| dc.description.abstract | Despite the decreasing rate of enrollment for face-to-face classes at higher education institutions, the last decade has seen an increase in enrollment for classes online. However, online learners suffer from several challenges, which may hinder desirable learning outcomes. These challenges include feelings of isolation as well as a lack of academic and social interaction, which has led to higher dropout rates and lower retention rates. All of these challenges have been associated with social presence, or a lack thereof, in online educational environments. Previous researchers have identified three indicators that influence social presence (i.e. affective association, interaction intensity, community cohesion). However, the way in which these factors were measured calls into question the validity of the research. Moreover, recent research has re-examined the original three indicators and added two additional indicators theorized to affect social presence (i.e. instructor involvement and instructor knowledge and experience), forming the social presence model (SPM). This study: 1) evaluated three scales to measure the three indicators of social presence; 2) examined the relationships of these three indicators to social presence; 3) developed two additional scales to measure the two additional indicators of social presence; and 4) inspected the full SPM. The participants of the study consisted of 411 students taking fully online courses in the United States. For the measurement models, the author conducted six item factor analytic models to evaluate the social presence construct scale as well as the five construct scales of the social presence indicators. The results indicated a good to excellent fit of the measurement models as well as high correlation coefficients between social presence and the original three indicators of social presence. Conversely, the results indicated lower correlation coefficients among the three original social presence indicators with the two additional indicators. To examine the structure of these relationships, the author conducted two structural equation models (SEMs). The first SEM inspected the three original factors, and the second SEM inspected the full SPM. The results of the first SEM indicated that the three social presence indicators are highly associated with social presence as well as with one another. They are statistically significant predictors of social presence; social presence is mostly affected by affective association (ß = 0.522, P < 0.001), community cohesion (ß = .226, p = .001), and then interaction intensity (ß = 0.163, p = .027). For the second SEM, the results indicated that affective association was found to be a significant predictor (ß = .507, p < .001); interaction intensity was not found to be a significant predictor (ß = .120, p = .080); group cohesion was found to be a significant predictor (ß = .173, p = .009); instructor involvement was not found to be a significant predictor (ß = .092, p = .120); and, finally, instructor knowledge and experience was not found to be a significant predictor of social presence (ß = .085, p = .098). Affective association, as demonstrated through means such as humor and self-disclosure, is the most critical among the five social presence indicators. Sharing attitudes, feelings, personal experiences, and interests with one another is a highly effective way to increase social presence. The higher degree to which online community participants feel part of a cohesive group, the greater the degree of social presence. When instructors deliberately design activities to encourage affective association, they will also, in effect, enhance the levels of community cohesion and intensity of interaction. Although the involvement of instructors and their previous knowledge and expertise in the field are contributing factors to positive outcomes in online education, they were not found to contribute as much as the three original factors. The author concludes by discussing the implications as well as the limitations of these findings and suggests future research. | |
