Minerva Psychology

Is GenAI Too 'Human’ to be Used in Human Research? 🤖 It looks like there's a social desirability bias in ChatGPT and its friends... Not surprising since they're trained on human data. Consequently if there’s an inference of personality appraisal, responses are bent in a 'desirable' direction. Research suggests: 🔶 If personality evaluation is *inferred* by a LLM there’s skewing towards increased extroversion, lower neuroticism etc. 🔶 Evidence of bias is found if questions are randomised, paraphrased (to control for memorisation), and across temperature ranges. 🔶 Reverse coding questions reduces bias but doesn’t eliminate it - this suggests it’s not entirely a result of ‘acquiescence’ (yea-saying). 🔶 LLMs show a high degree of internal consistency in response - Alpha 0.80 for sub-scales, and 0.93 for all items. 🔶 Effect sizes are large: GPT-4 responses change by 1.20 (human) standard deviations, and Llama-3’s by 0.98 SD. Takeaway: Obi-Wan has trained you well. Not quite! But social desirability in LLMs is an issue because it seems to increase as models become larger and more advanced. Paradoxically, LLMs may be becoming less useful as proxies for human participants - in research, surveys and elsewhere - as they learn to be more like a *prototypical* human. There are also interesting implications for those who use GenAI to help them apply for jobs :) 👉 Salecha et al (2024): https://lnkd.in/e4WghZb3 ♻️ Repost if you think this would interest your colleagues and network #psychologytheinterestingbits #iopsych #personality #genai

Should We Stop Asking Introverts To Be More Extraverted? 🥳 Society seems to over-value extraversion. And many feel social pressure to act more extraverted - but does it actually give you an advantage in life… A recent synthesis of 97 meta-analyses and a clever experiment (monitoring levels of extraversion and vitality, 6 times a day, over 5 consecutive work days) give us some clues. First, the meta-analyses: 🔸 Over a broad range of life and work areas, being more extraverted confers a small, persistent and positive advantage - grand mean ρ̄ = 0.14. 🔸 Focusing on four distinct areas of work ‘advantage’ - motivational, emotional, interpersonal, performance - boosts the figure to ρ̄ = 0.20. 🔸 So there’s something there, but perhaps not as much as might have been expected - given the perceived premium on being an extravert. Now, the experiment*: 🔸 When introverts act more like extraverts they typically feel more energised and mood is boosted (increased vitality) - in the short term. 🔸 Energy levels decline substantially after an hour and this can reduce the benefit gained from acting out of character. 🔸 This suggests the costs may sometimes outweigh the benefits - and can leave you with a so-called ‘extraversion hangover’. Takeaway: If you’re more of an introvert, is it time for greater networking, public speaking and contributions in work meetings? To ditch the reflective alone-time and delay recharging activities... well, contrary to received wisdom, there does seem to be a downside to acting like an extravert. Yes there is a positive effect - not least because social presence leads to being noticed more by others - something that’s valuable in a work context. However, the research suggests things are more nuanced. And for those who are more introverted the best advice is to pick your battles: step out of character when it serves your purpose, not just to satisfy other peoples’ extraversion bias. 👉 Wilmot et al (2019), meta-analyses: https://lnkd.in/eV93avmU 👉 Pickett et al (2020), experiment: https://lnkd.in/e5Jr9NSv ♻️ Repost if you think this would interest your colleagues and network. * Small sample, n=67, observations = 1664. However, the same effects have been demonstrated by other researchers. ** Table, below. Figure-1 shows the concurrent within-person fluctuations in trait extraversion are positively associated with vitality, and this holds across different levels of extraversion. Figure-2 shows that when looking at delayed effects (1 hour later), deviations in trait extraversion are related to decreases in vitality, and this also holds across different levels of extraversion. #psychologytheinterestingbits #iopsych #personality

