Analyzing List-style Open-ended Questions: Combining Texts From Individual Answer Boxes Improves Classification with Language Models

Abstract

List-style open-ended questions allow for multiple answers. Previous research on the design of such questions found that providing multiple small answer boxes yields more and richer answers than providing one larger answer box. Using a series of classifiers based on the Bidirectional Encoder Representations from Transformers language model, we empirically study how this design choice affects the classification of such answers. We design a 2 × 2 factorial experiment: (i) analysis with a multi-label versus single-label classifier and (ii) answers obtained from one larger answer box versus multiple smaller answer boxes. We find that the multi-label classifier gives more accurate results than the single-label classifier (1 percent versus 9 percent misclassification of individual labels), regardless of how the answers were obtained. Surprisingly, analysis with a multi-label classifier is preferable. We attribute this success to the classifier’s ability to use label correlations. We conclude that list-style open-ended questions should continue to provide multiple answer boxes due to better data quality. However, answer boxes should be concatenated for analysis to improve classification performance.