost among all other options (except the <a href="https://www.udemy.com/course/wonderlic-exam-questions-practice-test/">Udemy course</a> I took; but that course was <strong>terrrible!)</strong></li></ul><p>You don’t need to but I got their “All Access Bundle” since Prepclubs also had <a href="https://prepclubs.com/free-practice-tests/ccat">Criteria’s Cognitive Aptitude Test</a> &amp; the <a href="https://prepclubs.com/free-practice-tests/pi-cognitive-assessment">Predictive Index Aptitude Test</a> too — and since they’re all “aptitude tests” and the question types are similar I just wanted to get as much practice as possible, and the all access bundle came out cheaper for me than buying each individually. Combined; these three had I think over 1500+ questions (yes, I was a nerd!).</p><p>Anyways, onto the next platform.</p><h3>2. BoostPrep, the best free clock on this page</h3><p><strong>Rating: </strong>8/10.</p><p><a href="https://boostprep.com/wonderlic-practice-test/">BoostPrep</a> has blocks of 52 questions though they don’t seem to be <em>as difficult </em>as the actual exam I sat in, a genuine 720 second limit that runs properly, the figure items rendered as real images rather than described in text, and you can start without registering.</p><p>What it will not do is tell you anything. I counted 101 explanation containers across its free tests and 100 of them render literally empty. The “detailed answer explanations” line comes from its own meta description rather than from the product. So take it for the clock tonight and get your reasons somewhere else.</p><p>Two honest marks against it. The blocks run 52 questions rather than 50, so your pacing feel is slightly off. And its “100% Pass Guarantee” is, per its own policy page, void for anyone who studied, <strong>which is a strange promise to make a prep customer.</strong></p><h3>3. wonderlictestpractice.com, a perfect clock attached to nothing</h3><p><strong>Rating: </strong>3/10.</p><p>(Just ignore this one) The simulation is excellent but the site feels like … I don’t know how to describe it. It just suddenly throws random ads at you.</p><p>There is no paid tier, which explains the shape of it. The site earns from AdSense and Amazon affiliate links, and roughly 26 of its 41 blog posts are NFL quarterback score bait. Worth one sitting for the clock. Not worth a second.</p><h3>4. beatthewonderlic.com, real test, decorative pricing</h3><p><strong>Rating: </strong>2/10.</p><p>(Ignore it) The free test is real: 50 questions, a 720 second limit, and it runs, which earns it a place above everything below.</p><p>Then the details. The <strong>$99.99 crossed out above the $49.99 price is hard-coded into the page</strong>, so nothing was ever sold at it. Three question images are hotlinked from a raw IP address and are broken, so you get items you can’t answer. The sitemap dates sit in 2020 and 2021.</p><h3>5. The official Wonderlic sample, honest and tiny</h3><p>Take it. It’s literally just 5 questions.</p><p>Five cognitive questions, no answer key, no explanations, no gate. It is the only material here written by the people who write the exam, so the phrasing is exactly right, and its single image item is a graph-reading question rather than the shape rotation everyone expects. Take it early to calibrate the real wording, then go somewhere with a clock.</p><h3>6. NurseHub, where the free test has the timer removed</h3><p><strong>Rating: </strong>6/10. <br>(An acquaintance took it; didn’t like the quality of paid questions much. But you can still take their free test just for extra practice).</p><p>The free 50 question Wonderlic test has its <strong>time limit set to zero</strong>, which turns the whole thing into an untimed sample. The paid tests are configured properly at 50 questions in 12 minutes, so the capability is there.</p><p><strong>One more thing: the policy page says the very first charge is never returned, while the homepage advertises a three day free trial. Those two sit badly together.</strong></p><h3>7. testprepreview.com, a different exam behind an email wall</h3><p>Its practice tests are for the <strong>Wonderlic Basic Skills Test</strong>, which is not the cognitive screen most employers send. Practising it won’t hurt you. It also will not prepare you.</p><p>The free tier is five questions and then a hard email wall, and its optional timer counts up rather than down, training the opposite instinct to the one you need. It funnels into Mometrix at roughly $72.99 to $75.99, with flashcards at $43.99. Credit where due: Mometrix will take a purchase back for up to a year, which is genuinely generous.