For low prior proficiency students, unsolicited hint usage rates correlate negatively with posttest solution length, while hint frequency does not correlate¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r2 · q3 · i0–i3
Currency · newest study 2020 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i3 HJR and HNR are significantly negatively correlated with posttest solution length for low prior proficiency students, with a stronger correlation for HNR (r = -0.48) than HJR (r = -0.37). → Mehak Maniktala 2020 q3 i0 The number of unsolicited hints given does not correlate with posttest solution length, suggesting condition differences are not attributable to hint frequency. → Mehak Maniktala 2020 (2)

Evidence¶

Mehak Maniktala 2020¶

Mehak Maniktala, Christa Cody, Tiffany Barnes, Min Chi. (2020). Avoiding Help Avoidance: Using Interface Design Changes to Promote Unsolicited Hint Usage in an Intelligent Tutor. https://arxiv.org/abs/2009.13371

q3 · i3 · causal · r?

Correlational analysis of the same 100-student study (Low group N = 41), reported in Table 3: HJR r = -0.37 (p < 0.001) and HNR r = -0.48 (p < 0.001) against posttest solution length. "significantly and negatively correlated" is the article's own wording. The authors interpret this as low-knowledge students learning more from needed hints.

"However, both HJR (second row) and HNR (third row) are significantly and negatively correlated to posttest solution length for students with Low Prior Proficiency (HJR: p < 0.001, HNR: p < 0.001), with a stronger correlation to posttest solution length for HNR than HJR"

Mehak Maniktala 2020 (2)¶

Mehak Maniktala, Christa Cody, Tiffany Barnes, Min Chi. (2020). Avoiding Help Avoidance: Using Interface Design Changes to Promote Unsolicited Hint Usage in an Intelligent Tutor. https://arxiv.org/abs/2009.13371

q3 · i0 · causal · r2

Table 3 top row: correlations between unsolicited hints given and posttest solution length are non-significant for the entire population (r = -0.13, p = 0.18), Low (r = -0.10, p = 0.26) and High (r = -0.03, p = 0.83) groups; printed values are below r = .10 in magnitude for the subgroup analyses.

"the number of unsolicited hints given does not correlate to posttest solution length, suggesting that differences in posttest solution lengths between conditions cannot be attributed to the frequency of unsolicited hints."

Discussion¶

Related Claims¶

• Assertions significantly increase unsolicited hint usage (justification and needed rates) over Messages for students regardless of prior proficiency — related
• Aptitude-treatment interaction: Assertions led low prior proficiency students to form significantly shorter posttest proofs than Messages — related