Does Your Team Conduct ‘After Action Reviews’?💥 It happens in the military*, but after each project do you *systematically* discuss what was expected, what happened, why there was a difference, and what you’ll do differently next time… In the jargon this is called ‘Team Reflexivity’ and it feeds into how a team changes its approach, revises communications, and flexes its decision making over time - basically, how it evolves. The latest meta-analysis (5766 teams; 79 independent effect sizes) suggests the following relationships between reflexivity and team performance. It also explores the influence of team size and tenure; and results relating to antecedents (e.g. leadership, team composition) and emergent states (e.g. trust, cohesion): 🔸 The uncorrected correlation between reflexivity and team performance is 0.30 (0.35 corrected). 🔸 And with team innovation (0.37), creativity (0.40), and satisfaction (0.46). 🔸 Team type and industry do not significantly influence the relationship. 🔸 Team size increases the positive relationship between reflexivity and performance. 🔸 Team tenure also increases the positive relationship between reflexivity and performance. 🔸 Leaders (participatory, transformational, task) are impactful, but…** 🔸 Leadership ‘style’ effects are all fairly similar, e.g. participatory (0.41) v. task (0.43). 🔸 Team composition is influential, e.g. learning oriented (0.35), goal oriented (0.27). 🔸 Structure appears to matter, e.g. autonomy (0.43), goal interdependence (0.39). 🔸 And emergent state effects are telling: psychological safety (0.46), cohesion (0.44), efficacy (0.38), and trust (0.44). Takeaway: Team Reflexivity - that’s regularly and actively reflecting on team performance *and* initiating change and improvement - impacts work performance. The relationship is more pronounced in teams that are large and established. Team attributes like psychological safety, cohesion and trust are also positively related to reflexivity. It appears that even if time is at a premium, team review activities are a simple and powerful way of influencing performance. The question is how many businesses or organisations, of whatever size, (really) encourage and support this type of team behaviour… 👉 Pierre-Marc Leblanc et al (2024): https://lnkd.in/eZ2926-v ♻️ Repost if you think this would interest your colleagues and network. * And in some other organisations, for example in the health sector; plus specific contexts, like safety. ** Leadership data is limited. However the fact that all the results are similar might imply that it’s really just a question of having an engaged and empowering leader who can engender trust. #psychologytheinterestingbits #iopsych #teamwork

Are there (Really) Different Leadership ‘Styles’? 💼 Authentic, Charismatic, Servant, Transactional, Transformational etc. There’s lots of them but are they bringing anything useful to the party… A new study exploring 12 dominant leadership styles (7 samples, 5 countries, multiple organisations, 4K participants) suggests a significant overlap between different ‘styles’. In brief: 🔸 The 12 different leadership assessments studied shared significant amounts of variance - mostly attributable to a general leadership factor. 🔸 Assessments do not systematically capture unique leadership-related variance - take out the general factor and what’s left may not be useful. 🔸 Shared variance seems to mostly equate to the affective quality of the leader-follower exchange relationship - this could simply be ‘likeability’! Takeaway: Leadership styles have much more in common than differences. Looks like we’re back with behaviour and the proliferation of ‘styles’ is actually a distraction. They’re just not distinct enough to be useful. Sorry, management consultants. The authors suggest a taxonomic approach that groups meaningful behaviours. This is broad stuff like having a task or strategic orientation, or if we think about the affective aspects of leadership, the relational stuff… 👉 Eva et al (2024): https://lnkd.in/eQuQUfzn ♻️ Repost if you think this would interest your colleagues and network. * Thanks to Timo Lorenz for surfacing this article. ** Lots of stats in the article. Some of the results for the US sample, below. #psychologytheinterestingbits #iopsych #leadership

Can You Meaningfully Change Your Personality? ⭕️ Is this possible and what could it mean in a work context… There’s debate about setting goals designed to shape traits - for example, a candidate consciously deciding to become more organised - and how this could impact selection processes. Practically, should these Personality Development Goals (PDGs) influence hiring decisions? For instance, when presented with two “identical” candidates - except that one has a PDG aligned with an important aspect of the role - who would be the better choice? In this new article the authors explore key aspects of applying PDGs to HR practices. They conclude: 🔸 Personality development through PDGs is possible but has been little tested in organisational settings. 🔸 Shifts in personality have been observed over a number of months, rather than years. 🔸 Personality development effects are modest, at best. The biggest shifts appear to be for Emotional Stability. 🔸 Better outcomes are achieved if PDGs are paired with personality development interventions - i.e. ultimately they need to involve employee and employer. 🔸 Multi-rater approaches provide better visibility of progress towards goals, and the opportunity for more effective feedback. Takeaway: Does personality development represent untapped potential? And from an employer’s perspective, does this equate to meaningful change? Achieving a different score on a personality questionnaire is one thing, but how does this translate into work behaviour… It should also be recognised that research is often conducted on student samples. These are typically younger people, and what we do know is most personality development takes place up to about 30 years of age. So, are PDGs just nudging change that could happen anyway, or are they influencing state-to-trait change more profoundly - both in younger and older people? 👉 Perossa et al (2024): https://lnkd.in/e3H3RGCq ♻️ Repost if you think this would interest your colleagues and network. Table, below. Cohen’s d is a measure of effect size: 0.2 is considered small, 0.5 medium, and 0.8 large. The table also appears to contain a few rather wild figures! #psychologytheinterestingbits #iopsych #personality

