There are research situations in which making an alpha error is worse than

# There are research situations in which making an

• 43
• 100% (1) 1 out of 1 people found this document helpful

This preview shows page 38 - 40 out of 43 pages.

There are research situations in which making an alpha error is worse than making a beta error. For example, in a situation in which there has already been considerable exploratory research, you may want to confirm that some alleged effect really exists. If the area is controversial, or if you believe the health of others or your professional reputation is at stake, then you want to be very confident that some effect really exists before you publicly claim it does. Under these circumstances, set alpha low, to .01 or less. This will minimize the risk of an alpha error, but at the expense of increasing beta and decreasing power. SUMMARY OF KEY CONCEPTS FOR SECTION 1-B of WEEK 4 1 Hypothesis testing is a process used to decide whether or not a sample result is due to sampling error. A simple application is to decide whether a sample with a certain mean occurs by chance from a specified population. You begin by assuming your sample result did occur by chance from the general population, and then you calculate the likelihood of getting the sample mean you obtained. If this turns out to be so unlikely as to contradict the initial assumption, you can conclude that your initial assumption is wrong and that the sample comes from a population with a different mean than that of the general population. 2 The four steps of the formal hypothesis-testing process are: a. Determine H 0 and H A . H 0 generally implies your results are due to sampling error. b. Set a value for alpha. Usually this is .05 or .01. c. Assuming H 0 is true, calculate the probability of getting your results by chance. d. Either reject or retain H o . If the probability of your results is less than alpha, reject H 0 and conclude your results are not due to sampling error. If the probability of your results is greater than or equal to alpha, you have not gathered sufficient contradictory evidence against H 0 , so you should retain H 0 . 3 Probability is a value between 0 and 1 and is abbreviated with a lowercase p . that gives a measure of likelihood. The probability of a particular event is defined as the number of outcomes in which the particular event occurs divided by the total number of possible outcomes.
Page 39 ࠵?࠵?࠵?࠵?࠵?࠵? ࠵?࠵? ࠵?࠵?࠵?࠵?࠵?࠵?࠵?࠵? ࠵?࠵?࠵?࠵?࠵?࠵?࠵?࠵? ࠵?࠵? ࠵?ℎ࠵?࠵?ℎ ࠵?ℎ࠵? ࠵?࠵?࠵?࠵?࠵?࠵?࠵?࠵?࠵? ࠵?࠵?࠵?࠵? ࠵?࠵?࠵?࠵?࠵?࠵? ࠵?࠵?࠵?࠵?࠵? ࠵?࠵?࠵?࠵?࠵?࠵? ࠵?࠵? ࠵?࠵?࠵?࠵?࠵?࠵?࠵?࠵? ࠵?࠵?࠵?࠵?࠵?࠵?࠵?࠵? 4 Independent events are events that do not influence the probability of occurrence of one another. 5 The multiplicative rule is used to calculate the probability of two or more independent events occurring together. The probability of two or more independent events occurring together is the product of their individual probabilities. 6 An alpha error occurs when you erroneously reject a true H 0 . The probability of making an alpha error is the level of alpha you set.
• • • 