Predictors of faculty sentiment on their transition to online teaching

In the spring and summer of 2020, we collected data on faculty experiences during their transition to online teaching as a result of the COVID-19 pandemic. Data on the participants' institutions, job security and position, preparation time, and anxiety were collected, as well as a free text response to add anything the participants felt was relevant. In total 364 text responses were collected. Using natural language processing tools and multiple lexicons, sentiment scores were calculated for each response. The sentiment was found to be positive for three of the four lexicons used. Each variable in the dataset corresponds to a particular answer option to a question in the survey, for a total of 66 variables. To determine which variables may correlate with participants' responses, a lasso regression was run for each lexicon's sentiment. This presentation will include a discussion of this process and any variables that may be well correlated with sentiment.