<?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 Noam Shemla on Medium]]></title>
        <description><![CDATA[Stories by Noam Shemla on Medium]]></description>
        <link>https://medium.com/@noamshemla?source=rss-7ead78d50714------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*vzDixycd9HzI7PGq_yJD0A.png</url>
            <title>Stories by Noam Shemla on Medium</title>
            <link>https://medium.com/@noamshemla?source=rss-7ead78d50714------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 19:45:31 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@noamshemla/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[It Doesn’t Hurt to Ask: Why Follow-Up Questions Make People Like You]]></title>
            <link>https://medium.com/exprea/it-doesnt-hurt-to-ask-why-follow-up-questions-make-people-like-you-3951fd710fee?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/3951fd710fee</guid>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[public-speaking]]></category>
            <category><![CDATA[relationships]]></category>
            <category><![CDATA[communication]]></category>
            <category><![CDATA[dating]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Tue, 17 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-17T14:31:01.553Z</atom:updated>
            <content:encoded><![CDATA[<p>It Doesn’t Hurt to Ask: Why Follow-Up Questions Make People Like You</p><p>Two experiments and a speed-dating study found that people who ask more questions are liked more. Follow-up questions did most of the work.</p><h3>TL;DR</h3><ul><li>Across two experiments and a speed-dating study, people who asked more questions were liked more by their partners.</li><li>Follow-up questions drove most of the effect. They made up 40.5% of all questions and were the only type linked to liking.</li><li>The askers did not expect this. In Study 1 they predicted liking of 5.27 versus 5.19, almost no difference.</li><li>The partner felt more heard. Responsiveness explained part of the effect (indirect effect .12 in Study 1 and .07 in Study 2).</li><li>The experiments are strong. The speed-dating result is correlational, and a reanalysis questioned it.</li></ul><p>This is the short version. Read the full article on exprea.io: <a href="https://exprea.io/blog/huang-2017-it-doesnt-hurt-to-ask-question-asking-increases-liking">It Doesn’t Hurt to Ask</a>.</p><h3>Why this study matters</h3><p>Many people worry that too many questions feel like an interview. Huang and colleagues (2017) tested this fear. They asked a simple thing: if you ask more questions, do people like you more or less?</p><h3>The study in short</h3><p>In Study 1, <strong>398 people in 199 pairs</strong> chatted by text for 15 minutes. In Study 2, <strong>338 people in 169 pairs</strong> did the same online. People were told to ask either at least nine or at most four questions. These numbers were the 25th and 75th percentiles in a pilot. The outcome was partner liking, rated on four items from 1 to 7. Then <strong>612 observers</strong> read transcripts and rated them. In Study 3, <strong>110 graduate students</strong> went on 15 to 19 speed dates of four minutes each (1,961 date observations). The outcome there was whether the partner wanted a second date.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*gDi6EMHze5vqV38m.png" /><figcaption>Diagram by Exprea. Data: Huang et al. 2017.</figcaption></figure><h3>What they found in the chats</h3><p>In Study 1, the many-question group asked 10.2 questions on average and the few-question group asked 4.3. Partners liked the heavy askers more, by about one-third of a standard deviation. But the askers did not see it coming: they predicted liking of 5.27 versus 5.19.</p><p>In Study 2, liking was 6.02 for heavy askers and 5.79 for light askers. The effect held no matter how many questions the partner had been told to ask. The observers who read the transcripts did not rate heavy askers higher. They tended to prefer the people who were asked many questions and did most of the talking.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*S_tEP1JjdlHNZ4Hn.png" /><figcaption>Chart by Exprea. Data: Huang et al. 2017.</figcaption></figure><h3>Which questions matter</h3><p>Three coders classified the questions in 368 chats. Six types were in the scheme, and four were common enough to analyse. Follow-up questions, which build on what the partner just said, were the most common (40.5%). They were the only type linked to liking (effect .89). Full switches, which change the topic, were 27.6% of questions and lowered responsiveness. When people were told to ask many questions, the follow-up share rose from 28% to 42%, and this shift explained part of the effect.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Pv40oVsqPErN9XtH.png" /><figcaption>Chart by Exprea. Data: Huang et al. 2017.</figcaption></figure><h3>Why it works</h3><p>The authors tested a mediation idea: asking makes the partner feel heard, and feeling heard makes the partner like you. The standardised indirect effect was .12 in Study 1 and .07 in Study 2.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*rh-szLjygy3TNmoT.png" /><figcaption>Diagram by Exprea. Data: Huang et al. 2017.</figcaption></figure><h3>The speed dates</h3><p>For Study 3, the team trained a classifier on 4,209 labelled question turns. It reached 87% accuracy on held-out examples. On average, each person asked about 9.8 questions per date, roughly 22% of turns, and about 4.5 of these were follow-ups. Second-date “yes” rates were 57% for men and 37% for women. A difference of 8% in follow-up rate was linked to about one extra “yes” over 20 dates.</p><h3>Has it held up?</h3><p>Partly. Kluger and Malloy (2019) reanalysed Study 3 with a social relations model. They found that only about one-third of the variation in question-asking was a stable trait of the asker, and that follow-up and switch scores were hard to separate. They also argued that a second-date “yes” is not the same as liking. Yeomans and colleagues (2019) replied that the reanalysis tested the wrong variables and ignored results from all three studies. In supplementary models, once each asker’s general desirability is controlled, the speed-date follow-up effect shrinks and is no longer significant. A 2025 correction by the APA found only minor reporting errors, for example 5.79 reproducing as 5.78. The audit confirmed all conclusions.</p><h3>How far to trust it</h3><p>The experiments (Studies 1 and 2) are the strongest part. But they were 15-minute text chats between strangers, so there was no tone of voice. The paper did not test more than nine questions. Liking is a rating, and nobody checked whether heavy askers formed lasting relationships. The speed-date link is observational, so it cannot prove cause alone.</p><h3>What to take from this</h3><ul><li><strong>Ask more.</strong> In short talks with new people, more questions did not backfire.</li><li><strong>Ask follow-ups.</strong> Build on what the person just said. Do not jump to a new topic.</li><li><strong>Expect to be surprised.</strong> Askers underestimate how much it helps, so practise it on purpose.</li></ul><p>Next in the series: <a href="https://exprea.io/blog/ireland-2011-language-style-matching-predicts-relationship-initiation-and-stability">Language Style Matching: How the Smallest Words Predict Who Dates and Who Stays Together</a>.</p><h3>Source</h3><p>Huang, K., Yeomans, M., Brooks, A. W., Minson, J., and Gino, F. (2017). It doesn’t hurt to ask: Question-asking increases liking. <em>Journal of Personality and Social Psychology, 113</em>(3), 430–452. DOI: <a href="https://doi.org/10.1037/pspi0000097">10.1037/pspi0000097</a></p><p>Practice these skills with feedback at <a href="https://exprea.io/">exprea.io</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3951fd710fee" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/it-doesnt-hurt-to-ask-why-follow-up-questions-make-people-like-you-3951fd710fee">It Doesn’t Hurt to Ask: Why Follow-Up Questions Make People Like You</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Your Voice in a Job Interview Predicts the Verdict: What DeGroot and Motowidlo Found]]></title>
            <link>https://medium.com/exprea/your-voice-in-a-job-interview-predicts-the-verdict-what-degroot-and-motowidlo-found-a86506fe733a?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/a86506fe733a</guid>
            <category><![CDATA[public-speaking]]></category>
            <category><![CDATA[career-advice]]></category>
            <category><![CDATA[communication]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[job-interview]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Sun, 15 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-15T14:31:01.662Z</atom:updated>
            <content:encoded><![CDATA[<p>In 1999, two researchers tested whether how managers sounded in a mock interview could predict their job ratings. The link was real but small.</p><h3>TL;DR</h3><ul><li>DeGroot and Motowidlo (1999) tested whether voice and looks in a mock job interview predict judgments and job performance.</li><li>Study 1 had 60 managers. Study 2 had 110 managers. They answered interview questions on video as if applying for their own jobs.</li><li>A combined vocal score correlated r = .32 with the interview judgment and r = .20 with supervisor and colleague performance ratings. The looks score was linked to the judgment (r = .21) but not to performance (r = .14).</li><li>Listeners’ personal reactions (liking, trust, credibility) to the voice carried the link to performance.</li><li>The effects are small, the interviews were simulated, and the authors did not recommend using voice in hiring.</li></ul><p>This is the short version. Read the full article on exprea.io: <a href="https://exprea.io/blog/degroot-motowidlo-1999-visual-and-vocal-interview-cues-job-performance">Your Voice in a Job Interview Predicts the Verdict</a>.</p><h3>Why this study matters</h3><p>Interviewers say they judge answers. But many studies show that looks and sound also change the verdict. DeGroot and Motowidlo asked a harder question: do these cues only bias the interviewer, or do they also predict how well the person later does the job?</p><h3>The study in short</h3><p>In Study 1, <strong>60 managers</strong> from four utility companies were interviewed on video as if they applied for their own jobs. Their supervisors rated their performance. In Study 2, <strong>110 managers</strong> at a news-publishing company answered nine structured questions. <strong>26 company employees</strong> judged the interviews. Performance ratings came from supervisors and colleagues.