Sentence Completion as a User Experience Research Method: Recommendations From an Experimental Study

Abstract

The aim of this study is to investigate the use of the sentence completion technique (SCT) as a user experience (UX) research method. We conducted an online experiment (N = 400) to test the effect of sentence stem variations on sentence completion outcomes. Using a between-subjects design, half of the participants were exposed to impersonal sentence stems that did not include pronouns while the other half were exposed to stems formulated using first-person pronouns (PR). Additional hypotheses around stem formats (use of redundant stems, imaginative stems, two blanks stems, generic stems, stems prompting others’ perception) were tested using a within-subject design. The results do not support hypothesized differences between the pronoun and no pronoun condition. Findings however show that varying the format of the stem influences response behavior, as measured by variety, quantity and novelty of ideas, as well as the length of response. This study contributes to consolidating the use of SCT as a user research method and proposes actionable recommendations on how to create optimal sentence completion surveys in Human-Computer Interaction (HCI).