Leveraging Emotional Learning Process (ELP) Data-based Interventions in Undergraduate Computing Education

Carla Marcela De Lira

doi:10.7273/000006564

The increasing demand for a diverse pool of computing talent combined with a persistent shortage of skilled workers, particularly from underrepresented groups, has engendered a need to support students pursuing computer science (CS) careers. This dissertation presents the results of an empirical study on the effectiveness of using emotional learning process (ELP) data to support community building in CS courses, particularly introductory programming courses. Building community, such as communities of practice, plays a part in retention, as students often cite social isolation and lack of support as reasons for withdrawing from computing programs. This is especially true for those from underrepresented groups. We designed and implemented the HELPd Empathy Tool (HELPd), integrating a customized IDE to gather programming behavior data and a private website to collect ELP data. HELPd used both sets of data to generate ELP data-based interventions. Interventions occurred but, unfortunately, participants did not act on them and, thus, we were unable to evaluate the usage of HELPd for community building. However, the study did result in some useful data. Participants were recruited from a class of 73 students enrolled in a semester-long CS0 programming course and were offered extra credit and a $10 gift card for completing the study. At the conclusion of the study, we were able to evaluate data from 19 participants. Data were gathered over an 8-week period divided into two parts. The first four weeks were the control period when programming behavior data were collected to establish a baseline. The treatment period included ELP data collection and ELP data-based interventions in addition to the programming behavior data collection. Survey data were collected at pre-, mid-, and post-time points, that is, before the start of the 8-week period, before the start of the treatment period, and at the end of the treatment period. One- and two-way repeated measures ANOVAs were conducted to explore the relationship of dependent variables, such as classroom community, empathy, and intention to persist, and the demographics of students, such as age or major. The results of our one-way repeated measures ANOVAs indicated no significant change in pre-, mid-, and post-scores for each dependent variable. The two-way repeated measures ANOVAs, however, showed computing and non-computing majors differed significantly in their intention to persist scores throughout the study. Although we did not see significant differences with our ANOVA analysis in general, we used ELP, programming behavior, and demographic data to produce counts for changes in survey responses over time. While changes were noted between pre/mid and mid/post scores, no patterns emerged most likely because of the limited number of study participants. Based on an analysis of our results, we recommend design approaches for future iterations of ELP data-based interventions as well as the direct integration of these interventions into classroom coursework. We hope these recommendations will enable other researchers to ascertain the effectiveness of ELP data-based interventions in community building through help-seeking and help-giving actions. We also discuss further customizing interventions to engage populations from underrepresented groups and from groups with other demographics.

Leveraging Emotional Learning Process (ELP) Data-based Interventions in Undergraduate Computing Education

Files and links (1)

Abstract

Metrics

Details