</p><h3>The one I would skip, plus one price worth checking</h3><p><strong>wonderlictestprep.com</strong> shows a countdown that does not count. I pulled the page twice, 72 seconds apart, and the timer read the same value. A frozen clock on a site selling timed practice is the product failing in public.</p><h3>What Wonderlic says about paying for prep, taken both ways</h3><p>Wonderlic tells candidates they <strong>“should never have to pay”</strong> for preparation material. Read it twice.</p><p>The first reading is generous and it is fair. Much of the paid Wonderlic market is thin. This page alone has an unmaintained shop with a fake strikethrough, a frozen timer, and a $79 SKU that expires in a week. If the vendor is warning you off that, the vendor is right.</p><p>The second reading is structural. A timed cognitive screen keeps its predictive validity partly because candidates walk in cold. If everyone drills the format, the instrument measures preparation as well as ability, and the vendor has a noisier signal to defend to the employers buying it. So Wonderlic has a real reason to discourage prep regardless of whether prep works. Both readings are true at once.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*komCnrFTsiYbyNLm9vm0QQ.png" /></figure><h4>What a real explanation looks like</h4><p>This is the actual difference between the entries above, so here’s a real item from the PrepClubs bank. The Wonderlic includes attention to detail work, which catches people out because it looks easy.</p><p><strong>How many of the five pairs below are exact duplicates?</strong></p><ol><li>Kowalski 386 Dunmore / oKwalski 386 Dunmore</li><li>Whitfield 722 Crestwood / Whitfield 722 Crestwood</li><li>Nakamura 462 Kingsbridge / Nakamura 462 Knigsbridge</li><li>Castellano 580 Bellamy / Castellano 580 Bellaym</li><li>Bartlett 674 Kingsbridge / Bartlett 674 Kinsgbridge</li></ol><p>The answer is <strong>1</strong>. Only pair 2 matches. Pair 1 has the first two characters swapped, pair 3 swaps characters 15 and 16, pair 4 swaps 21 and 22, pair 5 swaps 17 and 18.</p><p>Now the part that matters more than the answer. Every planted error is a transposition, two adjacent characters trading places, which is exactly what a fluent reader cannot see. You read “Knigsbridge” and your brain serves you “Kingsbridge”, because reading is shape matching and the shape barely changed. Most people answer 4 or 5 here and feel confident doing it.</p><p>Here is the same item type on a live clock, which is the version you will actually meet.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*I9kjUG0S2IvhcmgBXUNBOQ.png" /></figure><p>The answer is <strong>4</strong>. Granger on the left is Grainger on the right, one inserted letter buried inside a word you have already recognised. Same mechanism as the transpositions above: your eye completes the word from its shape and never reaches the individual letters, which is why scanning a pair and feeling fine about it means nothing.</p><p>Comfortable is the wrong feeling here.</p><p>The fix is mechanical rather than clever: read left to right in fixed chunks of two or three characters, comparing chunk against chunk rather than entry against remembered entry. It feels slower and it is faster, because you stop re-reading.</p><p>That’s what a worked explanation owes you, and a score of 38 gives you none of it: the trap, the reason it works on you specifically, and a habit you carry into the next item.</p><h3>If your test is in 72 hours</h3><ol><li><a href="https://prepclubs.com/free-practice-tests/wonderlic"><strong>Take one full-length Wonderlic Aptitude timed test today.</strong></a> Not tomorrow. The score matters far less than the map of where your 720 seconds went.</li><li><strong>Sort every wrong answer into four buckets:</strong> misread it, knew it but was too slow, didn’t know the concept, never reached it. Those have four different fixes, and lumping them together is why people plateau.</li><li><strong>Drill only your worst bucket.</strong> Three days is not enough time to be thorough, so be selective.</li><li><strong>Take a second full-length test the day before.</strong> Pacing practice, not learning. Practise abandoning questions.</li><li><strong>On the day: every tab closed, and no blanks behind you.</strong></li></ol><p>A caveat I’d rather state than skip:</p><p>Wonderlic does not publish a scratch paper policy, a retake policy, or a go-back policy for the cognitive section. What it does tell candidates is <strong>“do not skip around or spend too much time on any one question.”