Registered learners modify the teacher-provided base code more than anonymous learners, both in share of sessions and in Levenshtein distance¶

Claim · All claims
Evidence · 1 study (2 entries) · associational r1 · q2
Currency · newest study 2020 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? 62.5% of registered learners' sessions involved code changes versus 41.34% of anonymous learners' sessions. → Gallego-Romero 2020 q2 i? Registered learners' final code was less similar to the base code (Levenshtein distance 95.53 vs 98.52; t=7.74, p<0.0001), meaning more changes. → Gallego-Romero 2020 (2)

Evidence¶

Gallego-Romero 2020¶

Gallego-Romero, J.M., Alario-Hoyos, C., Estévez-Ayres, I. et al. (2020). Analyzing learners' engagement and behavior in MOOCs on programming with the Codeboard IDE. Education Tech Research Dev 68, 2505–2528 (2020). https://doi.org/10.1007/s11423-020-09773-6

q2 · i? · associational · r1

Descriptive comparison of sessions filtered to those where learners modified the base code, from the same Codeboard log analysis in three Java MOOCs. The article reports 41.34% of anonymous sessions and 62.5% of registered sessions involved code changes.

"An interesting fact is the percentage of sessions where learners did changes in the code compared to the total number of sessions: 41.34% of the sessions for anonymous learners, and 62.5% of the sessions for registered learners."

Gallego-Romero 2020 (2)¶

Gallego-Romero, J.M., Alario-Hoyos, C., Estévez-Ayres, I. et al. (2020). Analyzing learners' engagement and behavior in MOOCs on programming with the Codeboard IDE. Education Tech Research Dev 68, 2505–2528 (2020). https://doi.org/10.1007/s11423-020-09773-6

q2 · i? · associational · r1

t-test over coding files (1,069 anonymous, 298 registered) comparing learner code with teacher base code using Levenshtein similarity. The article reports means of 98.52 vs 95.53, t=7.74, p<0.0001 (Table 5).

"Table 5 shows the Levenshtein distance in the case of anonymous learners and registered learners, obtaining a lower value on average in the case of registered learners, which means that registered learners did more changes in the code on average."

Discussion¶

Related Claims¶

• When comparing only sessions in which the base code was actually modified, there are no significant differences between anonymous and registered learners in actions or time per session — related
• Registered learners have a higher ratio of compilation errors to total compilations, consistent with deeper code modification — a narrower finding that bears on this claim
• Registered learners perform more actions and spend more time per Codeboard working session than anonymous learners — related
• Registered learners in programming MOOCs using Codeboard show greater engagement than anonymous learners, compiling and executing code more times per access — related