This morning, I happened upon @mervatim's article describing "5 myths of personality tests." Unfortunately, it relies on old literature and selective misreadings of the newer findings on personality. In this thread, I debunk the mythbusting. 1/ washingtonpost.com/outlook/five-m…
1) Personality traits are about 50% heritable: emilkirkegaard.dk/en/wp-content/… In my course, I answer "nature or nurture?" with "yes"/ "How much?" "Half." However, this doesn't mean that personality is completely stable across the lifespan. 2/
Among many others, @BrentWRoberts has shown how personality changes across the lifespan. This study is a descriptive classic describing how (aspects of) the Big Five change over time. psycnet.apa.org/fulltext/2006-… He's even posited a theoretical framework. 3/ journals.sagepub.com/doi/abs/10.117…
This general "maturation of personality" was anticipated in the writings of Gordon Allport. study.com/academy/lesson… Personality psychologists also embrace a plurality of methodologies and informants (see @DanMcAdams and @siminevazire's work, respectively). 4/
2) The author cherry-picks some of the most egregious historical examples of non-empirical personality test development. In addition to ignoring Allport's work that happened contemporaneously with the article's cited sources, 5/
the article doesn't mention Hans Eysenck's Extraversion-Neuroticism-Psychoticism model or Raymond Cattell's 16-PF (16 traits, reducing Allport's adjectives substantially!) from the mid-20th century. In the 1980's, Costa and McCrae's NEO-PI variants assessed the 5 factor model 6/
Auke Tellegen substantially revised and improved Eysenck's model, and Robert Cloninger made a first (though ultimately unsuccessful) attempt to separate heritable temperament and environmental character. In the 1990s until now, psychological science has refined these, 7/
and the work of @hardsci, @ColinDeYoung, @cjsotomatic and others have used modern psychological science to make more economical, variegated assessments of personality traits. Check our Lew Goldberg's IPIP for a huge range of self-report measures! ipip.ori.org 8/
3) Emre's most trenchant critique of personality measurement surrounds the nature of personality items. Historically, our field hasn't been culturally sensitive in writing items. However, we've also learned that there's meaningful cultural variation in traits. 9/
For example, Openness to Experience is hard to replicate cross-culturally, for some of the reasons Emre addresses in her second myth. journals.sagepub.com/doi/abs/10.117… A narrower "Intellect" factor (how much people like to think, play with mental puzzles) emerges more often. 10/
4) However, personality tests (especially modern ones, not the Meyers-Briggs Temperament Inventory the author's book focuses on) are more reliable and almost invariably use dimensional assessments rather than auguring person "types". Dimensions are more reliable than types. 11/
Emre links to a study that shows by random chance alone, people may have at least one elevated score on the MMPI-2-RF. It's a good point related to the multiple comparison problem, sbenning.faculty.unlv.edu/2018/03/01/how… & she's made me think about how to address it in my assessment course. 12/
5) Nevertheless, personality traits are helpful in predicting job performance jwalkonline.org/docs/Grad%20Cl… , intimate relationship satisfaction sciencedirect.com/science/articl… , and other important domains of life. They're hardly perfect, and observer measures like Jack Block's Q-sorts 13/
and narrative features of personality (see the McAdams work noted earlier, especially with respect to generativity in mid-life, agency and communion, and other work) predict important outcomes above and beyond self-reports. However, traits are easier to measure quickly. 14/
Thus, personality is a useful, imperfect framework for understanding consistencies in human behavior. We lose specificity when we assess broad traits (see Seymour Epstein's rebuttal of Walter Mischel's critique of personality), but we can make broad inferences 15/
about patterns in people's functioning. Not only can it predict good things; it also has robust associations with psychopathology and other untoward outcomes (e.g., my student Vincent Rozalski's dissertation) along with neurobiological processes. onlinelibrary.wiley.com/doi/full/10.11… 16/
In short, I consider 4/5 myths busted, though Emre does point the way toward challenges in unifying etic (what's common across all cultures) and emic (what's built from the ground up within each culture) personality research. /17x
