Statistics questions. Alpha levels. Please answer all parts. Need help.

0 50. Suppose you work for the company that manufactures the Old Drug and you nd out that the above pharmaceutical company manipulated the sample size. You are still convinced that the Old Drug is a better option because it costs almost 75% less than the New Drug. How could you convince doctors that the difference between the effectiveness of the Old Drug and that of the New Drug is not really important, even though it was statistically signicant in the studies conducted by the above pharmaceutical company? 1. Compute the z scores and show them that the z is much larger for the New Drug than the Old Drug. 2. Compute the exact p value and explain that this very small p value shows that the difference between the two drugs is actually very small. 3. Compute the effect size and explain that the actual difference between the drugs is very small. - 51. Suppose a pharmaceutical company has developed a new drug to treat a severe mental disorder that currently has no effective drug treatment available. The statistician responsible for analysis of the data suggests that they use an alpha level of .10 rather than .05 or .01. Is this a legitimate option in this case? 1. Yes, the researchers are justiably concerned about making a Type 11 error and using an alpha of .10 reduces the probability of making this type of error. 2. Yes, the researchers are justiably concerned about making a Type I error and using an alpha of .10 reduces the probability of making this type of error. 3. No, the researchers are justiably concerned about making a Type II error and using an alpha of .10 increases the probability of making this type of error. 4. No, the researchers are justiably concerned . 52. about making a Type I error and using an alpha of .10 increases the probability of making this type of error. A researcher in another company reads the above study and sees that they used an alpha of .10 and decides to use an alpha of .10 in a study his company is currently conducting to test the effectiveness of a new cough drop. Is this a legitimate option in this case? 1. In this case, the researchers are investigating a treatment for a relatively minor problem and treatments already exist. Using an alpha of .10 is a good idea because it will help maximize the statistical power and minimize the probability of making a Type 11 error. Using an alpha of .10 is not acceptable because it would increase the Type I error rate. Although Type II errors would go down, there is little reason to be overly concerned with Type II errors in this study. Because current treatments exist for coughs, incorrectly failing to reject the null should not be the primary concern. 3. It does not matter if the researcher uses an alpha of .10 or .05. Either is acceptable as long as they use a very large sample size. Suppose that the National Institutes of Mental Health have the money to fund two research programs. Although there is money to fund only two research programs, more than 30 research teams have applied for this grant money. A group of experts read all the grant applications and narrow it down to the ve best ones. A number of factors are being considered, but one is the researchers' past success in developing treatments. Each of the researchers submitted information about the type of mental disorder and population they have studied, the sample size, the result of the signicance test, and the effect size. Use the information provided in the following chart to answer the next question: Study Number Type of Cancer/ Population N Significance d 1 Depression in teenagers 500 p <.05 reject ho obsessive compulsive disorder in adults p anxiety the elderly bipolar .05 antisocial teenagers> .05 0.67 Fail to reject Ho . 53. Based on these data, which two researchers would you recommend receive the research grant? Explain your answer using effect sizes, sample sizes, and the results of the hypothesis test