Testlets are groups of locally dependent items, such as the questions about the same topic.

 

If two or more items are highly dependent (so that their residuals are highly correlated in Table 23.99) , then it may be better combine them into one "testlet" item.

 

Testlet Response Theory is  technically complex and require special software, but we can approximate it by combining the dependent items into partial-credit items

 

The easiest way to do this is:

Export the RFILE= to Excel.

Use Excel to add together the dependent items into partial-credit items: ISGROUPS=0

Use the Winsteps Excel/RSSST menu to create a Winsteps control file from the Excel file.

Estimate the person measures

 

Cross-plot the person measures from the original analysis and the partial credit analysis. Are there any differences? Which analysis produces more meaningful results?