Predicting Latent Proficiency From Student Responses
The Wright maps (Figures 7–9) show that the logit range of any given EBRAS threshold across all the items of a construct is less than the logit range of the accuracy items. In many cases, the threshold range is very narrow, on the order of 1.0 logit, compared to the 5.5 logit range for the accuracy items. This is evidence that the latent variables measured by the EBRAS proficiencies have a stronger effect on item responses than the latent variable measured by the accuracy items, resulting in the same response category being observed across most of the items. In contrast, the wide range of difficulty of the accuracy items indicates that the items have a stronger effect on responses, resulting in responses that cannot be predicted without knowing the details of the item context. A practical upshot of this is that the EBRAS can be employed in formative assessment contexts more confidently, in which the number of items is small. Given the stability of the EBRAS responses, teachers can better rely on student responses to reflect latent student proficiency.
Predicting Latent Proficiency From Student ResponsesThe Wright maps (Figures 7–9) show that the logit range of any given EBRAS threshold across all the items of a construct is less than the logit range of the accuracy items. In many cases, the threshold range is very narrow, on the order of 1.0 logit, compared to the 5.5 logit range for the accuracy items. This is evidence that the latent variables measured by the EBRAS proficiencies have a stronger effect on item responses than the latent variable measured by the accuracy items, resulting in the same response category being observed across most of the items. In contrast, the wide range of difficulty of the accuracy items indicates that the items have a stronger effect on responses, resulting in responses that cannot be predicted without knowing the details of the item context. A practical upshot of this is that the EBRAS can be employed in formative assessment contexts more confidently, in which the number of items is small. Given the stability of the EBRAS responses, teachers can better rely on student responses to reflect latent student proficiency.
การแปล กรุณารอสักครู่..
