CODEWITHSUNDEEP
statisticsconfidence-intervalp-valueab-testingsignificance
xcentralx
xcentralx

Reputation: 25

Which p-value thresholds for a two-sided 90% confidence interval?

I'm running an A/B test, and I want to check for statistical significance with 90% confidence two-sided. I've calculated standard errors, z-scores and p-values.

I'm saying that I have significance when my p-value is lower than 0.1 and greater than 0.9. Am I right? I'm using this tool https://vwo.com/blog/ab-testing-significance-calculator-spreadsheet-in-excel/

I'm doubting whether it should be lower than 0.05 and greater than 0.95.

I think I'm mixing things up in my head. Because, I have the p-value, and I'm saying that my alpha is 0.1. I'm not calculating the alpha/2 and p-value/2, nor I need it. So, should I just check if the p-value is lower than 0.1 and that's all? Not even greater than 0.9?

Upvotes: 0

Views: 1882

Answers (2)

Sara Kuntumalla
Sara Kuntumalla

Reputation: 1

If you have 90% confidence level and your test is 2-tailed, the p-values should be less than 0.05 or greater than 0.95 to reject the null hypothesis and conclude that there is insufficient evidence to support the alternate hypothesis.

Upvotes: 0

Deepak Tatyaji Ahire
Deepak Tatyaji Ahire

Reputation: 5309

90% confidence interval means you have 10% as the Region of Rejection, which is divided into 2 equal halves on both sides in the tail area.

You are correct! Just check if the p-value is less than 5% or 0.05 and is greater than 95% or 0.95 as this range represents the Region of Rejection

If the p-value lies in the above range, you can reject the Null Hypothesis.

Upvotes: 0

Related Questions