The Three Personalities of America

Sigh … more junk science. When you ignore the constituent properties of measurement and meaning, and just play with numbers and ranks, this is what you get. No wonder Chris Ferguson published:
Ferguson, C.J. (2015). “Everybody knows psychology is not a real science”: Public perceptions of psychology and how we can improve our relationship with policymakers, the scientific community, and the general public. American Psychologist, 70, 6, 527-542.

Anyway, the reality of that Rentfrow work follows below. And yes, I did contact the authors (with more detailed information) for comment back in 2008 when this work first appeared. No response; much as you’d expect from psychologists intent on telling a good story rather than more carefully understanding the data they acquired.

Never mind; it’s just that some us took take Richard Feynman’s words and parable of cargo-cult science seriously.

Rentfrow, J., Gosling, S.D., & Potter, J. (2008) A theory of the emergence, persistence, and expression of geographic variation in psychological characteristics. Perspectives on Psychological Science, 3, 5, 339-369.

Data is presented in the paper which seems to show States differ in the magnitudes of personality scores. But the data reported is actually ranks and z-scores – not the actual raw scores.

The key question here is do the actual State mean scores on the BFI personality test reflect the magnitudes implied by the ranks of the means (and the separation indexed by z-score/percentile transformation of the mean scores?)

A previous Wall Street article reads …
”The Geography of Personality
Even after controlling for variables such as race, income and education levels, a state’s dominant personality turns out to be strongly linked to certain outcomes.

Amiable states, like Minnesota, tend to be lower in crime.

Dutiful states — an eclectic bunch that includes New Mexico, North Carolina and Utah — produce a disproportionate share of mathematicians.

States that rank high in openness to new ideas are quite creative, as measured by per-capita patent production. But they’re also high-crime and a bit aloof.

Apparently, Californians don’t much like socializing, the research suggests.

As for high-anxiety states, that group includes not just Type A New York and New Jersey, but also states stressed by poverty, such as West Virginia and Mississippi. As a group, these neurotic states tend to have higher rates of heart disease and lower life expectancy.”

Bottom line for me
The authors are basically asserting that a 1.2 -to 3.3 point average score difference in a 32-40 score point range on a scale accounts for distinguishable personality differences between states. This doesn’t seem to make a lot of sense to me; because I assume that more than a 1 or 2-point raw mean scale score difference (each item is scored 1-5) is required to differentiate “high vs low” State scale scores, and the behaviors these might imply as “State Stereotypes”.

So, what I did was “open-up” this score range back into the metric of the actual questionnaire 9-score range for each item). If we multiply each mean score by the number of items in the scale, we obtain the raw test-score mean.

For Extraversion we have a 1.2 mean disparity score range across all countries in an 8-40 point possible score range.

For Agreeableness we have a 1.9 mean disparity score range in a 9-45 point possible score range.

For Conscientiousness we have a 1.6 mean disparity score range in a 9-45 point possible score range.

For Neuroticism we have a 1.8 mean disparity score range in an 8-40 point possible score range.

• i.e. the difference between the Highest and Lowest anxiety state mean scores is 1.8. This represents 6% of the possible measurement range.
• But the =ranked= difference between lowest and highest States is 50.

For Openness we have a 3.3 mean disparity score range in a 10-50 point possible score range.

This stuff is just embarrassing.