</p><p>The team measured five vocal cues by computer, from short speech samples. Students also rated visual cues from silent video: attractiveness, smiling, gaze, hand movement and leaning in. Other students said how much they would like, trust and believe each manager, once from sound alone and once from silent video alone.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*LiCpARPkJzADYMaW.png" /><figcaption>Chart by Exprea. Data: DeGroot and Motowidlo 1999 (as summarised on exprea.io).</figcaption></figure><h3>How the vocal score was built</h3><p>Each cue was turned into a standard score. The two cues expected to help (more pitch variability and faster speech) were added. The three others (higher pitch, more pauses, more amplitude variability) were subtracted. Women’s pitch values were adjusted by 97 Hz. Speech rate was estimated from pitch-period length, and it correlated r = .88 with syllable counts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*fstVUcTi1q2j_poE.png" /><figcaption>Equations by Exprea. Concept: DeGroot and Motowidlo 1999.</figcaption></figure><h3>What they found</h3><p>In Study 1, no single cue was significant on its own. But all pointed in the expected direction, and the combined score correlated r = .18 with performance. The authors saw this as a feasibility check.</p><p>In Study 2 the pattern was clearer. The vocal score correlated r = .32 with the interview judgment and r = .20 with job performance. The looks score correlated r = .21 with the judgment and r = .14 with performance, which was not significant.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*C1stxbgidnMFMDLc.png" /><figcaption>Chart by Exprea. Data: DeGroot and Motowidlo 1999.</figcaption></figure><h3>Why voice may matter: the reactions</h3><p>The authors tested a mediation idea. Voice changes how listeners feel about the speaker. Those feelings then shape the verdict, and they also relate to performance. For voice and performance, the link between voice and reactions was .20, and the link between reactions and performance was .23. After this step, the direct path from voice to performance fell to .16 and was no longer significant. For looks, the link to performance was not supported.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*SXBMSACd8RwsdQ_s.png" /><figcaption>Diagram by Exprea. Data: DeGroot and Motowidlo 1999.</figcaption></figure><h3>Has it held up?</h3><p>We found no replication in the material we used. The authors called theirs the first link between specific interview vocal cues and job performance, and said it needs extensive replication. The mediation result fits the reaction explanation, but it does not prove it.</p><h3>How far to trust it</h3><p>Be careful. The interviews were simulated: managers answered as if applying for the job they already had. The effects are small. A correlation of .20 explains about 4% of the variance. Performance ratings are judgments too, so raters may react to the same vocal cues. Some measures were proxies, such as speech rate from pitch-period length. The authors could not rule out rater bias and found no added value beyond established predictors. They did not recommend using voice in hiring.</p><h3>What to take from this</h3><ul><li><strong>Voice shapes first reactions.</strong> Liking, trust and credibility are formed fast, and the voice feeds them.</li><li><strong>Do not over-trust it.</strong> A voice score is a weak signal. It should not decide who gets hired.</li><li><strong>Train the voice you control.</strong> Pace, pauses and pitch movement can be practised and measured.</li></ul><p>This is the last post in the Communication and Wealth series. See all four parts: <a href="https://exprea.io/blog/series/studies-wealth">Communication and Wealth</a>.</p><h3>Source</h3><p>DeGroot, T., and Motowidlo, S. J. (1999). Why visual and vocal interview cues can affect interviewers’ judgments and predict job performance. <em>Journal of Applied Psychology, 84</em>(6), 986–993. DOI: <a href="https://doi.org/10.1037/0021-9010.84.6.986">10.1037/0021–9010.84.6.986</a></p><p>Practice these skills with feedback at <a href="https://exprea.io/">exprea.io</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a86506fe733a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/your-voice-in-a-job-interview-predicts-the-verdict-what-degroot-and-motowidlo-found-a86506fe733a">Your Voice in a Job Interview Predicts the Verdict: What DeGroot and Motowidlo Found</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What Salespeople’s Voices Reveal About Selling: Speaking Rate, Pauses and Pitch]]></title>
            <link>https://medium.com/exprea/what-salespeoples-voices-reveal-about-selling-speaking-rate-pauses-and-pitch-f9aada5f0814?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/f9aada5f0814</guid>
            <category><![CDATA[public-speaking]]></category>
            <category><![CDATA[sales]]></category>
            <category><![CDATA[communication]]></category>
            <category><![CDATA[career-advice]]></category>
            <category><![CDATA[psychology]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Thu, 12 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-12T14:31:01.838Z</atom:updated>
            <content:encoded><![CDATA[<p>A small 1995 study of 21 salespeople found that how they sounded was linked to how much they sold. Here is what it showed, and what it cannot show.</p><h3>TL;DR</h3><ul><li>A 1995 study compared 21 direct salespeople with 13 non-sellers and linked four voice measures to sales performance.</li><li>Speaking rate, pause length and pitch contour were linked to sales. Loudness variation was not.</li><li>Sellers spoke faster, and their pitch fell more at the end of phrases.</li><li>Listeners could hear the speed differences, but their impressions did not match the voice measures or sales.</li><li>It is a very small, exploratory study. It shows a link, not a cause.</li></ul><p>This is the short version. Read the full article on exprea.io: <a href="https://exprea.io/blog/peterson-cannito-brown-1995-voice-characteristics-and-selling-effectiveness">What Salespeople’s Voices Reveal About Selling</a>.</p><h3>Why this study matters</h3><p>Many people believe that a “sales voice” exists. Peterson, Cannito and Brown (1995) tried to measure it. They recorded real salespeople and asked a simple question: do the sounds of a voice relate to how well someone sells?</p><h3>The study in short</h3><p>The authors recorded <strong>21 direct salespeople</strong>. They also recorded <strong>13 people who did not sell</strong> as a comparison group. They measured four things: speaking rate, average pause length, pitch contour and loudness variability. Then <strong>26 women from the product’s target market</strong> listened to the voices and rated the salespeople. The outcome was each seller’s sales performance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*T5W2OZKMup5PTf1-.png" /><figcaption>Diagram by Exprea. Data: Peterson, Cannito and Brown 1995.</figcaption></figure><h3>What they found</h3><p>Three measures were linked to sales: speaking rate, pause length and pitch contour. Loudness variability was not. Sellers also spoke faster than non-sellers, and their pitch contours fell more.</p><p>The listeners could hear the differences in speed. But their impressions were not related to the voice measures, and not related to sales. In other words, how a voice felt to the listeners did not predict who sold more.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*vBHb_ylYfrWg5OVB.png" /><figcaption>Chart by Exprea. Data: Peterson, Cannito and Brown 1995 (abstract).</figcaption></figure><h3>What “pitch contour” means</h3><p>Pitch contour is how the pitch moves over a phrase. It is often summarised as the change over the end of the phrase, in semitones. A falling ending is a negative number. A flat ending is close to zero.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*SzMsNtmbTy7OFxmG.png" /><figcaption>Illustration and equation by Exprea. Concept: Peterson, Cannito and Brown 1995.</figcaption></figure><h3>Has it held up?</h3><p>Later work points in a similar direction. Vaughan-Johnston and colleagues (2026) ran three experiments. They found that falling intonation signals speaker confidence. At moderate attention levels it increased message processing, and it helped strong arguments more than weak ones. Xu and colleagues (2026) studied more than 8,000 real phone sales calls in China, plus an experiment. Brightness and loudness of the voice predicted whether new customers bought and how quickly. The features that helped with new customers were not the same as for existing customers.</p><h3>How far to trust it</h3><p>Be careful. Twenty-one salespeople is a very small sample, and the authors called the study exploratory. It is correlational. Fluent speech may come from experience, product knowledge or confidence, not the other way round. The listener result cannot tell us how voice influences buying. Also, the paper is subscription-only, so this summary uses the published abstract and later research. We did not read the full paper.</p><h3>What to take from this</h3><ul><li><strong>Pace:</strong> fluent speech usually comes from knowing your material well. Rehearsal builds it.</li><li><strong>Endings:</strong> let statements end on a falling pitch.</li><li><strong>Measure yourself:</strong> a recording of your delivery tells you more than asking how it came across.</li></ul><p>Next in the series: <a href="https://exprea.io/blog/degroot-motowidlo-1999-visual-and-vocal-interview-cues-job-performance">Visual and Vocal Interview Cues: How Your Voice in a Job Interview Predicts the Verdict and the Job</a>.</p><h3>Source</h3><p>Peterson, R. A., Cannito, M. P., and Brown, S. P. (1995). An exploratory investigation of voice characteristics and selling effectiveness. <em>Journal of Personal Selling and Sales Management, 15</em>(1), 1–15. DOI: <a href="https://doi.org/10.1080/08853134.1995.10754008">10.1080/08853134.1995.10754008</a></p><p>Practice these skills with feedback at <a href="https://exprea.io/">exprea.io</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f9aada5f0814" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/what-salespeoples-voices-reveal-about-selling-speaking-rate-pauses-and-pitch-f9aada5f0814">What Salespeople’s Voices Reveal About Selling: Speaking Rate, Pauses and Pitch</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[First Offers as Anchors: Why the Opening Number Sets the Price, and How to Break Its Pull]]></title>
