Uma análise sobre o uso de distratores em problemas de Parsons para o ensino de programação introdutória

Abstract

Parsons Problems are a type of programming exercise in which the student must order code snippets provided in a scrambled form. To evaluate the impact of distractor blocks inserted in Parsons problems on academic performance (grades) and task difficulty (number of block movements), this undergraduate thesis analyzed data produced by a quasi-experimental study (n = 249) that alternated Parsons exercises with and without distractors through an on- line judge. Statistical analysis reveals that distractors increase effort, requiring more movements (Rank-Biserial = 0.406 in sequential structures; 0.143 in conditional structures). Despite this greater practical demand, academic performance showed no statistical difference in sequential structures (Rank-Biserial = 0.105), with a small effect favoring the control group in conditional structures (0.139).