The good-old-days bias and post-concussion syndrome symptom reporting in a non-clinical sample

Abstract

Primary objective: To investigate the good-old-days bias, a psychosocial factor associated with post-concussion syndrome (PCS).Study design: Repeated measures comparison of PCS symptoms reported 6 months prior (retrospectively) and currently. A non-clinical sample was used to determine if this bias is a general recall bias.Methods and procedures: Fifty-seven university students with no history of brain injury or neurological disease completed the British Columbia Post-concussion Symptom Inventory. Symptoms were reported on two occasions, spaced 1 week apart, commencing with current symptoms.Main outcomes and results: Using PCS symptom frequency by severity product scores, there was no significant difference in the 13 PCS symptoms reported across occasions, nor the relevant summary score (p = 0.199). These data do not support the presence of a general recall bias. However, significant differences emerged when analysed using a simple count of the total number of endorsed symptoms (p = 0.002, d = 0.39, small-to-medium effect) or the sample percentage that endorsed each symptom (four symptoms were endorsed by fewer participants retrospectively than currently).Conclusions: There is only weak evidence of a general recall bias in this non-clinical sample. Further consideration of the methods used to study this bias and its role clinically is needed.