</strong> Build a run that goes forward only and you never need to know what the software would have allowed.</p><h3>Where I would actually send you</h3><p>If you want free and nothing paid, use BoostPrep. That’s the honest answer, and better said here than left for you to wonder about.</p><p>If you want the timed version with explanations attached, <a href="https://prepclubs.com/free-practice-tests/wonderlic?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=best-wonderlic-practice-tests-2026">PrepClubs</a> is built for exactly the 72 hour window this article describes. The terms, stated the way I stated everybody else’s: <strong>one free full-length test, your pick of cluster, signup required.</strong> That free run gives you the score and the benchmark rather than the explanations. The $49 opens the 400 question Wonderlic bank, where every item carries a written explanation of the trap as well as the answer, and the Pass Guarantee extends your access by 30 days, which is an extension rather than a refund and worth reading twice.</p><p>If your test turns out to be the CCAT, which happens because recruiter emails often do not name the instrument, <a href="https://ccattests.com/free-ccat-practice-tests?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=best-wonderlic-practice-tests-2026">CCAT Tests</a> has five full-length simulations at the real 50 questions in 15 minutes, and my longer <a href="https://mjunaidkhalid.com/ccat-ultimate-guide-with-practice-tests-free-resources-2025/?utm_source=medium&amp;utm_medium=referral&amp;utm_campaign=best-wonderlic-practice-tests-2026">CCAT guide</a> covers each question type.</p><h3>TLDR</h3><ul><li><strong>The clock and the explanations almost never ship together for free.</strong> That one pattern explains nearly every free Wonderlic practice test.</li><li><strong>The Wonderlic is 50 questions in 12 minutes, which is 14.4 seconds each.</strong> An untimed sample measures a skill the exam does not test.</li><li><strong>BoostPrep gives you the best genuinely free clock, and it costs nothing.</strong> Real 720 second limit, no account needed, and 100 of its 101 explanation slots render empty. Its pass guarantee is void if you studied.</li><li><strong>wonderlictestpractice gives a free simulation with all 50 explanations empty</strong>, testprepreview tests the Basic Skills Test rather than the cognitive screen, NurseHub sets its free time limit to zero, and wonderlictestprep shows a countdown that never moves.</li><li><strong>Wonderlic scoring is weighted by difficulty and accounts for guessing</strong>, so the CCAT habit of guessing everything does not transfer.</li><li><strong>Wonderlic says you should never have to pay for prep.</strong> Partly fair, partly self-interested: a drilled cognitive test is a noisier instrument to sell.</li><li><strong>Take one timed full-length test today, not more sample questions.</strong> PrepClubs gives you one free full-length test with signup, reporting your score rather than the explanations, and BoostPrep costs nothing at all.</li></ul><p>Stop reading comparison pages, including this one, and go put yourself under a real 12 minute clock.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=986c6e4bff19" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cross Out Every Word a Competitor Could Also Write. What Is Left Is Your Positioning Statement.]]></title>
            <link>https://medium.com/@Elena.Rivera/cross-out-every-word-a-competitor-could-also-write-what-is-left-is-your-positioning-statement-b5e5bc9ee187?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/b5e5bc9ee187</guid>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[copywriting]]></category>
            <category><![CDATA[branding]]></category>
            <category><![CDATA[positioning]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Wed, 26 Aug 2026 19:01:01 GMT</pubDate>