            <link>https://medium.com/exprea/first-offers-as-anchors-why-the-opening-number-sets-the-price-and-how-to-break-its-pull-03241c4a2f8e?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/03241c4a2f8e</guid>
            <category><![CDATA[public-speaking]]></category>
            <category><![CDATA[salary]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[career-advice]]></category>
            <category><![CDATA[negotiation]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Tue, 10 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-10T14:31:01.598Z</atom:updated>
            <content:encoded><![CDATA[<p>In three experiments, whoever made the first offer got the better price. Thinking about what does not fit the number cancelled the edge.</p><h3>TL;DR</h3><ul><li>In three negotiation experiments, whoever made the first offer, buyer or seller, got the better price. The opening number strongly predicted where the deal closed.</li><li>In the control groups, the first offer explained between 64% and 86% of the variation in final prices (r = .80 to .93).</li><li>Thinking about information that pointed the other way cancelled the edge: the other side’s alternatives, their walk-away price, or your own target.</li><li>A 2006 meta-analysis of 19 results put the typical link between opening number and outcome at about r = .50.</li></ul><p>Full article with all charts: <a href="https://exprea.io/blog/galinsky-mussweiler-2001-first-offers-as-anchors">exprea.io/blog</a></p><p>Every negotiation has an opening number, and someone has to say it. Folk wisdom says to let the other side go first, so you learn something before you commit. This paper, by Adam Galinsky and Thomas Mussweiler, is one of the most cited behind the opposite advice, and it explains why the opening number is so powerful. It is part two of our series on communication and wealth.</p><h3>The study in short</h3><p>The paper is Galinsky and Mussweiler (2001), published in the Journal of Personality and Social Psychology. The participants were MBA students in negotiation classes. The task was a single-issue price negotiation between a buyer and a seller over a pharmaceutical plant, face to face and by email. There were three experiments. In each, some negotiators were asked to focus on one piece of information before bargaining, and a control group was not. The focus was the other side’s alternatives to the deal (Experiment 1), the other side’s walk-away price (Experiment 2), and one’s own target (Experiment 3). The outcome was the final price, and how closely it tracked the first offer.</p><h3>Why a number sticks</h3><p>The authors built on Mussweiler’s earlier model of anchoring, called selective accessibility. When you hear a number, you do not simply adjust away from it. You start testing whether it could be right, and the easiest evidence to find is the evidence that fits. A buyer who hears a high asking price starts recalling reasons the thing might be worth that much. Those reasons are now the most available thoughts in the room, and the counteroffer and final price are built from them. The model makes a prediction a simple “people are lazy about adjusting” story does not: if you get the receiver to call up the thoughts that do not fit the number, the anchor should lose its grip.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*uJruU3Xr0o518ci1.png" /><figcaption>Diagram by Exprea. Data: Galinsky &amp; Mussweiler 2001 (selective accessibility account).</figcaption></figure><h3>What they found</h3><p>Across all three experiments, whichever side made the first offer got the better outcome, and the first offer was a strong predictor of the final price. A 2006 meta-analysis by Dan Orr and Chris Guthrie extracted the correlations from the paper. In the control groups, the correlation between first offer and final price was .93, .80 and .85. With a focus that pointed away from the offer, it fell to .30, .40 and .34.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*CnwbJRxZCoKWnszy.png" /><figcaption>Chart by Exprea. Data: Orr &amp; Guthrie 2006 (from Galinsky &amp; Mussweiler 2001). Intervals computed by Exprea.</figcaption></figure><p>A correlation is easier to feel as the share of variation it accounts for, which is its square. At .93 the first offer accounts for 86% of the variation in final prices. At .80 and .85 it is 64% and 72%. The lead author’s later review summed it up: over half the variance in outcomes explained by first offers alone.</p><p>The intervals in the chart are ours. With 16 to 20 negotiations per condition, a single correlation is uncertain. The standard way to show that is to move the correlation onto a scale where its sampling error is about the same size everywhere, build the interval there, and move back.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*iz6m_rF58BgpOLr-.png" /><figcaption>Equations by Exprea. Fisher’s transformation, the standard method, not a formula from the paper.</figcaption></figure><h3>Which thoughts break the anchor, and which do not</h3><p>The three focus conditions have one thing in common. Each points the receiver at information that contradicts what the offer implies. A seller’s high opening implies the plant is worth a lot. Thinking about what the seller could get elsewhere, where their real limit lies, or what you yourself are aiming for, all bring to mind numbers other than the one on the table. The lead author’s review adds the other half: focusing on information that fits the offer, such as the other side’s own target or your own alternatives, did not protect anyone.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*y6qvmQma_KVcdmUs.png" /><figcaption>Table by Exprea. Data: Galinsky &amp; Mussweiler 2001; Oesch &amp; Galinsky review.</figcaption></figure><h3>Has it held up?</h3><p>The broad finding has. Orr and Guthrie pooled every simulated negotiation they could find, 16 articles and 19 results, and found a correlation of .497 between the opening anchor and the final outcome. No study in their sample reported less than .30. By their file-drawer test, it would take 113 unpublished studies showing no effect to erase the result. The same analysis supports the paper’s second message. Anchors were strongest when the opening number was the only salient figure, and weaker when the receiver knew something about the other side’s position. Negotiators with professional experience were anchored too, but less (.37 against .51 for novices).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*lQ0JbXmR4PYOBxTs.png" /><figcaption>Chart by Exprea. Data: Orr &amp; Guthrie 2006.</figcaption></figure><p>Other research has drawn limits the 2001 paper did not test. An opening that is extreme enough can be read as an insult, or as a sign that no deal is possible, and end the conversation instead of anchoring it.</p><h3>How far to trust it</h3><ul><li>The samples are small. Each condition had 16 to 20 negotiations, which is why the intervals are so wide, and why every focus-condition interval includes zero.</li><li>The negotiators were MBA students in a classroom simulation with a single issue, price. Real negotiations usually have more issues and more information, and the meta-analysis shows that information weakens anchors.</li><li>A correlation is not a mechanism. As Orr and Guthrie point out, other things can drive both the opening number and the final price, not least how strong each side’s position was.</li><li>Our own limit: the paper is subscription-only. The design and headline findings come from its abstract and the lead author’s review, and the correlations come from the meta-analysis that extracted them.</li></ul><h3>What to take from this</h3><p>A negotiation is a conversation, and the first number spoken in it does a lot of the work. Two habits follow, and both are things you say out loud. If you can, say the number first. Say it plainly, and give the reason for it in the same breath, so the reasons that fit it are the ones they hear. That last part is our extrapolation from the mechanism, not something the paper tested. If you hear a number first, answer with what does not fit it. Before you counter, say your own target out loud, or name the other side’s alternatives. Thinking about exactly these things cancelled the anchor in every experiment here, and saying them out loud is the surest way to make sure you do. Rehearse the sentence beforehand. Under pressure, the words that come most easily are the ones the other side just gave you.</p><h3>Next in the series</h3><p>Next: <a href="https://exprea.io/blog/peterson-cannito-brown-1995-voice-characteristics-and-selling-effectiveness">Voice Characteristics and Selling Effectiveness: What Salespeople’s Voices Reveal About Who Sells More</a></p><h3>Source</h3><p>Galinsky, A. D., &amp; Mussweiler, T. (2001). First offers as anchors: the role of perspective-taking and negotiator focus. Journal of Personality and Social Psychology, 81(4), 657–669. <a href="https://doi.org/10.1037/0022-3514.81.4.657">https://doi.org/10.1037/0022-3514.81.4.657</a></p><p>Oesch, J. M., &amp; Galinsky, A. D. (2003). First offers in negotiations: determinants and effects. SSRN working paper 399722. <a href="https://doi.org/10.2139/ssrn.399722">https://doi.org/10.2139/ssrn.399722</a></p><p>Orr, D., &amp; Guthrie, C. (2006). Anchoring, information, expertise, and negotiation: new insights from meta-analysis. Ohio State Journal on Dispute Resolution, 21(3), 597–628.</p><p>Exprea is a community for speakers who want to practise with data and feedback. <a href="https://exprea.io/">exprea.io</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=03241c4a2f8e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/first-offers-as-anchors-why-the-opening-number-sets-the-price-and-how-to-break-its-pull-03241c4a2f8e">First Offers as Anchors: Why the Opening Number Sets the Price, and How to Break Its Pull</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Who Asks and Who Receives: What Negotiating Your Starting Salary Is Worth]]></title>
            <link>https://medium.com/exprea/who-asks-and-who-receives-what-negotiating-your-starting-salary-is-worth-e3d4140bbed6?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/e3d4140bbed6</guid>
            <category><![CDATA[salary]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[career-advice]]></category>
            <category><![CDATA[negotiation]]></category>
            <category><![CDATA[public-speaking]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Mon, 09 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-09T14:31:01.455Z</atom:updated>
