Question

Write a research question (that is, a research project) that you are interested in conducting. Write the main objective of the research project together with a null and alternative hypothesis.

Suggest a probabilistic and a non-probabilistic sampling method for your study population.

List your alpha significance level, and suggest at least 2 ways to lower the chances of making type 1 error.

Answer 1

One research question can be: Whether women and men differ in logical thinking abilities. Fistly, the response can be taken on the basis of an IQ test.

Thus, the main objective of the research is to find out whether women and men have indentical logical thinking abilities or not.

A probabilistic sampling method: We can randomly choose n many women and n many men. Ideally this choice should be from the set of 'all' mean and women. However, choosing from a smaller set would also suffice (for example men and women in a district, etc.)

A non-probabilistic sampling method: I will choose my n closest female aqquintants and n closest male aqquintants. (This might be biased and the results may not generalize to the population.)

I would go with the standard choice alpha = 0.05

Ways to lower type I error:

1. Lower the level of significance. This lowers down the probability of making type I error.

2. Increase sample size. This would reduce the variability in the statistic and hence the probability of falling into the rejection region, when the null is actually true would also go down.