Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

Fit statistic cut off rule of thmb

I was hoping that i might get advice on what to have as a cut off point to remove bad respondents from my data.  I'm trying to anaIyse some max-diff data (HB, Logit etc)  would use this in conjunction with time taken to fill out the survey.
asked Mar 28, 2018 by Gnd

1 Answer

0 votes
Hi Gnd,

The magnitude of the fit statistic depends on how many items were shown in each set.  If four items are shown in each set, then any random set of scores should be able to predict respondents' answers correctly with 25% likelihood (fit statistic=.25).  If two items are shown per set (paired comparison approach), then the expected likelihood given random data is 50% (fit statistic=.5).  Thus we can generalize that the fit statistic should be at a minimum 1/c, where c is the number of items displayed per set.


## copied from help manual
answered Mar 28, 2018 by Saurabh Aggarwal Gold (31,425 points)