            <content:encoded><![CDATA[<p>149 new hires were surveyed. Those who negotiated started on about $5,000 more. How they asked mattered too.</p><h3>TL;DR</h3><ul><li>Marks and Harold surveyed 149 people who had just started new jobs. Those who negotiated their starting salary started on about $5,000 more a year.</li><li>How they negotiated mattered. Competing and collaborating, which put your own position on the table, beat compromising and accommodating.</li><li>The biggest barrier was not skill but temperament: people who disliked risk were the least likely to ask at all.</li><li>It is a survey of recent hires, not an experiment. The full paper is subscription-only, so we use the published abstract and the figures the authors gave to the press.</li></ul><p>Full article with all charts: <a href="https://exprea.io/blog/marks-harold-2011-who-asks-and-who-receives-in-salary-negotiation">exprea.io/blog</a></p><p>Most people accept the first salary they are offered. It feels polite, it avoids an awkward conversation, and the stakes look small: a few thousand a year. Michelle Marks of George Mason University and Crystal Harold of Temple University asked what that choice is actually worth, and what separates the people who ask from the people who do not. This is the first part of our series on communication and wealth.</p><h3>The study in short</h3><p>The paper is Marks and Harold (2011), published in the Journal of Organizational Behavior. The team surveyed 149 newly hired employees across a range of jobs and industries. They asked whether each person negotiated their starting salary, which strategy they used, and what they were paid. They also measured individual differences that might explain who asks, including risk aversion and attitudes towards negotiation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*4GhjsBxA7VBvRRP5.png" /><figcaption>Diagram by Exprea. Data: Marks &amp; Harold 2011 (abstract).</figcaption></figure><h3>What they found</h3><p>Asking paid. People who chose to negotiate increased their starting salary by about $5,000 on average.</p><p>How you asked paid too. The strategies compared are the classic ways people handle a conflict of interest. Two are assertive and open: competing (pressing for your own position) and collaborating (laying out both sides’ interests to find a better deal for both). Two give ground: compromising (meeting in the middle) and accommodating (deferring to the other side). The fifth, avoiding, is not negotiating at all. People who used competing and collaborating strategies, which the authors describe as an open discussion of one’s positions, issues and perspectives, did better than those who compromised or accommodated.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*wAzP5q78_XvLDiLP.png" /><figcaption>Diagram by Exprea. Data: Marks &amp; Harold 2011. The two-axis layout is the standard model of conflict styles.</figcaption></figure><p>Who asked was predictable. Risk aversion and attitudes towards negotiation predicted both whether someone negotiated and which strategy they used. The people most averse to risk were the least likely to ask. The cost of that trait is paid, year after year, in salary. In a summary for strategy+business, the findings were also reported to show that men generally came away with more than women, a gap the authors linked to long-standing pay imbalances. Men and women did not differ significantly in the strategies they used.</p><h3>Why $5,000 is not small</h3><p>A difference in starting salary does not stay the same size. Raises are usually a percentage of what you already earn, so an early gap compounds. Take two people whose starting salaries differ by Δ and who then get the same raise r every year. The gap is Δ in the first year, Δ(1 + r) in the second, Δ(1 + r)² in the third, and so on. The total extra pay after n years is a geometric series. This is our own illustration, in the spirit of the example used when the study was presented to the press.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*EYiL4kn0-wQBOaTL.png" /><figcaption>Equations by Exprea. Our derivation, not a formula from Marks &amp; Harold 2011.</figcaption></figure><p>With Δ = $5,000 and r = 5%, the sum grows like this.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*FxTQ7xjF1lCRENbq.png" /><figcaption>Chart by Exprea. Our calculation for a $5,000 gap and equal 5% raises. Arithmetic, not a finding of the study.</figcaption></figure><p>About $604,000 over a 40-year career, before any interest on the difference. This is arithmetic, not a finding of the study, and real careers are not this tidy.</p><h3>How far to trust it</h3><ul><li>It is a survey, not an experiment. People who negotiated may have differed in ways the survey did not capture: stronger offers, scarcer skills, more options. The $5,000 is what negotiators ended up with, not a guaranteed return on asking.</li><li>Strategies were self-reported, after the fact, by people who knew what they had been paid.</li><li>The sample is small and specific: 149 recent hires in one national labour market.</li><li>Our own limit: we could not read the full paper, which is behind a subscription. We report only the findings the authors published in the abstract and described to the press, and we redraw no tables.</li></ul><h3>What to take from this</h3><p>Salary negotiation is a speaking task. The strategies that paid in this study are the ones that require saying your position out loud, clearly, to someone with more power than you: stating a number, explaining what you bring, asking what the other side needs. The ones that paid least let the other side do the talking. The study also names the real obstacle. It is not that people who stay quiet lack the words. It is that asking feels risky. That kind of discomfort shrinks with rehearsal: saying the sentence out loud, several times, before the conversation that matters.</p><h3>Next in the series</h3><p>Next: <a href="https://exprea.io/blog/galinsky-mussweiler-2001-first-offers-as-anchors">First Offers as Anchors: Why the Opening Number Sets the Price, and How to Break Its Pull</a></p><h3>Source</h3><p>Marks, M., &amp; Harold, C. (2011). Who asks and who receives in salary negotiation. Journal of Organizational Behavior, 32(3), 371–394. <a href="https://doi.org/10.1002/job.671">https://doi.org/10.1002/job.671</a></p><p>Palmquist, M. (2011). Getting a new job? It pays to speak up about salary. strategy+business, 62.</p><p>Exprea is a community for speakers who want to practise with data and feedback. <a href="https://exprea.io/">exprea.io</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e3d4140bbed6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/who-asks-and-who-receives-what-negotiating-your-starting-salary-is-worth-e3d4140bbed6">Who Asks and Who Receives: What Negotiating Your Starting Salary Is Worth</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Surprising Power of Weak Ties: Why Small Conversations With Acquaintances Make Better Days]]></title>
            <link>https://medium.com/exprea/the-surprising-power-of-weak-ties-why-small-conversations-with-acquaintances-make-better-days-a17e5408094c?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/a17e5408094c</guid>
            <category><![CDATA[happiness]]></category>
            <category><![CDATA[social-skills]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[relationships]]></category>
            <category><![CDATA[public-speaking]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Sun, 08 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-08T14:31:01.530Z</atom:updated>
            <content:encoded><![CDATA[<p>On days with more small interactions than usual, people felt happier and more that they belonged. A look at the evidence.</p><h3>TL;DR</h3><ul><li>Psychology mostly studies close relationships. Sandstrom and Dunn asked about everyone else: classmates, neighbours, baristas and colleagues we only half know.</li><li>Students felt happier, and more that they belonged, on days when they spoke to more classmates than usual.</li><li>Counting every interaction in a day, the casual ones still went with better social and emotional well-being.</li><li>A companion experiment found that chatting with a barista, instead of being efficient, lifted mood and belonging.</li><li>A 2021 study in Turkey found that even greeting and thanking a shuttle driver lifted mood.</li></ul><p>Full article with all charts: <a href="https://exprea.io/blog/sandstrom-dunn-2014-social-interactions-and-well-being-weak-ties">exprea.io/blog</a></p><p>Most research on relationships and happiness is about the people closest to us: partners, family, best friends. Gillian Sandstrom and Elizabeth Dunn noticed that this leaves out most of the people we actually talk to on a given day. The sociologist’s term for them is weak ties: acquaintances, the familiar faces we know a little. This is the last part of our series on communication and social life.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*AAJ0tyzvW2lr1h6J.png" /><figcaption>Diagram by Exprea. Solid lines are strong ties, dashed lines are weak ties. The examples are ours.</figcaption></figure><h3>The study in short</h3><p>The paper is Sandstrom and Dunn (2014), published in Personality and Social Psychology Bulletin. In Study 1, university students reported, day by day, how many classmates they had interacted with, however briefly, together with their happiness and sense of belonging. In Studies 2A and 2B, students and members of the community counted all their daily interactions, with close and casual ties alike, including simple greetings. The design uses repeated daily reports, so each person is compared with themselves on busier and quieter days.</p><h3>What the researchers did</h3><p>Daily reports allow a sharper question than a one-off survey. Instead of asking whether sociable people are happier, which could just mean that happy people are sociable, the analysis can ask whether the same person is happier on days when they talk to more acquaintances than they usually do. The standard way is to split each day’s count into the person’s own average and that day’s departure from it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*W3r7dQ-fWj-WuMbH.png" /><figcaption>Equations by Exprea. Data: Sandstrom &amp; Dunn 2014 (standard multilevel model for daily diaries).</figcaption></figure><h3>What they found</h3><p>In the first study, students had more happiness and a stronger sense of belonging on days when they interacted with more classmates than usual. In the second pair of studies, which counted every interaction, weak ties were again linked to both social and emotional well-being, not only the close relationships that research usually centres on.