1-The lower the level of significance, the harder it is to reject the null. In sum, you want to be REALLY sure that your alternative hypothesis is correct, so if you use a low significance level (say, .01), then you would only reject the null if you're REALLY sure about the outcome.
Most use a .05 level of significance, which means that you have a bad sample 5 out of every 100 samples. This is not necessarily bad.
You can see that if you use a .01 level of significance, you are saying that you would only have an error sample 1 every 100.
So, it would be much harder to reject the null when you use the .01 level. You also have more CONFIDENCE in your answer if it is .01.
2-Whenever I see an alpha of .10, I do start to question why they would use that high of an alpha. It does lead one to believe that either there are issues with the information, or the individual is trying is best they can to ensure that the null hypothesis is rejected.
Of course, it could work the other way too. If you use .01 when it's not necessary, you may find that you can't reject the null except in few exceptions. This would work for some individuals who do not believe in the alternative but feel they must perform the analysis.
What are your thoughts here?
The relationship between levels of significance and rejecting the null is analyzed.