How we Detect and Handle Incoherent Answers in Open-ends

Overview

The Algorithm for Incoherent Answers Involving Open-Ends is a crucial data quality check we implement to ensure consistency throughout the survey. This check helps us identify coherence between closed and open-ended questions, ensuring that respondents’ answers remain aligned.

For instance, consider a questionnaire where the first question is an open-ended prompt asking, “Which brand do you know?” followed by a closed question, “Among these brands, which one do you know?” If a respondent provides an answer to the first question that does not appear in the options given in the second question, this discrepancy indicates a lack of answer consistency, which the algorithm is designed to spot.

Fraudulent respondents often attempt to bypass standard quality checks, but coherence checks are significantly harder to avoid. These respondents typically lack knowledge of the questionnaire topic and the potential links between questions. Additionally, this quality check also flags “unengaged” respondents who answer randomly, resulting in incoherent responses.

By employing this algorithm, we enhance the overall quality of our data, ensuring that only consistent and engaged responses contribute to your insights.