Orchestrating Feedback and Help-Seeking in Computer Science Classrooms (Poster 3)

• In Event: Theory-Driven Development of Classroom Orchestration Tools to Support Learning at Multiple Social Planes

Abstract

Within a computer science lab, the goal is to provide students with a hands-on experience working on programming problems. A lab setting provides a structured environment where students can seek help and receive immediate, formative feedback. This formative feedback has been shown to be beneficial to the learning process (Hattie & Timperley, 2007). Through feedback, students can work towards reducing the discrepancies between where they are and their goal. When this feedback is immediate rather than delayed, it can be more effective (Scheeler et al., 2004), which lab sections can facilitate. However, there are two significant barriers to students receiving this feedback in labs. 1) Students struggle engaging in productive help-seeking processes that would support them in getting feedback. When programming, two unproductive help-seeking behaviors students engage in are help avoidance and unneeded help (Marwan et al., 2020). 2) Instructors have a limited amount of time and are not always immediately available to support groups. These behaviors can prevent students from getting beneficial feedback.
To address these issues, we are developing TUPLE (Technology sUpport for Programming Lab Environments), an orchestration system to support hands-on individual or paired programming during lab sessions. During these sessions students have programming tasks that they need to complete related to class topics. Students may seek help from the instructor during the session either because they are stuck or because they need feedback before continuing to pin point misconceptions early. Within TUPLE, students are able to indicate when they need help. Like with past orchestration systems (Alavi & Dillenbourg, 2012), this feature makes the help seeking more visible to the instructor. Students are also able to indicate when they have reached a checkpoint where they require feedback. For the instructor, these types of requests can then be differentiated to allow them to better balance their time. Finally, as students do not always seek help when they need it, TUPLE will alert the instructor when a group has not made progress over a certain set of time so that support can be given.
The implementation of these features in TUPLE align with findings from needs analysis sessions with instructors, lab assistants, and students. A total of six instructors, three lab assistants, and eight students were interviewed. All three groups discussed a disconnect between the help seeking and the help giving process. Instructors expressed wanting to provide more support but struggled to know when to provide this help, especially if not asked by students. Students wanted to be able to work efficiently and expressed that learning when to ask for help was a skill with which they struggled. Furthermore, they wanted to get support from their instructors or lab assistants but found that there was often a disconnect between when they asked and were able to receive support. From these findings, we see that there is a need for more support in the help seeking and feedback process. These results will be used to continue to refine TUPLE and evaluate its effectiveness in university computer science labs.

Author

• Jenifer Olsen, University of San Diego