Hi @KDan,
Very glad to know you're using our SEM platform! A couple things to clarify:
There is a by-group analysis option:
![LauraCS_0-1677075950841.png LauraCS_0-1677075950841.png](https://community.jmp.com/t5/image/serverpage/image-id/50327iDDB22BFF59A65A58/image-size/medium?v=v2&px=400)
And also a multiple-group analysis option:
![LauraCS_1-1677076043270.png LauraCS_1-1677076043270.png](https://community.jmp.com/t5/image/serverpage/image-id/50330i9D65CBD13CC0EA99/image-size/medium?v=v2&px=400)
My guess is you want to do multiple-group analysis (MGA). Is this correct? With MGA, you're able to use your full sample and test for group differences in parameters across groups, whereas the By-group option simply expedites the fitting of two separate models (one for each group), which effectively divides your sample size in two (or more depending on how many levels your grouping variable has).
The sample size error shows up when your sample size is smaller than the number of parameters you're trying to estimate (side note: you can get a description of the warning/error by clicking on it in the Status tab):
![LauraCS_2-1677076630647.png LauraCS_2-1677076630647.png](https://community.jmp.com/t5/image/serverpage/image-id/50338iB2CE3CDF294D2F7B/image-size/medium?v=v2&px=400)
Thus, the only way around this warning is (1) using a larger sample size or (2) reducing the number of estimates in your model. Are you able to specify a more parsimonious model?
Lastly, if you accidentally used the by-group option, that could lead to the sample size warning because your sample is being divided to achieve separate models for each of the levels of the grouping variable, so just switching to MGA might solve the problem.
HTH,
~Laura
Laura C-S