Back to Post-publication discussions

1
Honesty, Intelligence and Race

Submission status
Accepted

Submission Editor
Noah Carl

Authors
Sebastian Jensen
Emil O. W. Kirkegaard

Title
Honesty, Intelligence, and Race

Abstract

 

Research shows that honesty correlates positively with intelligence. Similarly, there are racial differences in honesty, with Europeans being more honest than various other ethnic groups. It is currently unknown to what degree race differences in intelligence can explain the differences in honesty. We investigated this question using data from the National Longitudinal Study of Youth 1997 (NLSY97), a large American longitudinal dataset. We replicate prior findings that honesty correlated with measures of intelligence (r = .38, 95% CI [.34, .41]) and that Blacks (d = -0.67, 95% CI [-.76, -.59]) and Hispanics (d = -0.4, 95% CI [-.50, -.31]) are less honest than Whites, and this holds whether honesty is measured by self-reports, interviewer-reports or by parent-reports. In addition, race differences in honesty remained between Blacks and Whites but not between Whites and Hispanics after controlling for intelligence. 

Differences between Blacks and Whites but not Whites and Hispanics were noticeably lower in self-reports (Blacks: d = -0.18 [-0.24, -0.11], Hispanics: d = -0.24 [-0.31, -0.17]) than parent-reports (Blacks: d = -0.43 [-0.52, -0.35], Hispanics: -0.24 [-0.33, -0.15]) and interviewer-reports (Blacks: d = -0.7 [-0.75, -0.64], Hispanics: -0.3 [-0.36, -0.25]). 


Cross-national comparisons were made using national IQ data and Hofstede’s cultural dimensions. Bayesian model averaging suggests that Hofstede’s individualism dimension (β = .64, PIP = 100%), national IQs (β = .25, PIP = 73.6%), and masculinity (β = -.35, PIP = 100%) predict differences in honesty between countries. Parking violations per diplomat were only predicted by national IQs (r = -0.28, p < .001), given that no other variable reached a posterior inclusion probability above 0% besides national IQs. Implications and theories concerning these findings are discussed.

Keywords
intelligence, IQ, race, Ethnicity, black, honesty, white, hispanic

Supplemental materials link
https://osf.io/vw4sa/

Pdf

Paper

Typeset Pdf

Typeset Paper

Reviewers ( 0 / 0 / 2 )
Reviewer 1: Accept
Reviewer 2: Accept
Public Note
Changes: -confidence intervals fixed -irrelevant sections removed -removed results from earlier regression analysis

Fri 09 Dec 2022 06:43

Reviewer

This is an interesting piece of research. I never thought about the link between honesty/dishonesty and intelligence, though it seems intuitively obvious once someone points it out. My experience in law enforcement would lead me to expect the black/white difference, but not the white/Hispanic difference.

The piece is well-argued with appropriate statistics, and the values are quite robust. It is also based on a good and well-used sample. I can find no argument with the conceptualization of the problem nor with its analysis.

However, the piece would benefit from running it through Grammarly, or something similar. I found stuff like "an SEM." The regression models don't always line up properly, but I guess that's a computer translation problem. The only other issue is that I got the impression sometimes that parts of the conclusion should have been in the literature review. Overall, I find this a good piece of research and recommend publication.

Author
 

This is an interesting piece of research. I never thought about the link between honesty/dishonesty and intelligence, though it seems intuitively obvious once someone points it out. My experience in law enforcement would lead me to expect the black/white difference, but not the white/Hispanic difference.

The piece is well-argued with appropriate statistics, and the values are quite robust. It is also based on a good and well-used sample. I can find no argument with the conceptualization of the problem nor with its analysis.

However, the piece would benefit from running it through Grammarly, or something similar. I found stuff like "an SEM." The regression models don't always line up properly, but I guess that's a computer translation problem. The only other issue is that I got the impression sometimes that parts of the conclusion should have been in the literature review. Overall, I find this a good piece of research and recommend publication.

The grammatical mistakes have been fixed in the new edition of the paper, as well as the table alignments. The section discussing why honesty and intelligence are correlated did belong in the introduction, and has been moved appropriately. 

