Assessing participant group affiliation and attitudes towards CTSI services

Abstract

Analyzing open-ended survey text responses holds the capacity to yield greater insight about participants’ perceptions of clinical translational science institute (CTSI) initiatives. Few translational research studies have explored their effectiveness. The aim of this mixed methods analysis was to assess participant perspectives of the impact and effectiveness of our CTSI program and services. We selected two open-ended survey question items (how CTSI benefitted research, and the most important impact of the research facilitated by the CTSI) from a larger set and compared responses by participant affiliations (clinical/non-clinical; lab/non-lab). We used a three-step analysis. First, nodes were generated using NVivo word frequency function. Next, with the aid of Python, we used sentiment analysis to classify each node (as positive, negative, or neutral) to indicate participant ratings toward their experiences with the CTSI and computed the average differences between groups. Third, we selected nodes that met pre-established criteria and report the qualitative distinctions. We recommend using precisely worded open-ended questions in future annual surveys or administering a survey using only opened-ended questions every six months.