Research on social decisions and gender bias

There is a strong belief that women are more generous, and care more about equality, than men. Also, many people believe that women tend to be underconfident. None of these beliefs are true, two recent papers find. One paper —nicely titled "Men are from Mars, and Women Too"— focuses on self-confidence. The paper starts by surveying hundreds of experts, finding that 77% of them think that women are underconfident whereas men are overconfident. Next they do a comprehensive survey of all experimental tests of self-confidence that have been published in the past twenty years. The key result: 72% of all studies find that *both* women and men are overconfident. Only 18% of published studies support the commonly held view. The other paper considers social preferences. "Using data from 15 studies and 8,979 individuals" the paper finds "that women are believed to be more generous and more equality-oriented than men. [...] Yet this believed gender gap is largely inaccurate." The authors "find little to no evidence for gender differences in behavior or attitudes relating to social preferences". Hence, women and men are much more similar than commonly thought. Sure, people differ in their social preferences and self-confidence, and actually differences can be quite substantially. But gender is not predictive of such differences. Why is this important to know? Why is it important to correct these misperceptions about women and men? One reason is that misperceptions may affect selection and sorting in the labor market, in politics, and in society at large, resulting in a misallocation of talents and traits. For instance, if people think that to be a good leader one needs to have sufficient self-confidence, they may be biased against women, as they mistakenly believe that women tend to be underconfident. Likewise, people may be biased against men when selecting for positions that require generosity and care, mistakenly thinking that these qualities are mainly or only present among women. Read the full papers here (all open access!): Oriana Bandiera, Nidhi Parekh, Barbara Petrongolo, and Michelle Rao (2022), Men are from Mars, and Women Too: A Bayesian Meta-analysis of Overconfidence Experiments, Economica: https://lnkd.in/e6_dMRyG Christine Exley, Oliver Hauser, Molly Moore, and John-Henry Pezzuto (2024), Believed Gender Differences in Social Preferences, Quarterly Journal of Economics, forthcoming: https://lnkd.in/ecZCS2er And see also: Esther-Mirjam Sent and Irene van Staveren (2019), A Feminist Review of Behavioral Economic Research on Gender Differences, Feminist Economics: https://lnkd.in/ejJ8S6iX

We all want to reward employees fairly, yet decades of research--and for many people, their lived experience--show that bias persists. In other words, for the same performance, people earn less or more due to managerial error. New research from researchers at our Stanford VMware Women's Leadership Innovation Lab shows that many interventions are only targeting half the problem. Bias shows up both in how managers describe (view) performance as well as how they reward (value) behaviors. Viewing biases often show up in how performance is described differently based on who is performing it. Men’s approach may be called “too soft,” thus “subtly faulting them for falling short of assertive masculine ideals.” Valuing biases can show up as the same behavior being rewarded when men perform it but not when women do. Examples from the research show that men benefitted when their project specifics were described, whereas women were not. So the same description and behaviors showed up in reviews, but they were only rewarded on men’s. What can be done to curb biases? ✅ Standardize specific guidelines for how managers should view employee behaviors and assign corresponding rewards when giving employees feedback and making decisions about their careers. ✅ Help managers catch bias in both viewing and valuing. ✅ Monitor these impacts from entry level to executive leadership. It turns out that as the criteria shift, so can the way these biases work. A key lesson from our research shows that the work takes discipline, consistency and accountability. These steps may seem like a lot of “extra” work, but at the end of the day, managers also benefit when they weed out biases and fairly promote the most talented employees. Article by Alison Wynn, Emily Carian, Sofia Kennedy and JoAnne Wehner, PhD published in Harvard Business Review. #diversityequityinclusion #performanceevaluation #managerialskills

New research from Aymara evaluated gender bias across 13 popular image generators and found a concerning pattern: most don't just reflect gender stereotypes, they tend to amplify them. When asked to generate images of professionals without specifying gender, the average model showed men 93% of the time for stereotypically male professions and 68% for gender-neutral roles. I'm encouraged that Amazon Nova Canvas showed the lowest bias and achieved representation close to parity. This feels like meaningful validation of our teams’ relentless focus on responsible AI, and I'm grateful to every team member who's contributed to this progress…though we know there’s always more work to do. You can read the full report here: https://lnkd.in/e3-eCXdi

