Predicting the Persuasiveness of Influence Strategies From Student Online Learning Behaviour Using Machine Learning...

Area of Science:

• Educational Technology
• Human-Computer Interaction
• Machine Learning

Background:

• Limited understanding exists regarding the impact of persuasive influence strategies on student behavior within online educational systems.
• Current methods for assessing system persuasiveness rely on static, subjective measures like questionnaires, hindering real-time adaptation.
• Automated, real-time prediction of system persuasiveness is crucial for dynamic personalization of online learning environments.

Purpose of the Study:

• To investigate the relationship between the persuasiveness of influence strategies and student behavior in an online educational setting.
• To determine if machine learning models can effectively predict the impact of persuasive strategies on student engagement.
• To explore the feasibility of using student learning data for real-time persuasiveness assessment.

Main Methods:

• Implementation and testing of Machine Learning (ML) classification models.
• Utilizing student learning session data as input for the ML models.
• Analyzing the predictive power of ML models on the persuasiveness of different influence strategies.

Main Results:

• Student learning data can be successfully used to predict the persuasiveness of various influence strategies.
• Machine learning classification models demonstrated a significant impact of persuasiveness on student usage patterns.
• The study confirmed the potential for automated prediction of system persuasiveness.

Conclusions:

• Machine learning classifiers, when trained on learning session data, can automatically predict the persuasiveness of influence strategies.
• Online educational systems can leverage these ML models to dynamically adapt persuasive tactics.
• This adaptive capability holds significant potential for improving student engagement and overall learning in digital environments.