Guidelines on Research
Updated on 15 December, 2021
REPORTING OF P-VALUES
REPORTING OF P-VALUES
This is a little explanation on how to report p-values when you do significance testing or test hypotheses. Very often, in dissertations and theses, you would test hypotheses by way of several statistical tests or techniques, the most common ones being the independent samples t-test, one-way analysis of variance (ANOVA), chi-squared test of independence, Mann-Whitney U test and Kruskal-Wallis H test, by using statistical software (SPSS, SAS, STATA, R, etc.). Hypotheses are also tested by conducting multiple regression, logistic regression and other even more advanced techniques like MANOVA, ANCOVA, MANCOVA, etc. What you need to know is that most software use a default significance testing level of 5%, which may be adjusted according to your preferences.
Definition of p-value
Essentially, a p-value indicates the probability that the significant result obtained when using your data occurred by chance. Thus, the smaller the p-value, the higher the probability that the result was obtained, based on statistical evidence. There is a traditional way of reporting the significance of a result in the form of p < 0.01, p < 0.05, p < 0.1, etc., as 1%, 5% and 10% are historical values that have been used at the time when "only limited tables of critical values were available" (American Psychological Association, 2010, p.114). For example, if you obtain a p-value of 0.032, you would say that the result is significant at the 5% level, since 0.05 is the closest traditional value that exceeds 0.032. Along the same line, if p = 0.001, you would say that the result is significant at the 1% level.
NOTE
When obtaining p-values statistical software, you should report the lowest significance level possible. For example, if p = 0.004, even though the result is significant at the 5% level (< 0.05), the default significant testing level in the software, good reporting practices recommend that it should be reported as significant at the 1% level.(< 0.01).
Let me explain why in layman's language. If you were to obtain 92 out of 100 in your exam, what would you tell your parents? You simply passed or you would jump with joy and announce to them that you got a distinction? Similarly, if you get a p-value of, say, 0.004, you would "proudly" report that your result is significant at the 1% level, i.e., that there is a probability of at least 99% that the result is significant, based on statistical evidence. That was the point of this entire discussion.