Tuesday, 7 August 2012

Is a taste for extreme answers distorting cross-cultural comparisons of personality?

Data on how personality varies around the world is puzzling. Take the dimension of conscientiousness. Among individuals within a particular country, those with higher conscientiousness tend to earn more money and live longer. This makes sense given the behavioural sequelae of conscientiousness, including diligence and attention-to-detail. Compare across countries, however, and what you find is that richer countries with longer life expectancy tend to have lower average conscientiousness. Now a new study has tested a possible explanation for this paradox - perhaps there's a systematic bias between countries in people's tendency to tick more extreme scores on questionnaires.

How do you tell if a population's higher scores are a reliable reflection of their underlying traits, or if they're caused by a proclivity for more extreme answers? One way is to ask them to rate not just their own personality, but also the personality of a number of fictional characters described in vignettes. Exaggerated scores for the fictional characters would be a sign of a skewed response style.

A small army of researchers around the world led by René Mõttus at the University of Tartu in Estonia has taken on this challenge, recruiting 2,965 people across 20 countries (including European, African, American and Asian nations) and asking them to rate their own personalities and the personalities described in vignettes.

Mõttus and his colleagues uncovered systematic differences between nations in people's proclivity for extreme responding. One pattern to emerge was that richer East Asian countries tended to avoid extreme scores, whereas poorer countries in Africa and SE Asia tended to give more extreme ratings. Adjusting for cross-cultural response styles, the puzzling negative correlation disappeared between average international conscientiousness scores and national longevity and wealth.

The researchers acknowledged that they haven't shown conclusively that extreme response tendencies cause higher conscientiousness ratings. Theoretically the causal direction could run backwards, although common sense suggests this is unlikely. You'd expect higher scorers on conscientiousness to avoid extreme scores, not embrace them. Another possibility is that another unknown factor is at play, inflating conscientiousness scores and encouraging extreme responding. However, it's difficult to imagine what such a factor might be. Taken altogether, the researchers think the most likely explanation is that a proclivity in some countries for extreme responding has had the effect of inflating their conscientiousness scores.

All this raises a further intriguing question ... why should people in some countries be more prone to giving extreme answers? The answer remains beyond the current study, but the researchers suggested one factor could be "dialectical thinking ... 'an emphasis on change, a recognition of contradiction and of the need for multiple perspectives, and a search for the "Middle Way" between opposing propositions'". Countries where dialectical thinking is more common would be expected to avoid extreme scores. Consistent with this, there's some evidence that dialectical thinking is higher in East Asian countries that were found in this study to refrain from giving extreme scores.
  _________________________________ ResearchBlogging.org

Mõttus R, Allik J, Realo A, Rossier J, Zecca G, Ah-Kion J, Amoussou-Yéyé D, Bäckström M, Barkauskiene R, Barry O, Bhowon U, Björklund F, Bochaver A, Bochaver K, de Bruin G, Cabrera HF, Chen SX, Church AT, Cissé DD, Dahourou D, Feng X, Guan Y, Hwang HS, Idris F, Katigbak MS, Kuppens P, Kwiatkowska A, Laurinavicius A, Mastor KA, Matsumoto D, Riemann R, Schug J, Simpson B, Tseung-Wong CN, and Johnson W (2012). The Effect of Response Style on Self-Reported Conscientiousness Across 20 Countries. Personality and social psychology bulletin PMID: 22745332

Post written by Christian Jarrett for the BPS Research Digest.


JoeDuncan said...

Wouldn't it be easier to simply scale or re-weight the responses of any given population based on the variability in the scores produced?

Surely those populations with a preference for "extreme" answers have a much higher variability?

Anonymous said...

In fact, extreme response style may result in *reduced* variability in responses. The thing is that responses to, for example, Likert-type scales are often very skewed, such that most people use, say, only the higher (more desirable) end of the scale. Now, if there is less pressure for extreme responding, all response categories on that higher end will be used, whereas extreme response style will push all responses to a single extreme high-end category.

Morgan said...

I think the using extreme answers will always lead to skewed results. I think the tests should be redone only with different choices to receive a more accurate reading.

Post a Comment

Note: only a member of this blog may post a comment.