11.7 An example: comparing true and estimated parameters

Huebner & Wang (2011) considered several criteria for assessing the accuracy of different classification methods, including the proportion of correctly classified attributes (PCA) and the proportion of correctly classified attribute vectors (PCV).

To find the PCA and PCV, you can use the following code:

Code

K <- ncol(Q15)
ClassRate(EAP, true.attributes)

## $PCA
## [1] 0.622
## 
## $PCV
## [1] 0.98 0.89 0.68 0.40 0.16

Code

ClassRate(MLE[, 1:K], true.attributes)

## $PCA
## [1] 0.638
## 
## $PCV
## [1] 0.99 0.90 0.71 0.40 0.19

Code

ClassRate(MAP[, 1:K], true.attributes)

## $PCA
## [1] 0.63
## 
## $PCV
## [1] 0.97 0.88 0.67 0.46 0.17

References

Huebner, A., & Wang, C. (2011). A note on comparing examinee classification methods for cognitive diagnosis models. Educational and Psychological Measurement, 71(2), 407–419.