How Do Psychometric Test Results Vary Across Age, Race and Gender?Nayomi ChibanaBlockedUnblockFollowFollowingJun 27Most of us have taken a personality quiz at one point or another.
Whether it was for a job, school or just for fun, you probably remember skimming over the results to find out more about yourself and how you compare to the rest of the population.
But have you ever wondered how these results are correlated to other demographic factors such as race, gender, age, religion and sexual orientation?I asked that myself recently and took a deep dive into the data to find out the answer to this question.
After exploring all of the publicly available data sets published by the Open Source Psychometrics Project, I chose three in particular (Experiences in Close Relationships Scale; Depression, Anxiety and Stress Scale; and the Rosenberg Self-Esteem Scale) with the results of more than 110,000 psychometric tests released with the consent of test takers.
Based on the answers recorded in these data sets, I found that:1.
There is a relationship between age and personality trait dimensions, as well as relationship attachment styles.
There is a relationship between religion and stress scores.
There is a relationship between race and measures of mental health, such as stress, depression and anxiety.
There is a relationship between gender and measures of emotional stability and agreeableness.
There is a relationship between sexual orientation and depression, anxiety and stress scales.
Although the data analysis yielded some interesting results, I must also stress that due to the nature of online personality tests and the inability to verify the accuracy of responses given, this analysis is meant to provide some food for thought and even motivation for further research on the topic rather than results that can be generalized to the larger population.
Having said that, here is a quality comparison of the data collected through the Open Source Psychometrics Project and data collected on Amazon Mechanical Turk.
The results indicate that the former “contains less than 25% the rate of invalid responding as AMT data.
”And for those interested in the process behind this analysis, the data and code can be accessed here.
How is religion related to stress scores?One of the tests available through the Open Source Psychometrics Project is called the DASS inventory, which stands for Depression Anxiety Stress Scales.
After filtering out all invalid responses (those who responded incorrectly to validity questions), I plotted the results for the remaining 34,583 test results and conducted a two-sample t-test for equal means.
I found that test takers who categorized themselves as Mormons registered the highest mean stress scores (higher scores translate to higher levels of stress), whereas those who labelled themselves Protestants had the lowest stress scores.
Due to the positive t-statistic ( t=4.
734, p < 0.
01), I can conclude that the mean of self-reported stress scores of Mormons is significantly different than that of Protestants.
Using the same test, I also found that there is also a statistically significant difference between the stress scores of Protestants and Atheists ( t=6.
113, p < 0.
You can compare these means to the conventional severity labels below.
Overall, the mean stress score for this population of test takers fits within the cut-off score for those with “moderate” stress levels.
Although the results are thought-provoking, we should be cautious about drawing any conclusions from this data and applying them to the general population as it merits a more thorough analysis, controlling for possible extraneous, confounding variables.
As we age, do we become more secure?Another interesting pattern uncovered in the data was the relationship between certain personality measures and age.
After plotting several variables across age, I found that there is a relationship between the age and gender of respondents and scores of emotional stability, openness, conscientiousness, agreeableness and extroversion, as measured by the Ten Item Personality Inventory (TIPI).
The mean scores of male respondents for emotional stability was higher than that of females across all ages.
Meanwhile, it appears that the older respondents are, the more likely they are to register higher emotional stability scores (higher scores indicate higher levels of emotional stability).
The results of a chi-squared test confirm that these two variables are not independent (χ2=982.
64, p < 0.
Source: TIPI NormsIf you compare these mean values to the norms published by the creators of this scale, you find that the mean emotional stability scores for this population of test takers (3.
185) is lower than that of the general population (4.
This may suggest that those who take online psychometric tests are not representative of the larger population, but this needs to be further investigated.
How is race related to personality measures?I also plotted mean results by race for the different personality traits, as measured by the TIPI and DASS scales.
Although the results might seem counter-intuitive in some cases, it is worth investigating further.
The results of a chi-squared test indicate that race and emotional stability are not independent (χ2=597.
83, p < 0.
Higher scores correspond to higher levels of the trait or behavior in question.
It must also be noted that the majority of test takers self-categorized themselves as Asian, White or Other.
What about sexual orientation?I also performed a chi-squared test to determine whether or not sexual orientation is independent of DASS measures such as emotional stability.
The results indicate that the two are not unrelated (χ2=509.
23, p < 0.
From the subplot below, you can see that heterosexuals registered the lowest anxiety and depression scores compared to the rest of categories (low scores correspond to low levels of anxiety and depression).
Meanwhile, those who labelled themselves bisexuals registered the highest depression and anxiety scores and the lowest emotional stability scores.
Country of origin and levels of stressI also plotted stress scores by country and found that stress scores were not on the whole higher for Western countries.
On the contrary, stress scores seem to be higher for test takers from Middle Eastern and African countries.
ConclusionsFrom this analysis, we can conclude that characteristics inherent to test takers’ identity such as age, gender, race, country of origin and sexual orientation are related to a variety of personality and behavioral measures.
However, we must look further into the difference between the population of online psychometric test takers and the larger general population.
According to the results of this analysis, online test takers seem to register lower scores for emotional stability and higher scores for anxiety and depression.