In the case of a chi-square test for one sample proportion (test-of-proportions), you would write this:

1prefsAB: \chi_{(df=1, N=60)}^2 = 17.07, p < .0001

where df represents the degrees of freedom, N the size of the test. After the equal we have the test statistics, and than the p-value. No one cares about the exact value of p-value, only if it is:

• < .5
• < .01
• < .001
• < .0001

For p-values between .05 and .1, these values are given the name of trend, or marginal results. However they are not stat. sig.

For p-values > .05 (non stat. sig.), we don’t report the exact p-value. We just report n.s (non-significant).

N.B. It’s important not to interpret p-values as representing the effect strength of a test. There are other statistics for that (effect size statistics).

Sig. results tell us there is a statistically significant difference between things we are comparing. A non sig. one does NOT tell use that there is no difference, but there is no detectable difference in the data we are analysing.

References

• https://www.coursera.org/learn/designexperiments
• test-of-proportions

Next -> exact-asymptotic-and-binomial-tests