What Works Best at Reducing Discrimination at Work? ⚖️ Latest research suggests ‘passive’ measures - like diversity training - are unlikely to be effective. Whereas those targeted directly at behaviour show more promise… A recent meta-analysis (70 studies, 208 effect sizes) explored 22 different interventions designed to reduce work-place discrimination. These were also split between those that were Cognitive and designed to disrupt stereotyping (e.g. Resume ‘blinding’), Affective and concerned with influencing feelings (e.g. Evoking empathy), Behavioural and targeted at stopping the formation of biases (e.g. Accountability), or general Education about bias processes(e.g. Diversity training). The key points: 🔸Measures like short-term education, reminders of bias processes, calculating mindset (analytical approaches) and egalitarian messaging do not tend to influence behaviour. 🔸Attempting to control the manifestation of biases, by targeting specific behaviours, for example by challenging/changing social norms, making people accountable and affirmative action are more successful. 🔸Generally, cognitive interventions appear most effective at disrupting stereotype processing (0.51)*, and least effective at inhibiting the manifestation of biases (-0.70). 🔸Affective interventions also work best at disrupting stereotype processing (0.62); whereas behavioural approaches are helpful at disrupting stereotypes (0.22), inhibiting bias manifestation (0.59) and impacting affective states (0.23). There’s much more detail, and a range of analyses and results tables, in the article. Takeaway: There are many ways of attempting to reduce discriminatory behaviour at work. Research suggests general-purpose diversity training has limited impact, as do bland non-discrimination policies and statements. Rather it looks like interrupting the expression of biases is more effective, through for example people use transparent processes, and having them explain and justify their actions. Unsurprisingly aligning the type of attitude targeted with a congruent intervention, and an obvious outcome, also looks important - i.e. interventions need to be well focused. 👉 Costa (2024): https://lnkd.in/dHEecEes ♻️ Repost if you think this would interest your colleagues and network. * All figures are Hedges' g. This is a measure of effect size: 0.2 represents a small effect, 0.5 a medium, and 0.8 a large effect. #psychologytheinterestingbits #iopsych #dei

How Does ‘Emotional Stability’ Impact Voluntary Turnover at Work? 🔄 In many businesses staff turnover is a big problem. Can Big Five personality insights help… A recent 5 year multi-wave study (N=1.6K) looked at how emotional stability affects the relationship between job performance and voluntary turnover. Why emotional stability? Research shows it’s related to job performance and turnover across different roles, organisations and cultures. And there’s also evidence that helps to explain how the level of emotional stability affects employee’s reactions to work performance ratings. The study found: ▪️Overall relationship between job performance and voluntary turnover is less negative for those with higher emotional stability (more likely to stay). ▪️At the high end of job performance it’s flat for those higher on emotional stability; and more positive for those who are lower (more likely to leave). ▪️Those who are higher on emotional stability - *regardless of their performance level* - are likely to stay longer in an organisation. Takeaway: In general, employees who show higher emotional stability are likely to stay longer in an organisation and are less likely to make turnover decisions based on their performance level. As lower ratings might be expected at the beginning of someone’s employment, before they fully acquire new knowledge and skills, this has implications for roles that are traditionally high (early) voluntary turnover. 👉 Robbins et al (2024): https://lnkd.in/epZkrErN ♻️ Repost if you think it would interest your colleagues and network. * The lagged correlations between job performance and turnover across the years are negative (-0.147 to -0.315, mean -0.244). Modelling the effect of job performance on turnover likelihood, moderated by emotional stability, demonstrates the main and quadratic effects are significant. This suggests the performance-turnover relationship is *curvilinear*; adding emotional stability shows it’s significantly and negatively related to turnover. #psychologytheinterestingbits #iopsych #personality