🔍 Gender Differences in Feedback at Work: What We Need to Know 🔍 Feedback is crucial for growth and development, but did you know that the way feedback is given—and received—can differ significantly between genders? Understanding these differences is key to creating a fair and supportive workplace. 🌟 What the Research Says: Feedback Tone 🎯: Studies show that 66% of women receive feedback focused on their personality traits, compared to only 1% of men (Harvard Business Review, 2023). Men, on the other hand, are more likely to receive direct, constructive criticism related to their skills and performance. Personality vs. Performance 👥: Research indicates that women are 2.5 times more likely than men to receive feedback that is vague or less actionable (McKinsey & Company, 2023). This often includes comments on their communication style or likability, rather than their actual job performance. Impact on Career Growth 🚀: The disparity in feedback can significantly affect career progression. Women who receive less actionable feedback are 20% less likely to be promoted into leadership roles (Lean In and McKinsey’s Women in the Workplace Report, 2023). 💬 How to Address It: Be Specific and Actionable ✍️: Whether you’re giving feedback to men or women, focus on specific behaviors and outcomes. Offer clear, actionable steps for improvement to ensure it’s constructive and beneficial. Awareness and Training 🧠: Encourage awareness about these biases in your workplace. Training managers to give balanced, objective feedback can help reduce gender bias and support equal development opportunities. Encourage Open Dialogue 🗣️: Create a culture where feedback is seen as a tool for growth, not criticism. Encourage employees to ask for clarification and examples, ensuring they understand how to improve. 🌍 The Bigger Picture: Addressing gender differences in feedback isn’t just about fairness—it’s about unlocking the full potential of your team. When everyone receives the feedback they need to grow, we all win. Let’s work together to create an environment where feedback is fair, balanced, and truly helps everyone succeed. #WorkplaceEquality #GenderBias #FeedbackMatters #InclusiveLeadership #CareerGrowth

For every role you're overthinking about, a man will confidently go ahead and apply for it. Studies reveal that men apply for jobs when they meet just 60% of the qualifications, while women often wait until they meet 100%. This hesitancy stems from social conditioning that tells women to strive for perfection, while men are encouraged to be bold and take risks. But the effects go deeper: 👽Women are less likely to negotiate their salaries, leading to the gender pay gap. Women earn 84 cents for every dollar a man makes in similar roles, and for women of color, the gap is even wider. 👽In leadership, women hold only 28% of senior management roles globally despite being nearly 50% of the workforce in many industries. 💡 Why does this happen? Social norms: From a young age, women are conditioned to play it safe, be agreeable, and not appear "too assertive." Bias in recruitment: Gender stereotypes can lead to women being undervalued or overlooked during hiring and promotions. The Motherhood Penalty: Women who have children often face a double bind—viewed as less committed to their jobs, they are less likely to be promoted or given leadership opportunities. Mothers are offered 7.3% lower starting salaries compared to non-mothers, while men experience a fatherhood bonus, where they are seen as more committed after having children. Lack of mentorship and role models: Women have fewer opportunities for mentorship, sponsorship, and career guidance compared to men. 💪 How can we improve? Encourage more risk-taking: It’s time to embrace imperfection and go for opportunities. Salary transparency and negotiation training: Advocate for clear salary ranges in job postings and provide negotiation training, especially for women starting their careers. Support working mothers: Companies should adopt flexible work policies, provide paid parental leave, and ensure that motherhood is not penalized in promotions or salary decisions. Mentorship and sponsorship programs: Companies should establish mentorship initiatives, especially for women, to build networks, confidence, and leadership skills. Bias training in hiring and promotions: Recruiters and employers need to be trained on recognizing and addressing unconscious biases to create more equitable hiring and promotion practices. We’re all responsible for changing the system. It’s time for us to close the gap.