Several other changes were made to the paper:

1. Adjusted-R2 values were reported instead of R2

2. Model fitting statistics from the SEMs have been displayed.

3. General intelligence was measured with 4 subfactors (crystalized ability, mathematical ability, nonverbal reasoning, and knowledge) in the SEMs, which improved the model fits but did not change the correlations.

4. 95% confidence intervals were reported for the race differences in honesty. 

Reviewer

Review of “Honesty, Intelligence and Race”

Pretty straightforward manuscript, but I do have some suggestions that may help improve the paper, and / or resolve mostly minor issues.

1. You have a rather innocuous sentence on page 2 (“This psychological profile involves emphasizing the importance of individuals over kin...”), but it made me think of how cultural differences across the various nations might lead to predictions about honesty differences.

The reference here would be Hofstede’s Cultural Dimensions Theory: https://www.mindtools.com/a1ecvyx/hofstedes-cultural-dimensions. Hofstede did his research in the 1970s by studying people who worked at IBM across something like 50 different countries. He identified six dimensions along which national culture varies. One seems highly relevant to the present study:

Institutional Collectivism. Institutional collectivism is the degree to which organizational and societal institutions encourage individuals to be integrated into groups and organizations. In high institutional collectivism countries, collective distribution of resources and collective action are encouraged. Group loyalty is encouraged, even if it undermines the pursuit of individual goals. Sweden, Japan, and Singapore are examples of countries that have high institutional collectivism; Germany, Argentina, and Italy have low institutional collectivism. In the United States, low institutional collectivism has resulted in debates on appropriate work-life balance.

Could you code countries that are high or low on this dimension and then run some simple analyses (I’m assuming you have honesty data across countries, given your discussion in this section of the paper). Perhaps test for both honesty and IQ differences across this dimension.

Assuming the results are interesting, please mention all this in the text, including any new analyses. With success here, you may also want to look at some of Hofstede’s other five dimensions, as you see fit. It may be a stretch, but I was thinking it might also be worth looking at Long- Versus Short-Term Orientation, as you do mention something similar in your discussion section (as a possible explanation for your original results).

If your first analyses with Collectivism fails, don’t worry about looking at any of the other five dimensions. If results are not interesting on this dimension, please just briefly report the null effects in the reply thread here.

2. Could you appease my OCD and use an Oxford comma in the title?

3. Re: The Israeli solder study. That no effect was found past the above-average IQ level made me think of something like a SLODR effect for honesty, but I haven’t thought this through. Mention this only if it makes sense to you / adds value to your intro section.

4. Do you have a cite for your operational definition of honesty?

5. Were there any differences across the three methods you used to classify individuals by race?

6. The correlation between self- and parent- reported honesty kind of sucks. Can you report reliabilities for these (or if the SEMs capture this, that’s fine).

7. Could there be race differences in social desirability that might confound the survey results for honesty?

This is really all I got. It’s a concise, sound manuscript, which I suspect can be published after minor revision.

Bryan

Author

Review of “Honesty, Intelligence and Race”

Pretty straightforward manuscript, but I do have some suggestions that may help improve the paper, and / or resolve mostly minor issues.

1. You have a rather innocuous sentence on page 2 (“This psychological profile involves emphasizing the importance of individuals over kin...”), but it made me think of how cultural differences across the various nations might lead to predictions about honesty differences.

The reference here would be Hofstede’s Cultural Dimensions Theory: https://www.mindtools.com/a1ecvyx/hofstedes-cultural-dimensions. Hofstede did his research in the 1970s by studying people who worked at IBM across something like 50 different countries. He identified six dimensions along which national culture varies. One seems highly relevant to the present study:

Institutional Collectivism. Institutional collectivism is the degree to which organizational and societal institutions encourage individuals to be integrated into groups and organizations. In high institutional collectivism countries, collective distribution of resources and collective action are encouraged. Group loyalty is encouraged, even if it undermines the pursuit of individual goals. Sweden, Japan, and Singapore are examples of countries that have high institutional collectivism; Germany, Argentina, and Italy have low institutional collectivism. In the United States, low institutional collectivism has resulted in debates on appropriate work-life balance.

