Designing salient and supportive systems for proximal behaviors.

Designers should create salient reminders and support mechanisms to help users stick to near-term plans by reducing participation costs and increasing belief in the ease of tasks.

About this paper

The author conducted two studies to understand how temporal distance affects planned behavior, finding that attitudes become more important for distant events while perceived behavior control influences intentions regardless of timing.

These findings advance the Theory of Planned Behavior and provide strategies for designers and event organizers to motivate behaviors over different timeframes.

Here are some methods used in this study:

Theory Of Planned Behavior Construal Level Theory

Which part of the paper did the design guideline come from?

“We hypothesized that people tend to have a higher intention to perform the behavior in the far future compared to near future (H5). Results of the paired-samples t-test show that the mean of willingness to attend the yoga class differs a month before the event (M=.80, SD=.41) and a few days before the event (M=.60, SD=.49) at the .01 level of significance (t=2.70, df=29, p<.01, 95% CI, for a mean difference .05 to .35, r=.62). We should point out that in the end, only 6 participants actually (...)” (‘Change in Intention Over Time’ section)

Suh, M. (Mia), & Hsieh, G. (2016). Designing for Future Behaviors. Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems.

Inspiration and scope

In this paper, the authors focused on characteristics of people looking to participate in events or behaviors related to encouraging participation in proximal behaviors and adherence to planned behaviors.

You are designing for university teaching assistants, aiming to assess performance and provide real-time feedback. The contexts differ: the paper targets individuals motivated by interest, social influence, or rewards, while you focus on assistants motivated by professional development, job performance, and educational outcomes. At the same time, both contexts benefit from incorporating effective feedback mechanisms. Feedback in the paper's context drives participation and behavior adherence. In your context, real-time feedback helps teaching assistants improve performance. Timely, relevant, and constructive feedback enhances the system's overall effectiveness.

Also, the paper's design covers a broader scope with less granular, one-time interactions for immediate actions, while your design involves a narrower scope with ongoing interactions for continuous teaching assistant improvement. At the same time, both designs must adapt to users' evolving needs. In the paper's context, adaptability maintains engagement by addressing changing motivations and interests over time. Similarly, in your context, the system must adapt to teaching assistants' stages and educational environments to remain effective.

Leveraging these similarities, design a system that provides timely, relevant, and constructive real-time feedback to teaching assistants, reducing uncertainty and enhancing their confidence and capabilities to support continuous improvement.

Your input

What: I'm designing a LLM-based system for assessing teaching assistant's performance and providing real-time feedback.
Who: Teaching assistants in universities
Design stage: Research, Ideation, Evaluation

Understanding users

The following user needs and pain points may apply to your design target as well:

Effective Feedback Mechanisms

Providing timely, relevant, and constructive feedback to teaching assistants in real-time can significantly improve their performance. Such a mechanism helps in identifying strengths and areas needing improvement instantaneously, fostering a more adaptive learning environment.

Adaptability to Evolving Needs

Designing a system that adapitates to the varying developmental stages of teaching assistants and the different educational environments they work in is crucial. This ensures that the system remains effective and relevant, providing personalized feedback and support tailored to the user's current context and needs.

Design ideas

Consider the following components for your design:

Implement a dynamic feedback mechanism that adjusts the type and tone of the feedback based on the teaching assistant’s current development stage and teaching context.

Incorporate user-friendly visual dashboards that clearly show progress and areas for improvement while celebrating small wins to boost self-efficacy.

Provide exemplary behaviors and concrete, actionable steps within feedback to guide teaching assistants towards best practices progressively.

Methods for you

Consider the following method(s) used in this paper for your design work:

Theory of Planned Behavior (TPB)

Using TPB can help in understanding and predicting target users' behavioral intentions by assessing their attitudes, subjective norms, and perceived behavioral control. Designers should focus on identifying and targeting these three components specific to teaching assistants when planning interventions.

Within-Subjects Field Experiment

Conducting a within-subjects field experiment can help in examining how teaching assistants' intentions and attitudes change over a period and across different contexts. Designers should ensure repeated measurements and consistent conditions to truly understand changes.

Metrics for you

Consider the following metric(s) used in this paper to evaluate your design work:

Behavioral Intention

Behavioral intention (BI) measures how hard users are willing to try and how much effort they plan to exert toward performing a behavior. It can help understand the role of users' effort and motivation to adopt system recommendations. Consider user engagement and context to maintain user interest.

Perceived Behavioral Control

Perceived behavioral control (PBC) assesses the ease or difficulty users perceive in performing a behavior. It can highlight obstacles or facilitators to intended behaviors. Designers should consider users' self-efficacy and external factors when evaluating the feasibility of the system's feedback.

[Figure 1] From this figure, you can understand the theoretical framework that might provide insight into the underlying factors influencing behaviors, which could be applicable when designing an LLM-based system for performance assessment and feedback.