</p><p>A companion experiment, published the same year, gave the causal half. Customers at a coffee shop were asked either to have a genuine interaction with the barista (smile, make eye contact, have a brief chat) or to be as efficient as possible (money ready, no unnecessary talk). Those who had the genuine interaction left in a better mood and with a stronger feeling of belonging (Sandstrom and Dunn, 2014b).</p><h3>Has it held up?</h3><p>The idea has been tested again at the smallest possible scale. In Turkey, Gul Gunaydin and colleagues studied people commuting on university shuttle buses. Among 856 commuters, those who more often greeted, thanked or wished their driver well reported higher life satisfaction and more positive feelings, even after accounting for personality and age. In an experiment with 265 more people, those asked to greet, thank or wish the driver well, mostly with about three words, felt better after the ride than those asked not to speak to the driver. Silence was the control because in Turkey the norm is not to talk to shuttle drivers, so as not to distract them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Yvjy4i_IB09SSiVJ.png" /><figcaption>Chart by Exprea. Data: Gunaydin et al. 2021, Study 2.</figcaption></figure><p>Put next to the other studies in this series, the picture goes from correlation to experiment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*m1U-yE178pAbqvTz.png" /><figcaption>Table by Exprea. Data: Sandstrom &amp; Dunn 2014, Epley &amp; Schroeder 2014, Gunaydin et al. 2021.</figcaption></figure><h3>How far to trust it</h3><ul><li>Daily reports are still correlational. A good day might lead people to talk to more classmates, rather than the talking causing the good day. The companion experiment and later studies address this, but this paper alone cannot.</li><li>Counts of interactions are self-reported, and a person in a good mood may notice or remember more of them.</li><li>The samples were students and a community group in one country. The Turkish study extends this to a different culture.</li><li>Our own limit: the paper is subscription-only. We describe its method from its abstract and from later research, and we do not reproduce its own statistics.</li></ul><h3>What to take from this</h3><p>We think of speaking as something that matters in big moments: the presentation, the interview, the difficult conversation. This research points at the small ones. A greeting that becomes a sentence, a question to the person at the counter, a real exchange with the colleague you normally nod to. These are not networking in any strategic sense. They are simply more moments in the day when you speak to someone and they speak back, and on the evidence, days with more of them are better days. They are also the easiest place to practise the thing that makes bigger conversations easier: starting.</p><h3>Next in the series</h3><p>This was the last study in the social series. Next we move to speaking and money: <a href="https://exprea.io/blog/marks-harold-2011-who-asks-and-who-receives-in-salary-negotiation">Who Asks and Who Receives: What Negotiating Your Starting Salary Is Worth</a></p><h3>Source</h3><p>Sandstrom, G. M., &amp; Dunn, E. W. (2014). Social interactions and well-being: the surprising power of weak ties. Personality and Social Psychology Bulletin, 40(7), 910–922. <a href="https://doi.org/10.1177/0146167214529799">https://doi.org/10.1177/0146167214529799</a></p><p>Gunaydin, G., Oztekin, H., Karabulut, D. H., &amp; Salman-Engin, S. (2021). Minimal social interactions with strangers predict greater subjective well-being. Journal of Happiness Studies, 22(4), 1839–1853. <a href="https://doi.org/10.1007/s10902-020-00298-6">https://doi.org/10.1007/s10902-020-00298-6</a></p><p>Sandstrom, G. M., &amp; Dunn, E. W. (2014b). Is efficiency overrated? Minimal social interactions lead to belonging and positive affect. Social Psychological and Personality Science, 5(4), 437–442. <a href="https://doi.org/10.1177/1948550613502990">https://doi.org/10.1177/1948550613502990</a></p><p>Exprea is a community for speakers who want to practise with data and feedback. <a href="https://exprea.io/">exprea.io</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a17e5408094c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/the-surprising-power-of-weak-ties-why-small-conversations-with-acquaintances-make-better-days-a17e5408094c">The Surprising Power of Weak Ties: Why Small Conversations With Acquaintances Make Better Days</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Mistakenly Seeking Solitude: Why Talking to Strangers Is Better Than We Expect]]></title>
            <link>https://medium.com/exprea/mistakenly-seeking-solitude-why-talking-to-strangers-is-better-than-we-expect-6890ad66b02b?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/6890ad66b02b</guid>
            <category><![CDATA[public-speaking]]></category>
            <category><![CDATA[social-skills]]></category>
            <category><![CDATA[communication]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[relationships]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Sat, 07 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-07T14:31:01.679Z</atom:updated>
            <content:encoded><![CDATA[<p>On trains, buses and in taxis, talking to a stranger beat staying silent every time. People predicted the opposite.</p><h3>TL;DR</h3><ul><li>Epley and Schroeder asked people on Chicago trains, buses and taxis to talk to a stranger, keep to themselves, or do as they normally would.</li><li>Talking gave the more positive trip every time, and it cost no productivity.</li><li>People who only predicted the trips expected the opposite: that solitude would be better.</li><li>Commuters thought fewer than half of strangers would want to talk. Nobody who tried was rebuffed.</li><li>Being talked to was as pleasant as doing the talking.</li></ul><p>Full article with all charts: <a href="https://exprea.io/blog/epley-schroeder-2014-mistakenly-seeking-solitude">exprea.io/blog</a></p><p>Commuter trains are full of people sitting inches apart in silence. The obvious explanation is that silence is what they prefer. Nicholas Epley and Juliana Schroeder, at the University of Chicago’s Booth School of Business, tested whether that is true. This is part of our series on communication and social life.</p><h3>The study in short</h3><p>The paper is Epley and Schroeder (2014), published in the Journal of Experimental Psychology: General. There were several experiments. On trains, 97 Chicago commuters did the task and 66 others predicted it. On buses, 75 did it and 61 predicted it. In taxis from Midway airport, 93 did it and 43 predicted it. A further 64 train and 65 bus commuters were asked why, and 179 people in pairs took part in a laboratory waiting room. People rated how happy, how sad and how pleasant the trip was, and how productive.</p><h3>What people expect</h3><p>The authors started with a survey. Asked whether they would talk to a friend in a waiting room, on a train, on a plane or in a taxi, almost nobody said no. Asked the same about a stranger, most people said they would stay silent. The taxi stands out because there a stranger, the driver, has to be spoken to.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*8CbDyDcwrXAEv252.png" /><figcaption>Chart by Exprea. Data: Epley &amp; Schroeder 2014. Share who said they would not talk, one condition of a 203-person survey.</figcaption></figure><h3>What actually happens</h3><p>Then they ran it for real. Commuters were randomly told to make a connection with a stranger, to enjoy their solitude, or to commute as normal. At the end they rated the trip. The ratings were combined into one positivity score, and the difference between groups is reported as a standardised effect size, Cohen’s d.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*sVNcrmdmjYdpK7ec.png" /><figcaption>Equations by Exprea. Data: Epley &amp; Schroeder 2014.</figcaption></figure><p>Talking won on every kind of transport, and the predictions got it backwards. Every prediction from people who rarely talk to strangers said talking would be worse. The only group whose prediction was right were taxi riders who usually chat with their driver: the people with experience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*3FHRXjciPzqE6RgL.png" /><figcaption>Chart by Exprea. Data: Epley &amp; Schroeder 2014. Green is experienced, grey is predicted.</figcaption></figure><p>Talking cost nothing in productivity. On trains, buses and in taxis, the connection and solitude groups rated their trips as almost equally productive. The commuters asked to predict it expected talking to cost them productivity. On the train, conversations lasted 14 minutes on average and were rated pleasant (5.1 out of 6). The longer the conversation, the better the commute.</p><h3>Why people get it wrong</h3><p>If talking is better, why does nobody learn it? The authors tested two explanations. One is that people remember their worst conversations with strangers and expect more of the same. The other is that people assume others are not interested, so they never try, and never find out. The second fits the data. Commuters rated their own interest in talking higher than they thought other people’s interest was. They predicted that fewer than half of strangers would be willing to talk. In reality, every participant who tried to start a conversation had one. Not one was rebuffed. When people simply imagined a conversation, their predictions sat close to the positive version, which is evidence against the bad-memories idea.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*f5bVaW01JBIWmOdy.png" /><figcaption>Table by Exprea. Data: Epley &amp; Schroeder 2014.</figcaption></figure><p>The result is a loop that keeps itself going. Psychologists call it pluralistic ignorance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*s7dqF6p4cFLtvK5p.png" /><figcaption>Diagram by Exprea. Data: Epley &amp; Schroeder 2014.</figcaption></figure><h3>Being talked to is just as good</h3><p>A fair worry is that only the person who starts the chat benefits, and the stranger would rather be left alone. The waiting-room experiment tested it. One person in each pair was told to talk or to keep to themselves. The other was told nothing. Connection was more positive than solitude for both people, the one who started and the one who was talked to. It made no difference if the instruction was optional or mandatory. People alone in a room did no better than people sitting in silence next to a stranger (−0.18 against −0.25 on the positivity index). Both did worse than people who were talked to (0.33).