CUTHI= cut off responses with high probability of success = 0, no

Use this if careless responses are evident. CUTHI= cuts off the top left-hand corner of the Scalogram in Table 22.

Eliminates (cuts off) observations where examinee measure is CUTHI= logits or more (as user-rescaled by USCALE=) higher than item measure, so the examinee has a high probability of success. Removing off-target responses takes place after PROX has converged. After elimination, PROX is restarted, followed by JMLE estimation and fit calculation using only the reduced set of responses. This may mean that the original score-based ordering is changed.

The CUTHI= value to USCALE=. So, if USCALE=10, then CUTHI=1 means "omit responses where the person ability is 1 user-scaled unit or more higher than the item difficulty."

Usually with CUTLO= and CUTHI=, misfitting items aren't deleted - but miskeys etc. must be corrected first. Setting CUTLO= and CUTHI= is a compromise between fit and missing data. If you loose too much data, then increase the values. If there is still considerable misfit or skewing of equating, then decrease the values.

Here are the usual effects of CUTLO= and CUTHI=

1. Fit to the Rasch model improves.

2. The count of observations for each person and item decreases.

3. The variance in the data explained by the measures decreases.

Example 1:        Eliminate responses where examinee measure is 3 or more logits higher than item measure, to eliminate their worst careless wrong responses:

       CUTHI= 3

               This produces a scalogram with eliminated responses blanked out:

               RESPONSES SORTED BY MEASURE:

               KID    TAP

                             111111111

                    123745698013245678

               ------------------------                      

                15  111       11100000 observations for extreme scores remain

                14  111     1110000000

                28  111   111010000000

                30  1111 1111000000000

                27  111111100000000000

Example 2. At www.rasch.org/rmt/rmt62a.htm the instigator of this approach suggests CUTLO= -1 and CUTHI=2