Combining statistics from two distributions

Click For Summary

Discussion Overview

The discussion revolves around the possibility of combining statistics from two distributions representing the same parameter from different populations. Participants explore how to derive a single combined upper tolerance limit (UTL) for a hypothetical combined population based on summary statistics, without access to the original data.

Discussion Character

  • Technical explanation, Mathematical reasoning, Debate/contested

Main Points Raised

  • One participant questions the definition of the upper tolerance limit (UTL) and suggests that if data about the distribution is available, such as mean and standard deviation, the two distributions can be combined.
  • Another participant clarifies that the summary statistics indicate a normal distribution and specifies that the UTL is the 95% upper tolerance limit, noting the absence of standard deviation.
  • A different participant proposes that the standard deviation can be approximated as half the distance between the mean and the UTL, and suggests using a weighted average to combine the values, with weights based on inverse variances.
  • A final participant expresses gratitude and indicates intent to follow up on the discussion.

Areas of Agreement / Disagreement

The discussion includes varying perspectives on how to combine the statistics, with no consensus reached on a definitive method or approach.

Contextual Notes

Participants rely on summary statistics and assumptions about the distributions, but there are limitations due to the lack of original data and standard deviation, which may affect the validity of the proposed methods.

mclovin
Messages
3
Reaction score
0
Is it possible to combine statistics from two distributions for the same parameter. For example I have one distribution for X from population A and a second distribution for X from population B. I want to assume all data is from the same population. I have calculated UTLs(Upper tolerance limits) for each population but need a single, combined utl for a combined population. The catch is I don't have the original data, only the calculated statistics.
 
Physics news on Phys.org
How is that upper tolerance limit defined?
If you have some data about the distribution of the expected parameter (like mean+standard deviation and the assumption of a Gaussian distribution), you can combine both.
 
According to the summary stats it is a normal distribution and the UTL is the 95% Upper tolrance Limit. i have the range of values and the artithmetic mean. No std dev.
 
In that case, the standard deviation is just ~1/2 of the distance between mean and UTL. You can combine both values with a weighted average (the formula should be somewhere at wikipedia or on other pages. The weights are just the inverse variances.).
 
Thank you. I will follow up on this.
 

Similar threads

  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 30 ·
2
Replies
30
Views
5K
  • · Replies 4 ·
Replies
4
Views
1K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 9 ·
Replies
9
Views
3K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 4 ·
Replies
4
Views
2K