</p><h3>Has it held up?</h3><p>The core finding has. In a field experiment with London commuters, people assigned to talk to a stranger again had a more positive journey, and reported learning more, than those assigned to solitude or a normal commute (Schroeder, Lyons and Epley, 2022). One part changed. The London commuters expected talking to be pleasant too. What they got wrong was the start of the conversation, which they expected to be harder than it was, because they thought others would not be interested. A mini meta-analysis of seven studies and 2,304 people (Sandstrom and Boothby, 2021) reached the same place: fears about talking to strangers are overblown, and conversations go better than expected.</p><h3>How far to trust it</h3><ul><li>The experiments lasted one trip. The authors found no lasting effect on how people rated their whole day, so the benefit is real but may be brief.</li><li>Participants were instructed. The waiting-room study addresses this, but a conversation you choose may feel different from one you were asked to have.</li><li>Chicago in the early 2010s is one city and one moment. The London replication helps.</li><li>The prediction groups were different people from the ones who did the task. That is the right design for comparing expectation and experience, but no single person was shown their own error.</li></ul><h3>What to take from this</h3><p>The hardest part of talking to a stranger is the first sentence, and this study says the fear attached to it is miscalibrated. People are more willing to talk than they look. The conversation usually goes well. The person you talk to tends to enjoy it as much as you do. That makes the opening line the skill worth practising: a short, easy question about something you are both looking at, said out loud without apology. Once it is said, the evidence suggests the rest mostly takes care of itself.</p><h3>Next in the series</h3><p>Next: <a href="https://exprea.io/blog/sandstrom-dunn-2014-social-interactions-and-well-being-weak-ties">The Surprising Power of Weak Ties: Why Small Conversations With Acquaintances Make Better Days</a></p><h3>Source</h3><p>Epley, N., &amp; Schroeder, J. (2014). Mistakenly seeking solitude. Journal of Experimental Psychology: General, 143(5), 1980–1999. <a href="https://doi.org/10.1037/a0037323">https://doi.org/10.1037/a0037323</a></p><p>Schroeder, J., Lyons, D., &amp; Epley, N. (2022). Hello, stranger? Pleasant conversations are preceded by concerns about starting one. Journal of Experimental Psychology: General, 151(5), 1141–1153. <a href="https://doi.org/10.1037/xge0001118">https://doi.org/10.1037/xge0001118</a></p><p>Sandstrom, G. M., &amp; Boothby, E. J. (2021). Why do people avoid talking to strangers? Self and Identity, 20(1), 47–71. <a href="https://doi.org/10.1080/15298868.2020.1816568">https://doi.org/10.1080/15298868.2020.1816568</a></p><p>Exprea is a community for speakers who want to practise with data and feedback. <a href="https://exprea.io/">exprea.io</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6890ad66b02b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/mistakenly-seeking-solitude-why-talking-to-strangers-is-better-than-we-expect-6890ad66b02b">Mistakenly Seeking Solitude: Why Talking to Strangers Is Better Than We Expect</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Deep Conversations, Small Talk and Happiness: What 23,689 Recordings Show]]></title>
            <link>https://medium.com/exprea/deep-conversations-small-talk-and-happiness-what-23-689-recordings-show-b4a6f36d1f22?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/b4a6f36d1f22</guid>
            <category><![CDATA[communication]]></category>
            <category><![CDATA[happiness]]></category>
            <category><![CDATA[relationships]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[public-speaking]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Fri, 06 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-06T14:31:01.606Z</atom:updated>
            <content:encoded><![CDATA[<p>79 students carried a recorder for four days. The happiest talked more, and went deeper. One finding did not replicate.</p><h3>TL;DR</h3><ul><li>Mehl and colleagues did not ask people how they spend their days. They listened. 79 students wore a recorder for four days.</li><li>Happier people spent less time alone (r = −.35) and more time talking (r = .31).</li><li>The happiest had about twice the share of deep conversations of the unhappiest: 45.9% against 21.8%.</li><li>A 2018 replication confirmed the link with time talking and with deep talk. The small-talk finding did not hold.</li><li>The design is correlational. Happy people may attract deeper talk, not only enjoy it.</li></ul><p>Full article with all charts: <a href="https://exprea.io/blog/mehl-2010-eavesdropping-on-happiness-small-talk-substantive-conversations">exprea.io/blog</a></p><p>Most of what we know about how people spend their days comes from asking them. That has a problem. Few people will say that their conversations are shallow. Matthias Mehl and colleagues solved it by recording what people really said, in small, random slices of ordinary life. This is part of our series on communication and social life.</p><h3>The study in short</h3><p>The paper is Mehl, Vazire, Holleran and Clark (2010), published in Psychological Science. 79 undergraduates (47 women) wore the Electronically Activated Recorder, or EAR. It is a pocket recorder that switched itself on for 30 seconds every 12.5 minutes, for four days. That gave 23,689 recordings, about 300 per person. Coders listened to each clip. Was the person alone or talking? Was the talk small talk or a substantive conversation? The students also reported their life satisfaction and happiness, and people who knew them rated their happiness too. The Big Five personality traits were measured to test whether personality explained the results.</p><h3>How you eavesdrop on a day</h3><p>The recorder samples about 4% of waking time, and nobody has to decide what to record. Small talk was defined as an uninvolved chat where only trivial information changes hands. The paper’s example is a remark about someone’s popcorn. A substantive conversation was an involved one where meaningful information is exchanged, for example a talk about someone’s parents falling in love and then divorcing. Across everyone, 17.9% of conversations were small talk and 35.5% were substantive. The rest were neither.</p><p>Each day became a few proportions. Time alone and time talking are shares of all recordings. Small talk and substantive talk are shares of conversations only. That way, people who talk more are not automatically counted as deeper.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*EMss-y5647K6PW9v.png" /><figcaption>Equations by Exprea. Data: Mehl et al. 2010.</figcaption></figure><h3>What they found</h3><p>Well-being went with more company and more depth. The happier the person, the less time alone (r = −.35) and the more time talking (r = .31). Of their conversations, a smaller share was small talk (r = −.33) and a larger share was substantive (r = .28). To make the correlations concrete, the authors compared the happiest participants (1.5 standard deviations above average) with the unhappiest (2 below). The happiest spent about a quarter less time alone, 70% more time talking, had about a third as much small talk, and twice the share of substantive talk.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*FrV0BVuXVFmq6v6v.png" /><figcaption>Chart by Exprea. Data: Mehl et al. 2010. The happiest (1.5 SD above average) against the unhappiest (2 SD below).</figcaption></figure><p>The results held on weekdays and weekends, and for each well-being measure alone. An extravert might be both happier and chattier, so the authors removed everything the Big Five predict about well-being. The links with time alone, time talking and substantive talk stayed about the same. The link with small talk got weaker. People who were happier than their personality predicts had more conversations, and deeper ones.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*X7-B2mlYSp3dU5pI.png" /><figcaption>Table by Exprea. Data: Mehl et al. 2010.</figcaption></figure><h3>Has it held up?</h3><p>Partly, and the partial answer is the interesting one. In 2018 the same lab pooled the original data with three new samples, from cancer patients and their spouses to adults in a meditation trial and recently divorced adults. That made 486 people. Time with others and substantive talk replicated, at somewhat smaller sizes. The small-talk result did not. The authors concluded that the original negative correlation, from an underpowered study, was probably a false positive. The replication also found no convincing sign that personality changes the pattern. For example, introverts did not seem to gain less from deep talk.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*LFP02AqHZCuKtETm.png" /><figcaption>Chart by Exprea. Data: Mehl et al. 2010 and Milek et al. 2018.</figcaption></figure><h3>How far to trust it</h3><ul><li>The study cannot tell which way it runs. Happy people may attract deeper talk, or deep talk may make people happier, or both.</li><li>79 students is a small and narrow sample. That is why one of four headline findings did not survive a larger test.</li><li>What counts as substantive is a coder’s judgment. It was reliable here (agreement .84), but it is still a category set from outside.</li><li>Four days of 30-second clips is a sample of a life, not a record of it.</li></ul><h3>What to take from this</h3><p>The lasting finding is about depth, not volume. People who were more satisfied with life spent more of their conversations on something that mattered. That held up when the lab checked again with six times as many people. It also removes a worry. Small talk is not the enemy of a good life. In the larger data it made no difference either way. It is just not where the benefit lives. The move worth practising is the one that turns a trivial exchange into a real one: a follow-up question, an honest answer, a story instead of a status update.</p><h3>Next in the series</h3><p>Next: <a href="https://exprea.io/blog/epley-schroeder-2014-mistakenly-seeking-solitude">Mistakenly Seeking Solitude: Why Talking to Strangers Is Better Than We Expect</a></p><h3>Source</h3><p>Mehl, M. R., Vazire, S., Holleran, S. E., &amp; Clark, C. S. (2010). Eavesdropping on happiness: well-being is related to having less small talk and more substantive conversations. Psychological Science, 21(4), 539–541. <a href="https://doi.org/10.1177/0956797610362675">https://doi.org/10.1177/0956797610362675</a></p><p>Replication: Milek, A., et al. (2018). “Eavesdropping on happiness” revisited. Psychological Science, 29(9), 1451–1462. <a href="https://doi.org/10.1177/0956797618774252">https://doi.org/10.1177/0956797618774252</a></p><p>Exprea is a community for speakers who want to practise with data and feedback. <a href="https://exprea.io/">exprea.io</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b4a6f36d1f22" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/deep-conversations-small-talk-and-happiness-what-23-689-recordings-show-b4a6f36d1f22">Deep Conversations, Small Talk and Happiness: What 23,689 Recordings Show</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Liking Gap: After a Conversation, People Like You More Than You Think]]></title>