            <atom:updated>2026-08-26T19:01:01.267Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="A printed positioning statement with every one of its words struck through in red, showing what is left of a template sentence after the deletion pass." src="https://cdn-images-1.medium.com/max/1024/1*ZyaKnMC8Rpcw7IEfsP6-mw.png" /></figure><p>Every article about writing a positioning statement shows you Nike.</p><p>Then Apple. Then a fill in the blanks template with four slots in it, one each for your target market, your category, your benefit and your reason to believe. It is a real format, taught by real people, and <a href="https://online.hbs.edu/blog/post/brand-positioning-statement">Harvard Business School Online</a> publishes a version of it.</p><p>I filled mine in one afternoon, in about four minutes. Here is what came back:</p><blockquote>For growing teams, our platform is the leading solution that helps you work smarter and unlock your full potential.</blockquote><p>True. Grammatical.</p><p>Also an accurate description of roughly two hundred other companies.</p><p><strong>That is the failure nobody warns you about.</strong> The template does not hand you a wrong answer. It hands you a sentence that is correct about your entire category and silent about you.</p><p>Here is what I do now instead. It takes an afternoon, and it ends with one sentence I can actually check.</p><h3>step one: fill in the template anyway, then treat it as evidence</h3><p>Do the four minutes. Fill in every blank, fast, without editing.</p><p>You are not writing the deliverable here.</p><p>You are collecting your own defaults, which is the only way to ever see them. Mine were <em>growing teams</em>, <em>platform</em>, <em>work smarter</em>. I had written all three without noticing I was writing them.</p><p><strong>The template’s real job is to show you the sentence you would produce on autopilot.</strong></p><p>Then you take it apart.</p><h3>step two: cross out every word a competitor could also write</h3><p>Print it. Go one word at a time and ask a single question.</p><p><strong>Could the company I lose deals to put this word on their own page tomorrow?</strong></p><p>If yes, strike it through. Mine did not survive contact:</p><ul><li><strong>“Growing teams” goes.</strong> Everyone says growing teams.</li><li><strong>“Platform” goes.</strong> It names a shape of software, not a job it does.</li><li><strong>“Leading solution” goes.</strong> Nobody has ever typed that phrase into a search box.</li><li><strong>“Unlock your full potential” goes</strong>, along with every other line you have read on an airport billboard.</li></ul><p>What survives is usually two or three words.</p><p>Quite often it is nothing at all.</p><p>An empty page after this pass is not a failure of the exercise. It is the actual finding: you have been describing your category and filing it as your position.</p><h3>step three: ask what they would use if you disappeared</h3><p>This is the step the templates skip. It is also the one that does the work.</p><p>April Dunford puts competitive alternatives first among the five components of positioning. She is blunt about the fill in the blanks exercise too, calling it <a href="https://www.aprildunford.com/post/an-introduction-to-positioning">“not only pointless but potentially dangerous”</a>.</p><p>The reason is simple. The template asks you to name your category before you have worked out what you are being compared against.</p><p>And what you are compared against is rarely a competitor. It is usually one of these:</p><ul><li><strong>A spreadsheet</strong>, with a tab somebody built in a hurry.</li><li><strong>A folder</strong> one person maintains by hand.</li><li><strong>A calendar reminder</strong> nobody has ever completed on time.</li><li><strong>Doing nothing</strong>, and hoping.</li></ul><p>So ask the customers who bought in the last month, in these words: <em>what were you doing about this before us?</em></p><figure><img alt="A numbered checklist card showing the four steps for writing a positioning statement, from filling in the template to reading the sentence back to a recent customer." src="https://cdn-images-1.medium.com/max/1024/1*a9_VucZZX3boP8HTkV_Eqw.png" /><figcaption>The whole afternoon on one card. Step three is the one everybody skips, and it is the only one that brings new information into the room.</figcaption></figure><h3>the answer arrives in their words, not yours</h3><p>The sentence that reset ours came off an onboarding call in late July.</p><p>A customer said they used to do it in a spreadsheet. That someone had to remember to update it. That nobody ever did.</p><p><strong>That is a positioning statement example worth having.</strong> Not because it is elegant, but because it is checkable.</p><blockquote>Our real competitor was a spreadsheet with an out of date tab and a person who felt guilty about it.</blockquote><p>You cannot get that sentence out of a template, because the template has no blank for it.</p><p>I had spent an entire walk that same week working out how to describe what we sell. Out loud, alone, with nobody to correct me.