My problem is that I want to run Fit Life by X. Relationship: Location and scale on a Lognormal distribution using a nested design. I need to do a test for difference in location and scale. Normally this works fine for 5 to 6 times and then JMP starts up using the Arrhenius Celsius relationship instead of Location and scale and then I don’t get the Nested models box. When this happens JMP can not test Location and scale because even if I change over to location and scale. JMP will not change and do the calculations. Anyone seen this? Best regards/Lasse .