            <link>https://medium.com/exprea/the-liking-gap-after-a-conversation-people-like-you-more-than-you-think-eba68c5c0b0e?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/eba68c5c0b0e</guid>
            <category><![CDATA[relationships]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[communication]]></category>
            <category><![CDATA[public-speaking]]></category>
            <category><![CDATA[social-skills]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Thu, 05 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-05T14:31:01.598Z</atom:updated>
            <content:encoded><![CDATA[<p>In five studies, people kept underestimating how much their partner liked them. The signs were visible, but they missed them.</p><p><strong>TL;DR</strong></p><ul><li>After a five-minute chat, people rated their partner 5.82 out of 7 on liking. They thought their partner rated them 5.17. That difference is the liking gap.</li><li>Observers watching the videos could tell how much people liked each other. The participants themselves could not.</li><li>The gap survived long conversations and lasted months among suitemates, closing only at the end of the year.</li><li>It was larger in shyer people, and it was carried by negative thoughts about one’s own performance.</li></ul><p>The full article: <a href="https://exprea.io/blog/boothby-2018-the-liking-gap-in-conversations">https://exprea.io/blog/boothby-2018-the-liking-gap-in-conversations</a></p><p>This is the first paper in our series on communication and social life. Everyone knows the feeling of walking away from a first conversation and replaying it. Did I talk too much? Was I boring? Did that joke land? Erica Boothby, Gus Cooney, Gillian Sandstrom and Margaret Clark asked whether that nagging doubt is accurate. Their answer, across five studies, is that it is not.</p><h3>The study in short</h3><ul><li><strong>Study 1a:</strong> 36 adults from the community, paired with a same-sex stranger for a five-minute conversation using ice-breaker questions.</li><li><strong>Study 1b:</strong> two coders, blind to the hypothesis, watched 15 of those videos and rated how much each person liked the other.</li><li><strong>Study 2:</strong> 84 students and recent graduates, five minutes of free conversation, plus the three moments each person thought shaped the impressions.</li><li><strong>Study 3:</strong> 102 people who talked for as long as they liked, anywhere from 2 to 45 minutes.</li><li><strong>Study 4:</strong> 100 members of the public at “How to Talk to Strangers” workshops in the UK, before and after a five-minute introduction.</li><li><strong>Study 5:</strong> 102 first-year students rating their suitemates five times across an academic year.</li></ul><h3>How you measure a mistake about someone else’s mind</h3><p>The design is simple. Two strangers talk. Then, separately, each answers two sets of questions: how much they liked the other person, and how much they think the other person liked them.</p><p>The logic that turns this into evidence of an error is the elegant part. In a pair, the liking I report giving is exactly the liking my partner receives, and the reverse. So averaged across everyone, the liking people give must equal the liking people receive. If people’s estimates of the liking they received are lower on average, those estimates are wrong on average. People are underestimating how much they were liked.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ZZnORQdOSX90QTaF.png" /><figcaption>Equations by Exprea. Data: Boothby et al. 2018.</figcaption></figure><h3>Five studies, one gap</h3><p>In the first study, people liked their partner 5.82 out of 7 on average, and thought their partner liked them 5.17. That difference, which cannot be real, is the liking gap. It appeared in every study that followed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*2_SjzwN1Af1-YCMr.png" /><figcaption>Chart by Exprea. Data: Boothby et al. 2018.</figcaption></figure><p>Every interval sits clear of zero. Studies 1 to 3 and 5 use 7-point scales. Study 4 asked how interesting each person found the other on a 5-point scale, so its gaps are on a different footing. In Study 4 the gap was there even before anyone spoke, and nearly doubled after they did.</p><h3>The signals were there; people missed them</h3><p>Why would people misjudge this? One possibility is that people hide their liking, so there is nothing to read. The other is that the liking is on display and goes unnoticed. To separate the two, the authors had two coders who did not know the hypothesis watch the videotaped conversations. They rated how much each person liked the other. If people hide their liking, observers should not see it either.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*vfeO3GbyLeucc7JJ.png" /><figcaption>Diagram by Exprea. Data: Boothby et al. 2018.</figcaption></figure><p>Observers’ ratings predicted how much people actually liked their partners, but not how much people thought they were liked. The liking was visible from outside. The people receiving it did not use it. The authors call this the neglected-signal account.</p><p>The second study showed what people were attending to instead. Asked to name the three moments that shaped each impression, people rated the moments they thought shaped their partner’s view of them as more negative than the moments that shaped their own view of the partner. That difference in tone accounted for part of the gap (indirect effect b = -0.14, 95% CI [-0.22, -0.07]).</p><p>In short, during a conversation you have a running commentary on your own performance, and the other person does not. They got the point, laughed at the joke, and thought it was fine. You remember the story you could have told better.</p><h3>Who has the biggest gap</h3><p>Shyness made it worse. In the first study, the shyest third of participants showed a large gap, the middle third a smaller one, and the least shy third none that could be told apart from zero.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*SJF3bm8WcOVMgu4a.png" /><figcaption>Chart by Exprea. Data: Boothby et al. 2018.</figcaption></figure><p>The authors point out that this also argues against the gap being mere modesty. A pure reporting habit would have no reason to track shyness. Rejection sensitivity, self-esteem and narcissism did not change the size of the gap.</p><h3>Longer conversations do not fix it</h3><p>A natural guess is that the gap is a first-minute effect that disappears once people really talk. In the third study, people could talk for as long as they wanted, up to 45 minutes. Longer conversations produced more liking and more enjoyment all round. But the gap between what people felt and what they thought their partner felt was about the same whether the conversation was short, medium or long.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*CtndbCXyAOoiftrE.png" /><figcaption>Chart by Exprea. Data: Boothby et al. 2018.</figcaption></figure><p>Among suitemates the gap lasted for months. It was significant in September, October, December and February, and disappeared only by May, when the students were deciding who to live with the next year. Offering a $100 prize for accurate estimates made no difference, which suggests people really believed their low estimates.</p><h3>Has it held up?</h3><p>The liking gap has since turned up in new settings. The same group later found it in small groups and in engineering project teams, and linked it to how well teams of working adults were able to work together. A study of 241 children found it emerging at about the age of five, when children begin to care about how others see them, and growing with age.</p><h3>How far to trust it</h3><ul><li><strong>Most participants were students,</strong> though the workshop study used members of the public in the UK.</li><li><strong>Liking is self-reported on both sides.</strong> The design depends on people reporting honestly. The observers’ ratings and the ignored cash incentive both support that.</li><li><strong>The gap is an average.</strong> Some people overestimate how much they are liked. The finding is that, on average, people err on the pessimistic side.</li><li><strong>The numbers come from the authors’ accepted manuscript,</strong> which may differ slightly from the published version.</li></ul><h3>What to take from this</h3><ul><li>Remember this study after a networking event, a first meeting or a talk that felt flat from the inside. The critic in your head is reviewing a performance the audience did not see.</li><li>They heard the point, not the sentence you would rephrase.</li><li>If you are liked more than you know, following up costs less than it feels. Send the message, suggest the coffee, ask the question.</li></ul><p>Next in the series: <a href="https://exprea.io/blog/mehl-2010-eavesdropping-on-happiness-small-talk-substantive-conversations">Eavesdropping on Happiness: Deep Conversations, Small Talk and Well-Being</a>.</p><h3>Source</h3><p>Boothby, E. J., Cooney, G., Sandstrom, G. M., &amp; Clark, M. S. (2018). The liking gap in conversations: do people like us more than we think? <em>Psychological Science, 29</em>(11), 1742–1756. <a href="https://doi.org/10.1177/0956797618783714">https://doi.org/10.1177/0956797618783714</a></p><p>Want to see your own numbers? Record a short clip at <a href="https://exprea.io/">exprea.io</a>.</p><p>Read the full version on our site: <a href="https://exprea.io/blog/boothby-2018-the-liking-gap-in-conversations">exprea.io/blog/boothby-2018-the-liking-gap-in-conversations</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=eba68c5c0b0e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/the-liking-gap-after-a-conversation-people-like-you-more-than-you-think-eba68c5c0b0e">The Liking Gap: After a Conversation, People Like You More Than You Think</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Social Relationships and Mortality Risk: The 148-Study Meta-Analysis Behind “Loneliness Is Like…]]></title>
