Skip to main content
. 2017 May 11;41(7):561–576. doi: 10.1177/0146621617707509

Table 3.

The Measurement Accuracy and Test Length for the DINA Model (Variable Length).

Stopping criteria Item selection Attribute
Pattern Test length
1 2 3 4 5 6 M SD
0.7 PWKL 0.942 0.941 0.961 0.965 0.963 0.948 0.774 8.04 1.804
SHTVOR 0.951 0.951 0.948 0.954 0.953 0.952 0.765 11.76 3.548
DBS 0.957 0.961 0.951 0.953 0.945 0.954 0.776 12.39 3.549
0.8 PWKL 0.963 0.971 0.973 0.979 0.967 0.971 0.843 9.52 2.297
SHTVOR 0.971 0.975 0.961 0.971 0.975 0.971 0.852 13.96 4.150
DBS 0.976 0.973 0.968 0.975 0.970 0.971 0.857 14.45 3.998
0.9 PWKL 0.984 0.988 0.993 0.993 0.984 0.983 0.929 11.72 2.863
SHTVOR 0.990 0.986 0.989 0.986 0.982 0.984 0.921 17.88 5.434
DBS 0.987 0.986 0.990 0.987 0.985 0.982 0.920 17.50 4.816

Note. DINA = deterministic input, noisy, and gate; PWKL = posterior-weighted Kullback–Leibler method; SHTVOR = Sympson–Hetter method, which comprises test overlap control, variable length, online update, and restricted maximum information; DBS = dynamic binary searching.