Online education is unique in part for the relatively high degrees of autonomy afforded learners. Self-direction and self-regulation, along with support, are essential for students to succeed. The site of this action research project was a new, small online public charter school for middle and high school students, Foothills Academy Connected (FAC). The purpose of this action research project was to develop an online learner support system that was built around mentorship and based on the four areas identified by the Educational Success Prediction Instrument (ESPRI) (Roblyer & Davis, 2008); thoroughly document the process; and examine its influence on students and the researcher. This study was focused on: (a) identifying students’ main challenges with online learning, (b) identifying students’ perceptions about additional supports that would improve their schooling experience, and (c) examining the process of engaging in mentorship by the emerging mentor, herself.

The study employed a mixed methods research design. Research instruments included a questionnaire adapted from the ESPRI that marked the start of the study period, visual autoethnographies, interviews, extensive research journaling to document interactions with students and parents/guardians, and a second questionnaire. The research results showed that the “emerging mentorship approach” was a worthwhile innovation for augmenting the FAC online learner student support system. In particular, developing individual student profiles based on this varied data and responding to those students’ needs were accompanied by detailed documentation to develop a mentoring approach that could be used subsequently. A finding of the research was that the ESPRI would not have been effective alone in determining a student profile and responding only on that basis. The ESPRI areas of inquiry were helpful when used in conjunction with the other data to frame students’ needs and formulate personalized plans to support struggling online learners. Online learner support literature provided scant detail on the personal experience of the individual adopting the mentor role. In this study, it was determined that the process of becoming a mentor was uncomfortable and nonlinear, and it challenged the self-directedness and boldness of the action researcher as she worked in this new role as mentor.

This dissertation explores different methodologies for combining two popular design paradigms in the field of computer experiments. Space-filling designs are commonly used in order to ensure that there is good coverage of the design space, but they may not result in good properties when it comes to model fitting. Optimal designs traditionally perform very well in terms of model fitting, particularly when a polynomial is intended, but can result in problematic replication in the case of insignificant factors. By bringing these two design types together, positive properties of each can be retained while mitigating potential weaknesses. Hybrid space-filling designs, generated as Latin hypercubes augmented with I-optimal points, are compared to designs of each contributing component. A second design type called a bridge design is also evaluated, which further integrates the disparate design types. Bridge designs are the result of a Latin hypercube undergoing coordinate exchange to reach constrained D-optimality, ensuring that there is zero replication of factors in any one-dimensional projection. Lastly, bridge designs were augmented with I-optimal points with two goals in mind. Augmentation with candidate points generated assuming the same underlying analysis model serves to reduce the prediction variance without greatly compromising the space-filling property of the design, while augmentation with candidate points generated assuming a different underlying analysis model can greatly reduce the impact of model misspecification during the design phase. Each of these composite designs are compared to pure space-filling and optimal designs. They typically out-perform pure space-filling designs in terms of prediction variance and alphabetic efficiency, while maintaining comparability with pure optimal designs at small sample size. This justifies them as excellent candidates for initial experimentation.