Zihan Wu
University of Michigan, Ann Arbor, Michigan, United States
This paper investigates using micro Parsons problems as a novel practice approach for learning Structured Query Language (SQL). In micro Parsons problems learners arrange predefined code fragments to form a SQL statement instead of typing the code. SQL is a standard language for working with relational databases. Targeting beginner-level SQL statements, we evaluated the efficacy of micro Parsons problems with block-based feedback and execution-based feedback compared to traditional text-entry problems. To delve into learners' experiences and preferences for the three problem types, we conducted a within-subjects think-aloud study with 12 participants. We found that learners reported very different preferences. Factors they considered included perceived learning, task authenticity, and prior knowledge. Next, we conducted two between-subjects classroom studies to evaluate the effectiveness of micro Parsons problems with different feedback types versus text-entry problems for SQL practice. We found that learners who practiced by solving Parsons problems with block-based feedback had a significantly higher learning gain than those who practiced with traditional text-entry problems.
doi.org/10.1145/3613904.3641910
The ACM CHI Conference on Human Factors in Computing Systems (https://chi2024.acm.org/)