Q1) You are starting to feel more comfortable in your position as Special Projects Director in charge of statistical analysis for Harry Hines Medical System (HHMS). CEO, Dr. Biggs Schott, has observed this and wishes you to recreate your analysis of HHMS's managed care provider contracts. In particular, Dr. Schott wishes to know whether utilization rate is higher than 60 days (per 1,000 enrolees per month) rate expected when HHMS entered into contracts.
Digging through your files, you retrieve following data. In population of physician groups, mean utilization rate for hospital patient days is 60 days (per 1,000 enrollees per month) with standard deviation of 12 days (per 1,000 enrolees per month. Random sample of 25 physician groups (where each group represents single observation in sample) found mean utilization rate be= 75 days (per 1,000 enrolees per month) with sample standard deviation of 16 days.
a) As you do not care if utilization rate is less than expected, use one-tail Z-test to test null hypothesis that true population mean is less than 60 day estimate with 5% Type - I error rate (i.e., alpha = 0.05).
1) Null hypothesis is give. In place provided, write alternative hypothesis.
Ho : µ ≤ 60
Ha :
2) Calculate the test statistic Z.
3) Determine critical value for boundary of the rejection region (i.e., Zα).
4) Assume answer in [3] above was critical value Zα = 2.0. Would you reject null hypothesis (using test statistic you computed in [2])?
5) Also, mention your conclusion in terms of situation in the problem.
b) Assume your assistant performed analysis for you, but only told you that one-tail p-value was 0.075. Would you reject null hypothesis at a 5% Type - I error rate?