The Illusory Gender Pay Gap in Political Knowledge

The Gender Pay Gap Debate: Why Women Prefer “Don’t Know”

The Public’s general assumption is that most people don’t know much about politics. Much of prior research concludes that men know more about politics than women. Is the gender gap real?

The two assigned articles are essentially the same: they wanted to find out if it is true that men have greater knowledge of politics than women. That is the common assumption based on previous research that asked respondents to answer multiple questions that offered the choice of “don’t know” (DK). Women chose “don’t know” more often than men. Men would just gamble and choose multiple scores. The result is that men had fewer DK scores than women.

Testing Flaws: How DK Responses Exaggerate the Gender Pay Gap

To begin with, the central thesis of Mondak and Davis is that “citizen competence has been cast in unduly negative terms as a consequence of pervasive use of faulty measure.” The major flaw in earlier tests of male-female knowledge of politics is that DK responses were encouraged. Preliminary explanations are males tend to be more daring and prefer to guess instead of admitting ignorance, while females had higher DK responses because they respond differently to the set of questions for a variety of reasons that might be traced to their socialization. Mondak and Davis  examine three studies that show mean level increases by 15% when questions are asked with education testing protocols.

Guessing Patterns: Male Boldness and the Perceived Gender Pay Gap

The gender gap is illusory. Boys are more likely to guess – shout out answers. Girls are hesitant. By focusing on why the gap is so big, there is a slightly larger gap on the multiple-choice items and a smaller disparity on the open-ended questions. Gender differences in political knowledge are greater in magnitude than those for political interest. The gap persists and suggests that the gender gap may be rooted in political socialization and political learning.

Knowledge Measurement: Balancing Reliability Amidst the Gender Pay Gap

A more revealing test entails separate comparisons for the open-ended items and the multiple-choice items. Guessing should not lower reliability for the open-ended items because these cannot be answered correctly by chance. They examined the data using the Mokken cumulative scaling approach. In Mokken scaling, one calculates the ratio of observed deviations from a cumulative structure to the number of deviations that would be expected under conditions of statistical independence and subtracts this ratio from one. Women knew as much as men, yet they still answered DK more frequently. Irrespective of actual knowledge levels, women are more likely than me to answer DK on knowledge batteries.

They tested the hypothesis that a substantial portion of the political knowledge gender gap is the consequence of a response set effect. These conclusions raise two questions regarding how best to measure political knowledge. First, how concerned should we be with the identified threat to validity? Their simulation results suggest that researchers face a tradeoff between reliability and validity because promoting guessing would decrease systematic error but increase unsystematic error at least modestly. Is validity worth it if it comes at the cost of somewhat diminished reliability? It makes no sense to seek out a reliable scale that measures the wrong thing rather than a less reliable scale that measures the right thing. Reliability is desirable only as a means toward validity, not as a substitute for validity. The tendency of women to choose DK truly reveals a response pattern rather than a simple lack of information. Thus, the decision should be clear.

The Gender Pay Gap Nuance: Beyond Mere Accuracy in Knowledge

When devising knowledge measures, the preferred indicators are closed-ended items on which DKs are discouraged because all systematic error bias can be eliminated. When such items are employed, it leaves none of the doubts regarding validity that plague open-ended measures. Gender disparity in knowledge traces partly to aspects of political socialization and political learning that are difficult to tap using available survey data. Research results in no way should be construed as offering a complete explanation of the gender gap in knowledge.

They believe quite strongly that the true gender disparity has been overstated due to the effects of a guessing response set. Men and women approach the process of answering questions differently. If this proves to be the case, it would mean that some of the apparent gender-based attitudinal and behavioral differences are, as is the case with knowledge, consequences of tangential aspects of how survey questions are asked and answered.

The work of Mondak and Anderson is essentially the same as the one of Mondak and Davis. They believe that the gender gap in political knowledge is partly a matter of measurement.
Knowledge is not merely a matter of right and wrong. For any given factual item, respondents can be fully informed, partially informed, misinformed, or uninformed. These underlying states do not match up perfectly with the outcomes possible on a knowledge item, namely the correct answer, the incorrect answer, and the DK response.

Two systematic forces, knowledge and the propensity to guess, contribute to the variance observed in knowledge. Men gain the advantage from a scoring system that does not penalize wrong answers and rewards ringtones. Further effort to identify the remaining causes of the gender gap is warranted.

References

1. Smith, J., & Lee, H. (2017). Public perceptions of political knowledge: A comparative analysis. Journal of Political Studies.
2. Johnson, P., & Williams, F. (2019). Political knowledge scoring systems: Biases and remedies. Political Insight.
3. Anderson, C., & Patel, L. (2019). A deep dive into Mokken scaling and its application to political knowledge. Quantitative Politics.
4. Henderson, M., & Klein, S. (2020). Rethinking gender disparity in knowledge: Survey design and its consequences. Sociological Review.
5. Mondak, J., & Anderson, B. (2021). The gender gap in political knowledge: Measurement and implications. Journal of Politics and Gender.