Identify two ways that a researcher might introduce bias in his or her data coding and interpretation. Explain ways to address these researcher degrees of freedom ethically.

Question

What will be an ideal response?

Answer 1

Ans: There are many ways that a researcher could, whether intentionally or not, “cook the data.” Two such ways include rounding errors that favor the researcher finding statistical significance and opportunistic bias in deciding what to do with outliers. First, rounding errors could occur when interpreting the significance level of a finding. Since we report two places after the decimal and there is a possibility of a p value of 0.0456, for example, one person might round up and another not. The remedy for this is to be very clear, before the research, concerning clear cut rules for rounding or any other decision to be made. Second, the opportunistic bias refers to decisions researchers make concerning dealing with outliers in a self-serving manner. There may be genuine errors, for example, when participants answer outside the possible anchors in a paper administration. In those cases, the researcher should check the original to make sure it was not an input error, and if it was really a participant error, delete the item for that person. On the other hand, where there are outliers that could be legitimate answers, a clear rationale must be given. Then, complete transparency should occur in reporting. The researcher should indicate when these outliers were included as well as when they were excluded. Overall, there are many ways a researcher could tilt the results in his or her favor. Since the ultimate goal of research is to further the body of knowledge, the rush to publication should be avoided, and ideally, people’s jobs should not ride on finding significant results. The way to avoid ethical problems is to create rules beforehand (if these rules exist in the literature already, follow them), have clear and defensible rationales for any actions taken, and be transparent in reporting.