Item discriminating value as coefficient of variation decreases monotonically as number of correct answers increases, giving a negative difficulty–discrimination relationship¶

Claim · All claims
Evidence · 1 study · theoretical r3 · q2
Currency · newest study 2021 · not yet checked against its sources · not reviewed by a person · overdue for a recheck since 2024: tool and model accuracy changes quickly

Subclaims¶

q2 i? Discriminating value of an item equals the coefficient of variation of item scores, is bounded 0 ≤ Disci < 1, and decreases monotonically with increase in k, so its relationship with item difficulty value is negative. → Chakrabartty 2021 q2 i? Analytically, the square of item discriminating value equals (1 − difficulty value) divided by the number of correct responses, implying a negative Diffi–Disci relationship. → Chakrabartty 2021

Evidence¶

Chakrabartty 2021¶

Chakrabartty, S. N. (2021). Assessment of item and test parameters: Cosine similarity approach. International Journal of Psychology and Educational Studies, 8(3), 28-38. https://dx.doi.org/10.52380/ijpes.2021.8.3.190

q2 · i? · theoretical · r3

Empirical verification on the 911-candidate, 50-item Selection Test data. The correlation between item difficulty values and item discriminating values "was (-) 0.579586", consistent with the derived negative relationship in equation (1.10).

"The graph of item difficulty values and item discriminating values suggest that the 𝑟𝐷𝑖𝑓𝑓𝑖𝐷𝑖𝑠𝑐𝑖be negative. Value of 𝑟𝐷𝑖𝑓𝑓𝑖𝐷𝑖𝑠𝑐𝑖was (-) 0.579586."

Discussion¶

Related Claims¶

• Test discriminating value equals the coefficient of variation of test scores (tan∅ = SD/mean) — related
• Point-biserial item-total correlation is negatively related to test discriminating value and can be expressed via item and test difficulty and discrimination parameters — related
• Intersection point k0 of item difficulty and discriminating curves provides a data-driven item-deletion criterion — related
• Test difficulty value defined as ratio of observed to maximum score vector length times cosine of their angle equals test mean divided by number of items — related
• Accuracy and the proportion of high bets decrease with increasing discrimination difficulty in both domains — related