A researcher plans to conduct a study to compare two groups. Based on previous research, she anticipates a medium effect size. She does a power analysis with alpha level of. 05. This power analysis tells her.

Alpha level of. 05 tells her that the results are reliable and correct.

Explanation:

The alpha levels are commonly used in statistical studies to process a system of hypothesis, which must have a principal hypothesis, alternatives hypothesises and a null hypothesis.

In this order of ideas, in every quantitative research we must prove our principal hypothesis by processing the null one. If it results true the null hypothesis we have to support the alternate hypothesis, if it results false we don't support the alternative hypothesis and the principal one results true.

So, in this process researchers need to have some reliability in their results, and there's where the alpha levels come in. These alpha levels represents the percentage of probability to the null hypothesis to be rejected. As minimum is this number, more reliable is the result.

In this case, she used. 05 (5%) alpha level, that means the null hypothesis has 5% of probability to be false, it's a small chance, but if it results the null rejected, it means that the result is reliable. Otherwise if we use 40% of alpha level (for example), that means the result isn't reliable because there was a high chance to reject the null hypothesis in the beginning.

It's like if you are participating in a contest and you have 100% of probability to lose, that means your participation is not worth it, because in the beginning you had odds completely against you.