Menu ENN Search
Language: English Français

What is the meaning of Z-test and P value in SQUEAC estimate calculator?

This question was posted the Coverage assessment forum area and has 1 replies. You can also reply via email – be sure to leave the subject unchanged.

» Post a reply

Anonymous 1310

Nutrition Manager

Normal user

6 Oct 2014, 11:06

In a SQUEAC survey, i found

Point coverage: 72.9% (CI, 62.0%-81.4%).
Z: 0.24
p: 0.8093

So, here what is the meaning of this Z and p value?
how should i explain this results?

Thanks

Mark Myatt

Consultant Epideomiologist

Frequent user

8 Oct 2014, 10:11

The estimate returned by the calculator (a Bayesian beta-binomial conjugate analysis) requires that there is no conflict between the prior and the likelihood. When there is a prior-likelihood conflict then the estimate may be biased towards the prior and should not be used.

One way of assessing the presence of the a prior-likelihood conflict is to examine the plot of the prior, likelihood, and posterior. If the prior and the likelihood overlap considerably then there is no prior likelihood conflict.

Another way of assessing the presence of the a prior-likelihood conflict is to use a formal test (in the BayesSQUEAC calculate the z-test is used). The null hypothesis is that the prior and likelihood do not conflict. If the p-value is small (i.e. p < 0.05) we would reject the null hypothesis of no conflict (i.e. there is a conflict). In your example p = 0.8093. This means that there is no evidence to support the existence of a prior-likelihood conflict and it is safe to use the estimate.

If you examine plot of the prior, likelihood, and posterior for the survey that prompted your question you should see that there is a large overlap between your prior and likelihood.

I hope this helps.

BTW : Congratulations on achieving a high coverage.

Back to top

» Post a reply