Non-Linear Relationships
Whole Model Test
This portion of the output compares the performance of our specified model with a naive default model with no predictor variables. The naive model essentially posits that, based on the sample ratio of 8/32 or 25% of the patients having no disease, there is a probability of 0.75 that a tested patient has PD. In this panel, the Chi-Square test is analogous to the customary F test and in this analysis, we find a significant result. In other words, this model is an improvement over simply estimating a 0.75 probability of PD.
We also find RSquare (U), which is analogous to R2 as a goodness-of-fit statistic. This model provides an unimpressive value of 0.34, indicating a lot of unexplained variation. Such ...
Get Practical Data Analysis with JMP® now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
