Improving data quality in online parenting research

Abstract

AbstractOnline surveys provide many benefits for researchers but are vulnerable to fraudulent responders that threaten data quality. We aimed to evaluate the effectiveness of current detection strategies after noting patterns in our data. We systematically reviewed responses (N = 180) from parenting studies utilizing two online platforms. Respondents were categorized as accepted (n = 45) or fraudulent (n = 135), and responses on standardized measures were examined to determine if fraudulent scores differed. Accepted participants were predominately female (98.4%) and White (67.9%). Average caregiver age was 40.61 (SD = 7.37) on Qualtrics and 59.88 (SD = 6.01) on REDCap. The majority of fraudulent responders reported being female (62.1%) and White (80.6%), with average caregiver ages of 37.83 (SD = 7.35) for Qualtrics and 56.16 (SD = 8.86) for REDCap. Fraudulent responders had significantly higher median scores on measures of caregiver distress (rDASS = 0.21) and child behavioural problems (rCBCL = 0.33). Previously recommended strategies were insufficient at preventing and identifying fraudulent responders. Recommendations for improving these strategies are discussed.Highlights The current paper includes experiences using multiple strategies to detect fraudulent responders in two online parenting studies. Improvements to previously recommended strategies and an example review form are included to aid researchers in detecting fraudulent responders.