Doctors rate women's pain as lower than men's—but talking to patients tells a different story. A study on "Gender biases in estimation of others’ pain" found that on a scale of 0-100, doctors rated women's pain 2.45 points lower than men's, even when the actual pain levels were identical (link to paper in comments). Doctors consistently underestimate women's pain. But what happens when we look at real-world data? When we analyzed actual patient experiences as reported directly by patients, the gap between physician perception and patient reality seems clear: Our data show that men report their symptoms as 6.3% less severe than women, so while doctors perceive women as exaggerating their pain, real-world data suggest that men actually downplay their symptoms. The paper suggests that this is the result of a stereotype that women are more expressive about pain than men. And because of this, male patients' expressions of pain were seen as more "credible," while those of female patients were "discounted" when assessing pain. This isn't just about perception - it's about patient outcomes: this bias shapes diagnostic timelines, it affects who gets treated and how quickly. Pain is subjective. But when medical decisions are based on biased perceptions rather than real data, the entire care pathway is flawed. We need to close the gap between physician perception and patient reality: Real-world evidence (RWE) and new technologies can reveal and eliminate unconscious bias. Data can drive better care—but only if we use it right. — If you're working on RWE, patient engagement or improving the patient experience - let's connect 👉🏼 Mattia

In epidemiologic studies, measurement biases between genders can distort our understanding of health outcomes. Measurement scales, diagnostic criteria, and even data collection methods often reflect historical biases that favor one gender over another (e.g., may not capture gender-specific symptomology). This skewed approach has deep roots. Instead of the default being inclusion, the default was exclusion when it came to women in clinical trials—a choice driven by societal, cultural, and scientific biases. Concerns about reproductive health, like potential risks to fetuses or hormonal shifts from menstruation, were cited to bar women of childbearing age, even when irrelevant to the study. Male physiology was treated as the "standard," with trials overwhelmingly designed for men under the false assumption that gender differences in drug responses or side effects were trivial. Women’s hormonal variability was framed as a problem to avoid, and the absence of women in medical leadership cemented their exclusion for decades. The fix goes beyond solidarity statements on women's day. We need more inclusive approaches in study design: 1️⃣ Stratify by gender—and age—when sampling in clinical studies: Stratifying by gender during recruitment ensures enough women are included. But in some cases, gender alone isn’t enough—older women are often underrepresented, missing issues like perimenopause or menopause. Stratifying by age (e.g., <50 vs. 50+) and gender creates four groups—older men, younger men, older women, younger women—letting us probe treatment effects or disease patterns across diverse groups. 2️⃣ Test for effect modification by gender: Analyzing whether gender alters an intervention’s impact can reveal critical biological insights. If a treatment helps everyone but benefits one gender more, that’s a key finding, for better or for worse. 3️⃣ Seek female co-authors deliberately: Especially for women’s health topics, diverse teams matter. An all-male group risks missing key variables only flagged late (say, in peer review) because no one saw the female perspective. This can introduce unmeasured confounding. Once the work’s done, don’t judge author contributions by nouns or pronouns (Jack, Jill, him, her)—that’s the wrong lens. Focus on verbs and adverbs (analyzed, wrote, thoroughly, expertly): what was done and how well. 4️⃣ Power Studies for Subgroup Analysis: Design trials with enough statistical power to detect gender-specific differences, avoiding the trap of underpowered, one-size-fits-all conclusions. Gender sensitivity isn’t just about methods—it’s also about language. 🗣️ Words shape perception, and outdated terms entrench exclusion. Small shifts matter: ❌ Chairman → ✅ Chair or Chairperson ❌ Mankind → ✅ Humanity ❌ Man-made → ✅ Synthetic or Artificial ❌ Manpower → ✅ Personnel ❌ Layman → ✅ Layperson ❌ Middleman → ✅ Intermediary It’s time our science mirrors reality—for everyone. 🌍 #Chisquares #GenderBias #InclusiveResearch