            <link>https://medium.com/exprea/social-relationships-and-mortality-risk-the-148-study-meta-analysis-behind-loneliness-is-like-9107efd2ca6b?source=rss-7ead78d50714------2</link>
            <guid isPermaLink="false">https://medium.com/p/9107efd2ca6b</guid>
            <category><![CDATA[mental-health]]></category>
            <category><![CDATA[loneliness]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[public-speaking]]></category>
            <category><![CDATA[relationships]]></category>
            <dc:creator><![CDATA[Noam Shemla]]></dc:creator>
            <pubDate>Wed, 04 Nov 2026 14:31:01 GMT</pubDate>
            <atom:updated>2026-11-04T14:31:01.701Z</atom:updated>
            <content:encoded><![CDATA[<h3>Social Relationships and Mortality Risk: The 148-Study Meta-Analysis Behind “Loneliness Is Like Smoking”</h3><p>Across 308,849 people, stronger social ties went with 50% higher odds of survival. Here is what that does and does not mean.</p><p><strong>TL;DR</strong></p><ul><li>Holt-Lunstad, Smith and Layton pooled 148 studies that followed 308,849 people for an average of 7.5 years. People with stronger social relationships had 50% higher odds of being alive at the end of follow-up (odds ratio 1.50).</li><li>Rich measures of social integration showed the strongest link (1.91). Simply living alone versus with others showed the weakest.</li><li>The link was as strong as some well-known risk factors, including moderate smoking, and stronger than physical inactivity or obesity.</li><li>The studies are observational. They show a strong association, not proof of cause.</li></ul><p>The full article: <a href="https://exprea.io/blog/holt-lunstad-2010-social-relationships-and-mortality-risk-meta-analysis">https://exprea.io/blog/holt-lunstad-2010-social-relationships-and-mortality-risk-meta-analysis</a></p><p>In the last paper, <a href="https://exprea.io/blog/anderson-2013-virtual-reality-exposure-therapy-social-anxiety-public-speaking">a trial of virtual reality exposure therapy</a>, we saw how repeated speaking lowers fear. Now we step back to a bigger question. Do relationships, built mostly through conversation, matter for more than mood?</p><p>This is the paper usually behind the claim that loneliness is as bad for you as smoking. Julianne Holt-Lunstad, Timothy Smith and Bradley Layton measured, across every suitable study they could find, how strongly people’s social relationships predict whether they die.</p><h3>The study in short</h3><ul><li><strong>Studies:</strong> 148 prospective studies, searched from 1900 to 2007, of how people’s social relationships related to whether they later died.</li><li><strong>People:</strong> 308,849 participants. Average age 63.9 at the start, half women, mostly from North America and Europe.</li><li><strong>Follow-up:</strong> 7.5 years on average (from 3 months to 58 years). 29% of participants died during their study’s follow-up.</li><li><strong>Measures:</strong> structural (marriage, networks, living alone, integration), functional (support received or perceived, loneliness), or both.</li><li><strong>Method:</strong> random-effects meta-analysis of log odds ratios, with metaregression to test what changed the size of the link.</li></ul><h3>What a meta-analysis does</h3><p>Each study measured people’s relationships at the start, then counted who had died some years later. The authors converted each result into the same unit, the natural log of an odds ratio, and averaged them. More precise studies got more weight. Because the studies differed in who they followed and how they measured relationships, the authors used a random-effects model. It assumes the true effect varies from study to study and estimates both the average and the spread.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Jkw6YERmwFZFNwwQ.png" /><figcaption>Equations by Exprea. Data: Holt-Lunstad et al. 2010.</figcaption></figure><h3>The headline result</h3><p>Across the 148 studies the pooled odds ratio was 1.50 (95% CI 1.42 to 1.59). People with stronger social relationships had 50% higher odds of surviving the follow-up period. Individual studies ranged from 0.77 to 6.50. They disagreed more than chance alone would explain, so the authors looked for what made the effect larger or smaller. Only nine of the 148 studies found lower survival with stronger ties. The authors’ own tests found no sign of publication bias.</p><h3>Which relationships matter most</h3><p>Not every way of measuring relationships predicted survival equally well. Complex measures of social integration combine things like marriage, the size of someone’s network and how much they take part in social life. They showed the strongest link. A single yes-or-no question about living alone showed the weakest, and its interval includes no effect at all.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*G5cotpAZjDusEjNY.png" /><figcaption>Chart by Exprea. Data: Holt-Lunstad et al. 2010.</figcaption></figure><p>The metaregression tested whether the link depended on who was studied. It did not depend on age, sex, initial health, cause of death, length of follow-up or region. What did change it was how relationships were measured, and whether the estimate was adjusted for other variables. Adjusted estimates were smaller than raw ones.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*P5quX9mkttcNuRI3.png" /><figcaption>Table by Exprea. Data: Holt-Lunstad et al. 2010.</figcaption></figure><h3>As strong as smoking?</h3><p>The paper’s most quoted figure puts social relationships beside other risks, using effect sizes from other meta-analyses. Social relationships sit alongside smoking fewer than 15 cigarettes a day, and above physical inactivity and obesity.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*0YCP-MslYqlsA6ou.png" /><figcaption>Chart by Exprea. Data: Holt-Lunstad et al. 2010 (Figure 6, approximate).</figcaption></figure><p>The values for the other risks are read from the paper’s figure, so treat the second decimal as approximate. The two social values the paper states exactly, 1.50 and 1.91, match. Comparing effects from different meta-analyses with different designs is a rough guide, not a ranking.</p><h3>What 50% higher odds does and does not mean</h3><p>An odds ratio of 1.5 is not the same as 50% more people surviving. Odds compare survivors with non-survivors, so the translation to people depends on how many survived in the first place.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*E7eMVbLXJVIOGRbu.png" /><figcaption>Equations by Exprea. Standard conversion from an odds ratio to a probability.</figcaption></figure><p>If half of the people with weaker ties are still alive at some point, an odds ratio of 1.5 means about 60% of those with stronger ties are. That is ten more people in every hundred, not fifty.</p><h3>Has it held up?</h3><p>The authors extended the work in 2015 with a meta-analysis focused on the absence of relationships. Counting only studies that adjusted for possible confounds, social isolation was associated with 29% higher odds of death, loneliness with 26% and living alone with 32%. Objective isolation and the feeling of loneliness mattered about equally. The links were, if anything, stronger in people under 65.</p><h3>How far to trust it</h3><ul><li><strong>It does not show cause.</strong> Every study was observational. Sick people may lose social contact, and some other factor, such as income or personality, may drive both. Adjusted estimates were smaller than raw ones, which is what confounding would produce.</li><li><strong>It does not say which relationships to build.</strong> The strongest measures were the broadest. It is the overall web of relationships, not any one of them, that tracks with survival.</li><li><strong>It does not show the mechanism.</strong> The authors discuss two: relationships buffering stress, and relationships encouraging healthier behaviour. The meta-analysis cannot choose between them.</li><li><strong>The comparison with smoking is approximate.</strong> It sets effect sizes from different meta-analyses, with different designs, side by side.</li></ul><h3>What to take from this</h3><ul><li>Relationships are built mostly in conversation. This is strong evidence that they matter for more than mood.</li><li>The measures that predicted survival best were not about one close tie. They were about a life woven into many: people seen, groups belonged to, roles held.</li><li>Every one of those starts with someone speaking to someone else. The skills that make a talk land are the same ones that turn acquaintances into the kind of network this paper measured.</li></ul><p>This is the last paper in the Communication and Health series.</p><h3>Source</h3><p>Holt-Lunstad, J., Smith, T. B., &amp; Layton, J. B. (2010). Social relationships and mortality risk: a meta-analytic review. <em>PLoS Medicine, 7</em>(7), e1000316. <a href="https://doi.org/10.1371/journal.pmed.1000316">https://doi.org/10.1371/journal.pmed.1000316</a></p><p>Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T., &amp; Stephenson, D. (2015). Loneliness and social isolation as risk factors for mortality: a meta-analytic review. <em>Perspectives on Psychological Science, 10</em>(2), 227–237. <a href="https://doi.org/10.1177/1745691614568352">https://doi.org/10.1177/1745691614568352</a></p><p>Want to see your own numbers? Record a short clip at <a href="https://exprea.io/">exprea.io</a>.</p><p>Read the full version on our site: <a href="https://exprea.io/blog/holt-lunstad-2010-social-relationships-and-mortality-risk-meta-analysis">exprea.io/blog/holt-lunstad-2010-social-relationships-and-mortality-risk-meta-analysis</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9107efd2ca6b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/exprea/social-relationships-and-mortality-risk-the-148-study-meta-analysis-behind-loneliness-is-like-9107efd2ca6b">Social Relationships and Mortality Risk: The 148-Study Meta-Analysis Behind “Loneliness Is Like…</a> was originally published in <a href="https://medium.com/exprea">Exprea</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>