angelabkim
commented
8 years ago If you're referring to today's discussion in class, the p-value is a statistic that tells you how significant your results are. It gives you the probability of your result, assuming a null-hypothesis.
To give an example in psychology, let's say you want to test whether drinking coffee makes you perform better on tests.
Here are the two possibilities:
- Null hypothesis- Drinking coffee has no effect on test performance
- Alternative hypothesis- Drinking coffee improves test performance
As a researcher, you want to disprove the null hypothesis (i.e. no effect). You run your experiment, analyze the results, and get a p-value of .05. This says that you have a chance of .05 of mistakenly rejecting the null (i.e. thinking there is an effect when there really is none). In psychology, anything less than .05 is accepted as a "significant effect". The smaller the p-value, the better. So in this case, you can reject the null, and claim you found a significant effect of coffee drinking on improving test performance.
As an aside, there are a lot of controversies with using the p-value to "prove" an effect...
Dear peers I am in complete ignorance of p-value. "Is it the value of related Gausian function in that point?" I was wondering whether someone could help me with this desperate problem