</p><p>The customer did more in one clause than the walk did in forty minutes.</p><h3>step four: read it back and ask who it excludes</h3><p>Two minutes, and it is the only quality check on this list I actually trust.</p><p>Read your surviving sentence to somebody who bought recently. Then ask: <strong>who is this obviously not for?</strong></p><ul><li><strong>They name a group instantly.</strong> The sentence is working. A position that excludes nobody is a description.</li><li><strong>They hesitate.</strong> You are still holding a category summary in a nicer font.</li><li><strong>They name the wrong group.</strong> The most useful outcome of the whole afternoon, and it costs you one rewrite.</li></ul><p>Two honest limits on all of this.</p><p>It does not hand you a homepage headline. An internal positioning sentence and a headline are different objects with different jobs, and confusing the two is how you end up with a homepage that reads like a memo.</p><p>And with no customers yet, step three is not available to you at all. You are guessing, which is allowed, as long as you write the date on the guess and come back to it.</p><h3>what actually changed</h3><p>The sentence on our page is shorter now.</p><p>One of the words in it came out of somebody else’s mouth.</p><p>I have built brand voice guides for about a dozen companies, and I have written before about how <a href="https://medium.com/@Elena.Rivera/i-read-thirty-brand-voice-examples-posts-they-all-show-you-the-wrong-thing-6272d023114f">the brand voice examples everyone shares show you the wrong thing</a>. Positioning has the same disease.</p><p>Every public example is a finished sentence from a company whose position was settled by something other than a template.</p><p>So I stopped trying to write mine, and started collecting other people’s.</p><p>Four minutes of template. One afternoon of crossing out. One question on a call.</p><p>The part I am least proud of is that I ran a company for years describing it in words I had never once heard a customer use.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b5e5bc9ee187" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ChatGPT’s Voice Button Is a Phone Call. Its Microphone Is a Keyboard.]]></title>
            <link>https://medium.com/@Elena.Rivera/chatgpts-voice-button-is-a-phone-call-its-microphone-is-a-keyboard-8aeb663a5103?source=rss-8f31d295e461------2</link>
            <guid isPermaLink="false">https://medium.com/p/8aeb663a5103</guid>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[writing]]></category>
            <category><![CDATA[productivity]]></category>
            <dc:creator><![CDATA[Elena Rivera]]></dc:creator>
            <pubDate>Tue, 25 Aug 2026 19:01:01 GMT</pubDate>
            <atom:updated>2026-08-25T19:01:01.765Z</atom:updated>
            <content:encoded><![CDATA[<p>Every ChatGPT voice mode tutorial ends at the same moment. Tap the button, start talking, marvel at how natural it sounds.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KaU-FtbXk8BaQ2BYRIJ-7Q.png" /></figure><p>None of them film what happens forty minutes later, when you walk back through your own front door holding a phone with nothing on it.</p><p>That was me, most of late July.</p><h3>there are two buttons and they are not the same feature</h3><p>This is the bit the tutorials skip, and it sits in OpenAI’s own documentation.</p><p><strong>ChatGPT does not have one voice feature. It has two, and they do opposite jobs.</strong></p><ul><li><strong>Voice mode.</strong> Opens a <a href="https://learn.chatgpt.com/docs/features/voice">live conversation</a>. You talk, it talks back, and the chat has to be started in voice mode from the beginning.</li><li><strong>Voice dictation.</strong> Turns your speech into prompt text in the box, so you can read it, edit it, and send it like anything else you typed.</li></ul><p>One of them gives you a conversation. The other gives you <em>text</em>.</p><p>Jennifer Jolly’s walkthrough has to warn readers about this outright: <a href="https://jjtechish.substack.com/p/how-to-use-chatgpt-voice">do not tap the little microphone</a> if what you want is the live conversation, because the microphone only turns what you say into typed text.</p><p>That warning exists because everybody presses the wrong one.</p><p>I pressed the wrong one for two weeks. In my case the wrong one was the <em>conversation</em>.</p><p>Both controls sit in the same message box, a thumb’s width apart. Nothing on the screen tells you that one of them hands text back and the other one never will.