Chapter 7: Sample Size, Effect Size, and Power

Chapter 7: Sample Size, Effect Size, and Power New Vocabulary: Alpha: the chance of making a type 1 error Beta: the chance of making a type 2 error Effect size: the extent to which a difference or relationship exists between variables in a population Power: the ability to find a difference or association one one actually exists Type 1 error: an error made when the researcher incorrectly rejects the null hypothesis. They think there is a difference when there really is not (false positive) Type 2 error: an error made when the researcher fails to reject the null hypothesis when it's false. They think there is no difference when there actually is. (false negative) Effect Size -​ Effect size is the difference between group means that exist within the population -​ Effect size and sample size are inversely related: as one increases, the other decreases -​ The sample size directly relates to the power of the study or the ability to find a difference when one actually exists -​ When sample size is increased, the power of the study is also increased, and the likelihood of rejecting the null hypothesis correctly is increased. -​ Power is defined as the likelihood of rejecting the null hypothesis correctly, meaning that there is a relationship and you are correct in your alternative hypothesis Type 2 Error -​ A type 2 error is made when you fail to reject the null incorrectly, thus missing an association that is really there -​ These errors usually occur because the sample size was not large enough, and the study did not have enough power to find a difference that really existed -​ The risk of a type 2 error is depicted by beta (b). Sample Size -​ The sample size you select should give you adequate power to reject the null hypothesis correctly. -​ Power analysis is how sample sizes are calculated -​ The sample size for your study depends on: the effect size, your alpha, and your power. -​ When your sample size is too small you have a greater chance of a type 2 error -​ When your sample size is too large you can waste time/money and have a greater chance of a type 1 error