Lack of Diversity in Social Science Research

In college, friends majoring in biology told me I should switch majors because social science isn’t real science. It’s too subjective. Your personal biases will cloud the data.

They had a point. But researchers from a wide array of backgrounds can question each other’s assumptions, which can mitigate personal bias somewhat. Over the past 50 years universities have done a laudable job of trying to encourage more women and minorities to enter white male dominated fields. And while fields such as physics still lack diversity, 60% of biology degrees go to women, and psychology has an even larger number of women.

But contrarians say we’ve overlooked something. What about political diversity? Yet, academia has spent the past half century trying to purge conservatives, or even those who are not die hard liberals.

Does social psychology really prove that conservatives are unethical dullards? Can we trust the objectivity of a field that has almost no non-liberals? (Non-liberal because not every alternative viewpoint is conservative, or even libertarian.) Imagine for a moment that almost all social scientists were evangelical Christians, and their research found that atheists really are nasty people. Would you think something is amiss?

Jonathan Haidt writes that a century ago, the social sciences were almost evenly split between liberals and conservatives. But the gap started to widen, slowly at first, but then rapidly after 1990. Today, the ratio of liberals to conservatives is almost 14 to 1.

Unchecked biases degrade the quality and validity of research. Chief among these biases are negative presuppositions and confirmation bias (failing to critically examine or search for contradictory evidence for something you already believe). This can lead to “mischaracteriz[ing] liberals and conservatives alike.”

This doesn’t affect most aspects of social science research, such as personality theory or the psychology of decision making. But these biases are notable with areas of liberal concern, such as sex and gender, race, inequality, and moral and political psychology. And it can leave unexamined areas outside of liberalism’s concerns.

In the social sciences, the narrative of liberal progress is like water to a fish – it’s everywhere but often goes unnoticed. But this can lead to misinterpretation of non-liberal value statements. For example, social scientists might label someone unethical for not siding with a coworker who has filed a sexual harassment claim. But without someone to question the assumption of misogyny, the judgment of moral inferiority is unexamined.

In a previous post I wrote about a friend who received a sexual harassment complaint for using the phrase “OMG.” I think her claim was frivolous. My reasons are that I think a person is innocent until proven guilty (and the burden of proof is on her), and her failure to present any evidence other than her personal opinion is not sufficient evidence. But my perspective contradicts the liberal notion that an alleged victim must always be believed. This is not misogyny, however. Due process is a human right.

Too often people present statistics from dubious sources or which lack context, often arguing that numbers don’t lie. But numbers do lie. Ever made a math error? And too often someone will cite one study as if that seals the case, failing to question the researcher’s methodology, possible biases, and (most of all) failing to understand that studies must be replicated numerous times before being accepted as true.

Social science has a significant blind spot, and any research findings with political implications should be approached with a healthy dose of skepticism.

Advertisements