By Jason Hreha

Glossary

Self-Report Bias

Published 3 min read

“How often do you exercise?” looks like a simple question. But does walking to the station count? What about a ten-minute stretch? Are you describing last week, an ordinary week, or the sort of person you believe yourself to be? The number comes after those decisions.

What is self-report bias?

Self-report bias is systematic error in reports of one’s own behavior, characteristics, beliefs, or experiences. It is not a single mental shortcut, and it does not mean that respondents are necessarily lying. A questionnaire can collect precise-looking answers to a question that people understood in different ways.

One possible source is social desirability bias: answers shift toward what seems acceptable or favorable. But forgetting an event, interpreting “regularly” differently, or applying a new standard of self-evaluation can also change an answer. These explanations call for different checks. Systematic errors in reconstructing past events are a form of recall bias.

When the question changes the report

Brenner and DeLamater studied exercise reports using a web survey, activity reports sent by text, and recreation-facility records. Participants sent activity updates for five days; instructions either mentioned the study’s exercise focus or did not. The authors found overreporting in direct survey measures and reactivity in the exercise-focused text condition. They interpret the pattern through identity as well as conventional impression management.

The useful distinction is between reporting behavior differently and changing behavior because it is being measured. Both can complicate an exercise study. This study does not establish that an unprompted diary is always accurate, or that facility attendance captures every kind of exercise.

A lower rating can accompany better understanding

Consider a hypothetical management course. Before it starts, a participant gives their listening skills nine out of ten. After learning to distinguish listening from waiting to speak, they give themselves six. The lower score could reflect worse listening. It could also reflect a stricter understanding of what good listening requires.

This is a possible response shift: the standard used to answer has changed. Rosenman and colleagues discuss this problem in evaluating interventions. A before-and-after rating cannot, by itself, tell you how much changed behavior and how much changed self-judgment contributed to the difference. Calling every inconvenient result “response shift” would be just as careless as ignoring the possibility.

Match the check to the claim

If you want to know whether someone attended a session, an attendance record can help. If you want to know whether the session felt intimidating, their account is directly relevant. A door-entry log does not contain that experience. Self-report is a method to evaluate for a purpose, not a reason to dismiss a finding automatically.

Start by asking respondents what they think a question means. Define the behavior and reporting period. Explain who can see the answers, especially for sensitive topics. Where another measure is appropriate, compare it with the report rather than assuming either is perfect. Althubaiti’s discussion of self-report and validation describes both the value of people’s accounts and the need to examine their accuracy.

For a personality questionnaire, the corresponding question is whether there is evidence for its intended interpretation and use. The guide to evaluating personality tests walks through that decision. A score can be informative without being an objective instruction about how to live.

References