Page - 255 - in Intelligent Environments 2019 - Workshop Proceedings of the 15th International Conference on Intelligent Environments

The influential variables are effort required to obtain a reward and rate (or size/amount), quality, and delay between responding and gaining the reward [26]. Reed and Kaplan [26] found that, albeit with limited populations, the option most likely to be rewarding in preferred ways, to a large extent, and immediately for the least amount of effort is most likely to be chosen. Participants in the applied research have typically been individuals with developmental or behavioral disorders affecting motivation. However, one relevant study explored how play decisions are made in American football [27] and another evaluated matching experimentally via a commercially available basketball video game [28]. Another study by Borrero and associates [29] predicted college students’ attention to a juvenile delinquency presentation. While most studies were conducted to enhance educational outcomes, none of them focused on how adults make educational decisions. As Reed and Kaplan [26] further pointed out, the extent to which each of the four factors (rate/amount, quality, delay, and effort) predicts choice varies per individual. That is, one student may emphasize the amount of reward, while another one may prioritize whichever reward comes soonest. The idiosyncrasy parallels the students in the PLE study [15], in choosing unique sets of Web 2.0 tools compared to each other and to previous research. The individual nature of preferences for rewards and Web 2.0 tools necessitates adopting an expansive view to account for as many preferences as possible. Some choice-making factors may match with predictive factors identified above. For example, the Perceived Ease of Use variable may map onto effort. Any variables related to tool performance, usefulness, functionality, sharing, enjoyment, or social interaction may align with quality. Three areas have not been addressed in statistical models: (a) the effect of having a choice (vs. instructor choice); (b) rate/amount of reward associated with each choice (e.g., a final grade vs. an assignment grade; single vs. multiple rewards); and (c) delay (e.g., how quickly students can find resources). Finally, the choice model discussed here is known to predict directly observable behavior. The extent to which it predicts verbal reports of preference on surveys is unknown, necessitating direct measures of dependent variables when using this model. 5. Conclusions Educational research about Web 2.0 is expansive in terms of geography, platforms studied, academic disciplines, and course delivery formats. Widespread integration of Web 2.0 in higher education is challenging, although PLE studies offer insights about potentially effective integration. Factors predicting the endorsement and use of Web 2.0 are many, and a synthesis of recent models suggests that only a few have generalized effects. Despite calls for student choice in the integration of Web 2.0 in their learning, no models have explicitly tested it to our knowledge. Therefore, we plan to create a Choice Model through a future empirical study, which will be based on the Matching Law that has shown promise in other areas but has seen very little testing in the context of higher education. Determining the value of Web 2.0 tools in academia has been a growing priority in educational research. Improving student engagement, academic outcomes, and lifelong learning are strategic goals. Based on those aims and our findings, we suggest several areas for future research. First, more study on the PLE is needed. Specific questions may test optimizing a PLE, how an optimal PLE affects academic and non-academic performance, and whether it facilitates lifelong learning. In that vein, it is also E.Damianoetal. /Bridging theDivide: TheCurrentStatusofWeb2.0 inHigherEducation 255