Self-Report

Self-report is one of the most commonly utilized methods of data collection. Surveys, questionnaires, and interviews are all forms of self-report that rely on individuals' subjective evaluations and reports of their thoughts, feelings, behaviors, or experiences. Most often, self-report is used to gather personal information that cannot be obtained objectively. An individual's self-report may also be of interest in circumstances in which some degree of objective evaluation is possible (e.g., a patient's subjective reports of symptoms). In many settings, such as medicine, policy making, and opinion polls, important decisions are made on the basis of self-report data.

There are two broad categories of self-report data: (a) unstructured, open-ended responses, and (b) structured, fixed-response questions. Open-ended reports have the advantage of soliciting detailed information that may not be captured using closed-response formats. In many cases, this approach is optimal because it provides a wealth of qualitative data. However, such responses do not easily lend themselves to statistical analysis without the use of coding procedures, which can be labor intensive and difficult to develop.

Closed-response formats generate quantitative data much more easily. However, decisions regarding the use of checklists, number scales, or categorical endorsements can have a dramatic impact on response distributions and the subsequent presence or absence of significant statistical effects. Research has demonstrated that by simply altering the response format of a questionnaire, one can generate different responses to the same questions. For instance, decisions regarding whether to anchor a response scale with a midpoint, whether to use a bipolar (-5 to +5) or unipolar (0 to 10) number scale, and whether to label responses with verbal quantifiers like “frequently” or “very often” can influence one's perception of the questions and subsequently the responses one endorses.

A number of cognitive constraints can also interfere with one's ability to generate accurate self-reports. Both the time scale and the regularity of a behavior can alter the memory retrieval strategy used to recall events, while the accessibility of recent experiences may influence more general estimates. Similarly, measures designed to evaluate emotional traits or memories are often biased by the influence of current mood states. Research on such contextual effects has also demonstrated that responses to self-report items may be directly influenced by the content and placement of previous self-report items (i.e., later [Page 878]answers are influenced by earlier answers). Furthermore, respondents' conscious beliefs regarding the confidentiality of their reports can affect their disclosure of personally sensitive information (especially regarding threatening or stigmatized topics).

Critically evaluating questions to ensure that they are presented clearly, framed in the proper context, and accompanied by appropriate response formats can help prevent self-report data from being compromised by measurement constraints or response biases. Clearly informing respondents as to the intended use, privacy, and protection of self-report information can also reduce self-presentation concerns and facilitate more veridical reporting. A number of innovative self-report methodologies, such as daily diaries and ecological momentary assessment, have addressed some of these concerns by considerably limiting the recall periods (i.e., to a day or even a few minutes) and providing an ecologically valid alternative to lengthy retrospective reporting. By carefully considering these issues, researchers can effectively use self-report as a fast, cheap, and practical method for collecting personal information across a variety of research and applied settings.

...