Setting the Record Straight: Assessing the Reliability of Retrospective Accounts of Change

Abstract

Ecological degradation is accelerating, reducing our ability to detect and reverse declines. Resource user accounts have the potential to provide critical information on past change but their reliability can rarely be tested, hence they are often perceived as less valid than other forms of scientific data. We compared individual fishers' catch records, recorded 1-50 years ago, with their memories of past good, typical and poor catches for the corresponding time period. Good and poor catches were recalled with reasonable accuracy, matching variability in recorded catch with no significant change observed over time. Typical recalled catches were overestimated and became significantly more exaggerated over time, but were more comparable to mean than median recorded values. While accuracy of resource users' memory varied with the type of information recalled, our results suggest that carefully structured interview questions can produce reliable quantitative data to inform resource management, even after several decades have elapsed.