What is the Greatest 2x2 Matrix of All Time? 🐐 Ansoff Grid, Boston Box, Change Quadrants, Eisenhower Matrix, Johari Window, SWOT Analysis, etc… The 2x2 matrix is based on two related variables and provides a visual representation of four actions or options. These can be used to prioritize and make decisions. A more serious question: these matrices look meaningful but are they useful in the real world - do people *really* use them to make decisions? #psychologytheinterestingbits #psychology #decisionmaking

Does Your Personality Change as you Get Older? ⏳ Most change occurs before age 30. After that, not so much. Important new review of personality change and what it might mean for development activities in organizations... Fresh off the press! Some of the highlights: ⦿ Most personality change happens before 30, and then it slows right down. For older employees personality remains highly stable*. ⦿ Life events and deliberate interventions have a moderate effect. The Big Five dimension most amenable to change is Emotional Stability (two-thirds SD). ⦿ Aligning personality and work design is the way to increase both employee potential and performance (trait activation theory). Takeaway: Personality is pretty stable. However, if you're an employer you should consider assessing younger employees more frequently and try to ensure they are in work environments that capitalize on their personality. This shifts the debate from personality change to the nature of work. This review of personality and age is a must-read for HR/TA Professionals, I-O Psychologists and Careers Advisors. 👉 Ones et al (2024): https://lnkd.in/e7npTbv6 ♻️ Repost if you think it would interest your colleagues and network. * Rank‐order stability tends to increase with age, peaking in middle adulthood. Mean‐level changes are most prominent during young adulthood, particularly in Conscientiousness, Emotional Stability and Agreeableness. #psychologytheinterestingbits #iopsych #personality

Does Culture Eat Personality for Breakfast? 🌍 People from the same culture often behave in similar ways, especially when they're in a group... This suggests there's a relationship between aspects of culture and the Big Five personality dimensions; and countries grouped together in terms of culture will show similar personality traits. A recent large scale study (N=130K) looking at Hofstede's cultural dimensions*, and the Big Five, across 22 countries, found the following**: ⦿ Masculinity correlates with Conscientiousness (0.43) and Extraversion (-0.39). ⦿ Power Distance correlates with Conscientiousness (0.46) and Agreeableness (-0.48). ⦿ Individualism correlates with Extraversion (0.31) and Agreeableness (0.56). ⦿ Uncertainty Avoidance correlates with Extraversion (0.53) and Neuroticism (0.30). ⦿ Long-Term Orientation correlates with Openness (-0.47). ⦿ Indulgence correlates with Neuroticism (-0.34). Two large clusters of countries also emerged***: ⦿ Cluster 1 had higher scores in Conscientiousness, Neuroticism, Power Distance, Masculinity, Uncertainty Avoidance and Long-Term orientation, and lower scores in Openness, Agreeableness, Individualism and Indulgence.    ⦿ Cluster 2 had higher scores in Openness, Agreeableness, Individualism and Indulgence, and lower scores in Conscientiousness, Neuroticism, Power Distance, Masculinity, Uncertainty Avoidance and Long-term orientation. Takeaway: There appear to be substantial relationships between many aspects of individual culture and personality - but not all. And countries can be clustered on the basis of personality and cultural dimensions, with culture having predictive power over certain aspects of personality. But this is psychology and not all agree: does culture really 'eat' personality, or is it the other way round? And more to the point, if you're in one of the clusters do you recognize the characterization of your country? 🤔 👉 Rocha et al (2024): https://lnkd.in/e6WwZ43D ♻️ Repost if you think it would interest your colleagues and network. * See The Culture Factor for more information: https://lnkd.in/eNASDNjg ** The correlations are Spearman rank correlations. The Adjusted R-squares are Masculinity (0.39), Power Distance (0.39), Individualism (0.39), Uncertainty Avoidance (0.49), Long-term Orientation (0.16), Indulgence (0.12). These indicate the variables that show the best improvement in model fit in a regression analysis. Collectively personality is a significant predictor of all the cultural dimensions apart from Long-term Orientation and Indulgence. *** Cluster 1: Albania, India, Germany, France, Hong Kong, China, Romania and Mexico. Cluster 2: Canada, New Zealand, South Africa, Australia, UK, Ireland, Norway, Philippines, Malaysia, USA, Netherlands, Sweden, Singapore and Finland. #psychologytheinterestingbits #iopsych #personality #culture