How do I define Population Distributions parameters

Question

Hello, I am trying to determine the characteristics of distributions within a dataset, I am hoping to try and use the number of data points (N) as a parameter like I can do when modeling in Prism, however I haven't seen a place where I can put this in. Does anyone have any suggestions? I would prefer an interactive explanation.

cea5320 · Answer

Thank you Thierry, I do have access to the full data set! I am trying to determine whether the modeled distributions reflect a characteristic of the population which I also have data for, so I guess I was trying to figure out if I could add in that data or how best to determine if the modeled distributions reflect that characteristic. Any help is much appreciated!

Carmen

Thierry_S · Answer

Hi,
At the best of my knowledge, JMP is designed to only work with actual first level data (all data points or for some platform data with frequency) rather than aggregated data (i.e. Mean, Stdev, and N). This provide a much better control of the type of analyses you can conduct (e.g. suitability of the model by assessing the distribution of the residuals). While PRISM is a popular software for creating quality data plots, the pre-canned statistical analyses may not be always appropriate especially when you use aggregated data (distribution of the data is lost when summarizing the data as Mean, Stdev, and N).
Let us know if you have access to the complete data: I or somebody else in this forum may be able to help.
Best,
TS