Could you code countries that are high or low on this dimension and then run some simple analyses (I’m assuming you have honesty data across countries, given your discussion in this section of the paper). Perhaps test for both honesty and IQ differences across this dimension.

Assuming the results are interesting, please mention all this in the text, including any new analyses. With success here, you may also want to look at some of Hofstede’s other five dimensions, as you see fit. It may be a stretch, but I was thinking it might also be worth looking at Long- Versus Short-Term Orientation, as you do mention something similar in your discussion section (as a possible explanation for your original results).

If your first analyses with Collectivism fails, don’t worry about looking at any of the other five dimensions. If results are not interesting on this dimension, please just briefly report the null effects in the reply thread here.

2. Could you appease my OCD and use an Oxford comma in the title?

3. Re: The Israeli solder study. That no effect was found past the above-average IQ level made me think of something like a SLODR effect for honesty, but I haven’t thought this through. Mention this only if it makes sense to you / adds value to your intro section.

4. Do you have a cite for your operational definition of honesty?

5. Were there any differences across the three methods you used to classify individuals by race?

6. The correlation between self- and parent- reported honesty kind of sucks. Can you report reliabilities for these (or if the SEMs capture this, that’s fine).

7. Could there be race differences in social desirability that might confound the survey results for honesty?

This is really all I got. It’s a concise, sound manuscript, which I suspect can be published after minor revision.

Bryan

1. The Hofstede cultural dimentions did relate to honesty and parking violations indepependent of IQ. These analysis will be available in the new edition.

2. Yes.

3. I mentioned it before I did the analysis because I thought it would be relevant for how IQ relates to honesty. This did not replicate , as the relationship appears to be linear. I think it is worth mentioning as it is an example of a finding being inconsistent in two decently sized datasets (n=2487 for ours, n=427 for theirs), though I suspect the true relationship is closer to a linear one as that is the one that occurred in the larger dataset.

4. No - though I edited that section and noted what the dictionary definitions are. Dictionaries seem to agree honesty relates to a consistency between held and expressed beliefs as well as a generally fair/non-criminal conduct. I don't think honesty encompases criminality, as a criminal who admits to his crimes is honest. 

5. I did not check, though I did check to see how consistently self-reports coincided with household data. Among those explicitly identified as one race, accuracy was about 97% for Hispanics and Whites, and approximately 99% for Blacks. 

6. Yes, it does. Fortunately, the SEMs both factor in the measurement error and the differing validity, so parent-reported honesty is weighed more heavily than self-reported honesty. The larger SEM models do use latent variables, though the lower-order variables weren't displayed as the graph got too convoluted.

7. It is possible, but the evidence suggesting Blacks are more likely to lie about drug use suggests that they are more likely to lie to preserve social desirability as well. 

I appreciate the idea of using Hofstede's cultural dimensions, as they were empirical evidence that supported the relationship between honesty and individualistic attitudes. 

Bot

Authors have updated the submission to version #2

Author
Replying to Forum Bot

Authors have updated the submission to version #2

A few other changes were made, besides those discussed earlier:

1. Changing the SEM estimation method from maximum likelihood to diagonally weighted least squares improved the fit of the models, as the honesty variable is not normally distributed.

2. The racial variables were removed from the SEMs as they needlessly complicated the analysis.

3. Subtests instead of subfactors were used to measure latent general intelligence as the latter also needlessly complicated the analysis. 

Bot

Authors have updated the submission to version #3

Author
Replying to Forum Bot

Authors have updated the submission to version #3

Apparently the paper could not be downloaded, so a new version was submitted. It has also been clarified that respondents and parents were notified to only report lying or cheating that ocurred within the last 6 months.

Reviewer

Accept, though you really should consider mentioning hofstede before the method section...

Bot

The submission was accepted for publication.

Bot

Authors have updated the submission to version #5

Bot

Authors have updated the submission to version #6

Bot

Authors have updated the submission to version #7

Bot

Authors have updated the submission to version #8