</p><h3>what I actually did with it</h3><p>The walk was supposed to be twenty minutes.</p><p>I was trying to work out how to describe what my company sells to a customer type I had only recently started taking seriously.</p><p>I talked at ChatGPT for about forty minutes. It pushed back twice. One of those two times it was right, which is a better hit rate than most meetings I sit in.</p><p>By the time I got home I had a position I believed in and a phone with no text on it.</p><p>So I sat at my desk and wrote it out from memory. Badly, and slower than if I had stayed in and typed all afternoon.</p><p>Here is what actually went missing:</p><ul><li><strong>Not the conclusion.</strong> I remembered that fine.</li><li><strong>The three or four sentences on the way to it.</strong> The ones where I said the thing plainly, before I knew I was being quoted.</li></ul><p>Those middle sentences are the ones that end up in the copy. They are also the ones nobody can reconstruct at a desk two hours later.</p><blockquote>The conversation was genuinely good. That is the frustrating part.</blockquote><p>Nothing about the hour was wasted, except that none of it survived contact with my desk.</p><figure><img alt="A decision flowchart for chatgpt voice mode versus voice dictation, branching on whether anyone is waiting for the output, and ending in either a live conversation or text in the box." src="https://cdn-images-1.medium.com/max/814/1*Hl2KUFJ101I29-2opGYytA.png" /><figcaption>The whole decision, drawn out. It took me two weeks and one wasted walk to work out that there was a fork here at all.</figcaption></figure><h3>voice mode is not competing with typing</h3><p>This is where the framing goes wrong, and the framing is doing real damage.</p><p>Voice mode is not a faster way to write. It is a faster way to <em>think</em> with somebody in the room.</p><blockquote>Its real competitor is the phone call. Nobody has ever finished a phone call and expected to find a document waiting on their desk.</blockquote><p>Dictation is the feature competing with typing. It is dull, it has existed for years, and it produces the only thing my working day runs on, which is words already sitting in the box.</p><p>Every guide I found treats these two as the same upgrade to the same problem.</p><p>They are not even solving the same problem.</p><h3>where it earns the hour</h3><p>I have not stopped using it. There is one job voice mode does better than anything else I own.</p><p><strong>It earns the hour when every one of these is true:</strong></p><ul><li><strong>Nobody is waiting for an artifact.</strong> No reply, no brief, no page that has to exist by the time I sit down.</li><li><strong>The work is deciding, not producing.</strong> I am working out what I think, not writing it up.</li><li><strong>I am away from the desk.</strong> A blank page invites me to perform. A conversation does not let me.</li></ul><p>I wrote a while ago about <a href="https://medium.com/@Elena.Rivera/i-stopped-using-ai-to-brainstorm-thats-when-it-got-useful-c620a00dc567">why I stopped asking AI to brainstorm for me</a>. This is the same principle on a different button: the machine is a weak source of first ideas and a very good thing to argue with once I have one.</p><p>So voice mode is my thinking tool.</p><p>It was never built to be my drafting tool, and I was the one who kept filing it under drafting.</p><p>The real catch is that it does not tell you which one you are doing until the hour is already gone.</p><p>The tutorials teach you the tap. Not one of them teaches you the <em>decision</em> underneath it.</p><h3>what changed on my screen</h3><p>Nothing dramatic. I moved one finger about an inch to the left.</p><ul><li><strong>If it has to end up somewhere,</strong> a reply, a brief, a landing page, I dictate it straight into the box it belongs in.</li><li><strong>If it is me working out what I think,</strong> it gets the conversation, and it gets it outdoors.</li></ul><p>That one distinction has done more for my week than any prompt I have ever written.</p><p>I have used ChatGPT since it could barely handle a date, and it took me until last month to notice that two controls an inch apart were doing unrelated jobs. In fairness, nobody making tutorials has noticed either.</p><p>Those forty minutes on the walk were a phone call with a very patient stranger.</p><p>For two weeks I had been logging calls like that as drafting.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8aeb663a5103" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>