Are Stereotypes Accurate? Some New Evidence and a Dialogue about My Texas Accent

My daughters and wife routinely make fun of my accent. I pronounce potato with a soft 'a' rather than hard 'o' at the end. The same will routinely go for 'pillow.' My 'i's' throw people for a loop, such as when I pronounce 'night.' And, I routinely end days of the week with 'ee' rather than 'day.' When they poke, I counter that I speak like a Texan, and given that Texas is "God's Country" or "The Land of Milk and Honey" why in the world would I want to change it? Further, "y'all" is a Southern staple and also one of the more inclusive words around. 

Melissa, however, is very much aware of her accent (or lack thereof) and the girls' because of stereotypes associated with accents. Unless there is strong countervailing evidence to suggest otherwise, people with Southern accents are stereotypically not viewed as bright or articulate. Thus, if you can help it, why would you choose to speak in a way that prompts people to develop those preconceptions of you?

I was reminded of this continuing dialogue when I read a recent article from Lee Jussim and colleagues.* The authors noted that most definitions of stereotypes position the beliefs as faulty. This has been the case for nearly 100 years and continues in recent texts. There is reason to suggest, though, that such a stance might not be accurate. 

They suggested that you can examine stereotype accuracy by: (1) assessing people's beliefs about other groups, (2) collecting separate data, such as Census data, that addresses the empirical question, and (3) comparing the two. They offered an overview of over 50 studies that had done just that. Results of their analyses showed:

  • Stereotypes concerning demographics are generally accurate, 
  • Beliefs associated with political orientation are directionally accurate but overstated, particularly among those who hold a strong political affiliation, and 
  • Stereotypes concerning countries as a whole are faulty, especially when compared to personality data. 

Given these findings, the authors offered several recommendations, three of which I found particularly relevant. First, we should define stereotypes in ways that allow for them to be accurate (luckily, I did so in the recent version of my book, which came out before this article). Second, use objective data to empirically scrutinize the legitimacy of stereotypes. Third, people should not declare that some stereotypes are inaccurate because they do not represent everyone in a group. Something can be true for the group as a whole and not for a particular individual in it. 

After reading the article, I was then interested in the relationship between Southern accents and intelligence. If we accept that Southern accents are most likely spoken by people living in the South and that educational attainment is one of many measures of intelligence, we can empirically examine this stereotype. To do so, I drew from state-by-state data from the US Census Bureau and examined high school graduation rates, bachelor's degree attainment, and advanced degree attainment. Across all three measures, the South under-performed relative to other states (83% v 88%; 24% v 29%; and 8% v 11%, respectively). Given the above assumptions, these data show there is a kernel of truth to the stereotypes. But, keeping in mind Jussim et al.'s recommendations, we also know it is not true for all (as Melissa, from the South, is one of the smartest people I know, y'all). 

***

* Jussim, L., Crawford, J. T., & Rubinstein, R. S. (2015). Stereotype (In) Accuracy in Perceptions of Groups and Individuals. Current Directions in Psychological Science, 24(6), 490-497.