The Effects of Starter Template Scaffoldings on Coding Anxiety in Introductory Programming Courses

Programming anxiety is a well-documented challenge in introductory computing courses, particularly among students with no prior exposure to programming. Scaffolding strategies such as pair programming, Parsons puzzles, and starter code templates are widely used to reduce anxiety and build confidence during the early stages of learning. Although starter templates often include explanatory comments and partially completed code, the relative effectiveness of these different scaffold types in mitigating anxiety has not been systematically examined. This paper reports findings from a controlled experiment conducted during the first laboratory session of an undergraduate C programming course for students from different engineering and science disciplines. Pairs of students were randomly assigned to one of three conditions: a control group using blank templates, a comments-based scaffold group, and a code-based scaffold group. Coding anxiety was measured using a questionnaire with subscales related to self-efficacy, partner interaction, and peer judgement. Contrary to our expectations, the control group showed a greater reduction in coding anxiety than either scaffolded group, and students in the comments-based condition reported larger decreases in anxiety than those receiving code-based scaffolds. These findings suggest that for beginners, even well-intentioned scaffolds can temporarily increase the perceived difficulty in task completion.