Type I and type II errors
An alpha value measures the amount of risk in interpreting results of an experiment and is equal to the risk of obtaining false positive results called Type I errors. False negative results, called Type II errors, are measured by a beta value. In order to reduce the chances of obtaining a false negative or a false positive result, the sample size must be increased. A power value measures the likelihood that statistically non-significant results are actually non-significant, rather than a Type II error. Effect size is used in determining the significance of results. In order to reduce chances of obtaining a false negative or false positive result, or to detect a smaller effect or accommodate for larger standard deviations, a larger sample size must be evaluated.