topic Re: Linearity and Bias study in Discussions

Linearity and Bias study

Scorpio — Sat, 10 Jun 2023 20:54:38 GMT

Hello Community,

While performing the Linearity and the bias study, if the P value for the Bias is less than 0.05--then that marks the presence of bias in the scale. Further, the average bias value is less than the resolution of my scale and secondly, the bias is not in increasing or decreasing fashion but, randomly fluctuated at different standard reference weights.

Can it be concluded that the scale has no bias(as it will be not reflected on the scale) or is there other way to interpret this results?

Re: Linearity and Bias study

statman — Mon, 03 Apr 2023 15:43:17 GMT

Sorry, it is impossible for me to comment without context.

Re: Linearity and Bias study

AcceptanceGnu69 — Mon, 03 Apr 2023 17:03:07 GMT

That clearly didn’t stop you from commenting anyways….

How about a more helpful response: Could you please add some more context so that we can try to help?

Re: Linearity and Bias study

P_Bartell — Mon, 03 Apr 2023 18:53:04 GMT

Can you share at a minimum the associated graphics that JMP produced? In general p-values are the last thing I look at...and the first thing I look at are the plots. They should tell you the story more than a p-value.

Re: Linearity and Bias study

statman — Tue, 04 Apr 2023 10:18:34 GMT

I apologize, I could have asked for more context in a more friendly way. Without knowing what you are measuring, how the study was conducted, what comparisons were being made that resulted in a p-value, what the sources of variation are in the study, what is the source of the potential bias, etc. Also, if you could, perhaps, attach a JMP data table or a coded data table if the content is not for public consumption that would be helpful as well.