I have 20-30 repeated measurements each of which produce a single number. My goal is to determine whether the mean of these numbers is statistically significantly greater than the number 1. (significantly less than 1 does not count, in this case)

I know, for example, that if I want to know whether the means of two populations are different, I can use a t-test, but this does not seem to apply when one of my "populations" is just the fixed number 1. Can anyone help suggest the correct test?

  • 1
    $\begingroup$ If you're prepared to assume independence (which may be doubtful with repeated measures) and common variance, and that the population is (sufficiently close to) normally distributed, see en.wikipedia.org/wiki/Student%27s_t-test#One-sample_t-test $\endgroup$ – Glen_b -Reinstate Monica Oct 23 '18 at 23:23
  • $\begingroup$ Erika was first, so I accepted their answer, but yes -- an upper tailed one-sample t-test is exactly what I'm looking for. @Glen_b I actually can assume independence in this case. $\endgroup$ – Bunji Oct 24 '18 at 3:06

I would suggest using a One-Sample T-test. This site explains the procedure and also has a working example - http://www.biostathandbook.com/onesamplettest.html


This is known as a one-sample t-test, as mentioned by Erika.

You can use the calculator here (and the information below the calculator for reference): https://stattrek.com/online-calculator/t-distribution.aspx

Make sure you select sample, in place of t-score.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.