Confirmation Bias Against Black Associates Shown in Memo Study - Greedy Associates
Greedy Associates - The FindLaw Legal Lifestyle and Career Blog

Confirmation Bias Against Black Associates Shown in Memo Study

Study after study on implicit bias has shown that, when it comes down to it, we're all racists. Some studies use word association and split-second reactions to show that people are subconsciously racist -- that is, they tend to associate positive words with Caucasians, and negative words with African Americans. Others test the confirmation bias -- the tendency to notice facts that confirm your inherent biases.

The Nextions leadership consulting firm wanted to test this latter bias in a law firm setting, so they sent out a mistake-ridden research memo to law firm partners for review. (H/T to She Negotiates.) The average grade for the Caucasian-labeled candidate was a 4.1 out of 5. The African American? A mere 3.2.

Methodology

Of course, no discussion of a study is complete without an examination of the methodology. The study's authors had a team of law firm partners draft the test memo, then inserted 7 spelling/grammar errors, 6 substantive writing errors, 5 errors in fact, and 4 analytic errors.

Sixty copies of the memo were sent out (with research materials that served as the basis for the memo attached), netting 53 responses. If these, 24 were from the "African American" Thomas Meyer, and 29 were of the "Caucasian." Other than race/ethnicity, Meyer's bio (3rd Year Associate, NYU Law School graduate) was the same for both groups.

The surveyed group included 22 law firms, 23 women, 37 men, 21 were racial/ethnic minorities, and were 39 Caucasians.

Results

As mentioned, the overall score for the Caucasian (4.1) was significantly higher then the African American's score (3.2). Other findings included:

  • Drastically different qualitative comments ("Caucasian" Meyer was described in generally positive terms, which heighted his "potential," while "African American" Meyer inspired the comment: "can't believe he went to NYU.");
  • For spelling and grammar errors, 2.9 out of 7 were found for "Caucasian" Meyer, while 5.8 were found in "African American" Meyer's memo.
  • For writing errors, 4.1 out of 6 were found in "Caucasian" Meyer's memo, while 4.9 were found for "African American" Meyer.
  • For factual errors, 3.2 out of 5 were found for "Caucasian" Meyer, while 3.9 were found in "African American" Meyer's memo.

It gets worse: when it comes to formatting suggestions, which wasn't even a requested form of feedback, partners chimed in anyway, in 41 out of 53 responses. Eleven were in response to "Caucasian" Meyer's memo, while 29 edits were made to "African American" Meyer's memo.

Significantly, there was no significant correlation between the grading partner's race or gender and the pattern of errors found in the memos, except that female partners tended to find more errors and write longer narratives than male partners.

Takeaway

In this study, the only variable was the memo author's race, and the results were clear. One might criticize the sample size, but the drastic difference in overall grades makes a statistical fluke unlikely.

Another interesting note: the paper mentions that in a different study, where blind neutral evaluation was done on real life law firm work, minority and female associates were evaluated more positively than majority men.

The study's results made us wonder: will this hold up over a much larger sample size? Also, would the results be the same across generations? (Partners tend to be older -- does one's age skew results?)

Is everyone really racist?

Related Resources: