Review of Protection Against Bots and Fraudulent Survey Submissions in Nursing Research

Abstract

BACKGROUND: Online survey-based research is common in nursing, but potentially fraudulent responses (e.g., bots or bad actors) threaten the validity of the data obtained from these studies. OBJECTIVES: The purpose of this methods paper was to quantify and describe the extent to which nursing investigators have protected against fraudulent submissions in online, survey-based research studies. METHODS: A random sample of articles published in nursing journals and reporting the results of online, survey-based research was obtained from PubMed in June 2024. Each article was audited using a standardized audit matrix. RESULTS: The included articles (n = 132) published in 51 unique journals and involved 56,159 participants. Studies were primarily cross-sectional and conducted worldwide, with most originating in the United States, multiple nations, or China. Investigators mentioned screening for response validity in only 21 articles; fewer explicitly described the screening processes used. Screening strategies included reviewing open-ended responses, checking for response-set biases, reviewing completion times, using computer-assisted tools, and examining responses for implausible values. In all cases where a potentially fraudulent response was identified, investigators excluded the response from analysis. There was no significant difference in frequency of fraud screening across journal impact factor tertiles. DISCUSSION: Fraudulent responses are an ever-present problem in survey research. The articles examined did not routinely report strategies for detecting potentially fraudulent responses or protecting data quality. Published online, survey-based studies that include methods for detecting fraudulent responses enhance reader confidence. Investigators are encouraged to develop an a priori data analysis plan that includes multiple means to identify and eliminate, or otherwise, process fraudulent responses. We suggest that investigators transparently detail the use of a standard checklist for online survey research, in addition to the Fraud detection strategies, Recruitment, Incentive, Excluded responses, Data collection (FRIED) checklist we propose in this article.

Ähnliche Arbeiten