How Accurate Are Self-Reported Recreation Participation Rates?

Why self-reported participation is a measurement problem

Participation surveys ask people to report how frequently they performed an activity during a past period. Responses depend on memory, question interpretation and how respondents estimate frequency. These processes can introduce errors even when questionnaires are administered consistently.

The 1983 research note by D. R. Chase and G. C. Godbey examined whether respondents accurately recalled participation when external records were available. Such comparisons help identify discrepancies between reported and documented behavior.

What the 1983 research note tested

The research examined two recreation-club settings in central Pennsylvania. The first study, conducted in spring 1981, investigated tennis-club members' recall of playing frequency. A second study in spring 1982 replicated the approach with swimming-club members.

Club settings allowed researchers to investigate reported participation alongside independent attendance information.

How later research fits the same measurement question

A related 1984 paper by David R. Chase and Munehiko Harada reported substantial overestimation of recreation participation. Most reported frequency estimates, except very small values, ended in 0 or 5. The study also found mixed results when self-reports predicted actual swimming participation.

These findings demonstrate the importance of recall accuracy but belong to a separate investigation. They should not be treated as numerical results from the 1983 note.

Evidence from recreation and fisheries surveys

Later recreation research also compared retrospective responses with diaries or records. Connelly and Brown's 1995 Lake Ontario study found that anglers overestimated fishing days by 44–45% on a 12-month recall questionnaire after adjustment for changes in fishing activity.

A 2018 fisheries study by Andrews, Papacostas and Foster compared one- and two-month reference periods. It examined telescoping, when events are assigned to the wrong period, and omission, when events are forgotten. The study found no significant differences between its main two-month survey estimates and experimental alternatives, although a bounded multiple-period approach showed potential to reduce bias in one-month estimates. Together, these findings indicate that recall errors depend on activity, reference period and survey design.

What this means for survey design

The 1983 study highlights an important question: whether reported participation reflects remembered events, estimated habits or rounded frequencies. Self-report remains useful, but its accuracy should be assessed when participation counts influence research conclusions. 

• Practical safeguards include:

• Use shorter, clearly defined reference periods where feasible.
• Compare survey responses with diaries, attendance logs or independent records.
• Include memory cues or bounded questions when appropriate.
• Pilot-test whether respondents recall events, estimate typical rates or round their answers.
• Distinguish self-reported estimates from objectively recorded participation.

These measures can improve interpretation, although none guarantees error-free reporting..

Key limitations when interpreting the 1983 study

The research involved tennis and swimming club members in central Pennsylvania during the early 1980s. Its findings cannot automatically be generalized to other activities, populations or modern survey methods.

The two club settings provided opportunities to examine recall accuracy, but the results do not establish a universal level or direction of reporting error. Later methodological studies offer broader evidence that memory, response strategies and reference periods influence behavioral-frequency estimates.

Frequently asked questions