Skip to main content

Table 5 P(Li) predicted for the testing sets using the three calibration schemes

From: A new adaptive testing algorithm for shortening health literacy assessments

Li

Leave-one-out

Scheme (a)

Scheme (b)

 

Actual

Predicted

Actual

Predicted

Actual

Predicted

Low

0.29

0.23

0.31

0.29

0.52

0.43

Moderate

0.46

0.52

0.45

0.47

0.43

0.52

High

0.25

0.25

0.24

0.24

0.05

0.05

  1. In scheme(a) the testing set only included subjects not used for calibration; in scheme (b) the testing set is the clinic sample.