In his book The Chosen, 1 Jerome Karabel describes the actions of Harvard University President A. Lawrence Lowell with respect to the performance and behaviour of Jewish students at that university. In 1922, Lowell established a subcommittee to develop techniques for identifying who was Jewish (measurement methodology) and then to examine differences between the performance of Jews and non-Jews. At first, the committee reported that the increasing enrolment of Jews did not pose an academic threat and, in fact, that Jews outperformed non-Jews in an academic sense. The committee went on to study “moral offenses” and reported that Jewish students were more likely to be “underdisciplined” (4.7% v. 3% no p value reported) or have committed “offenses involving dishonesty” (3.7% v. 2.7%), but were less likely to be “guilty of drunkenness” (0.1% v. 0.5%). Looking back at these actions and the reported research nearly 90 years later makes them seem very strange indeed.
The purpose of academic inquiry is to ask and answer questions. The questions are primarily derived from the individual investigators’ curiosity. The principle of academic freedom attempts to establish few boundaries on what can and cannot be studied. However, research, like all other human endeavours, is subject to many psychological and sociological influences. It is not performed in a vacuum. Some topics of research are off limits because they will offend the investigator’s supervisors or bosses. In other cases, the backlash comes from people whose position in the hierarchy is below the offender (e.g., President Lawrence Summers in his comments about women scientists). 2
What are the elements of research that make it socially sensitive, unacceptable or offensive? In our view there are two. The first is the actual or perceived purpose of the research. Karabel reports that the actual purpose of Lowell’s research was to justify reducing enrolment of Jewish students at Harvard. Lowell is neither the first nor likely the last person to undertake research in the hopes of supporting the advocacy of a certain action. When that action has the potential to harm people, simply asking the question is perceived as threatening. Suppressing the result when the answer fails to support a desired action is another example of research misconduct.
The second characteristic of socially sensitive research concerns the prior beliefs of groups within society. When the CMAJ reported that breast self-examination was of no value in screening for breast cancer, 3 investigators who had invested considerable efforts in promoting this strategy were upset. They perceived the research as being methodologically flawed and, more importantly, biased against women. They invoked the analogy that men would never be told to stop screening themselves for testicular cancer by self-examination. 4 It is a natural human phenomenon to be skeptical of results that conflict with either prior beliefs or have the potential to have negative distributional consequences (e.g., the productivity of female physicians). 5
How should academics respond to these issues? We believe there are four approaches that might protect investigators when they consider undertaking potentially controversial research. First, they should fully examine their own motives and the motives that others might perceive in asking the question. Second, they should ensure that their methodology does its best to protect against biased interpretation of the answer. Third, they should enlist a broad group of co-investigators, particularly individuals who represent a group that might feel threatened by the study. Finally, they should seek the views of organizations and people who have the potential to be affected or offended by the result.
There are many important socially sensitive questions where data and evidence could help improve policies that have been left unstudied due to their controversial nature. These include the impact of race on police and judicial actions, age and competence in the work place, the dynamics of HIV transmission in homosexual and lower socioeconomic status groups, and abortion.
Academics must feel free to pursue enquiries without fear of retribution from others simply for asking the question. However, they must be sensitive to the human milieu in which they operate and recognize ways to mitigate those influences. The bottom line is that being afraid to ask questions because the answers may be unpopular is not a good thing; however, we also must realize that we live in the real, opinionated world.
Footnotes
-
Previously published at www.cmaj.ca