HW 4
Note: All of your statistical tests must be done so that you reject the null hypothesis
at the alpha=. 05 or less. You must provide your raw data.
1. You are working for a pharmaceutical company and have been asked to determine
if any of the three new antidepressants are effective in reducing the magnitude of
depressive symptoms. You are also concerned if any of the new antidepressants
are more efficacious (more effective) than the others. Patients were randomly
assigned to one of four conditions. Neither the experimenter nor the patient knew
what type of compound they were given. Group 1 was given a placebo, Group 2
was given the drug paroxetine, Group 3 was given the drug, fluvoximine, and
Group 4 was given the drug fluoxetine. Patients were given the drug for two
months and then given an assessment of overall depression. This is a 50­point
scale, where low numbers indicate less depression and higher number indicate
more depression. YOU NEED TO CREATE YOUR OWN DATA SET.
PLEASE BE SURE TO GIVE ME YOUR DATA AS WELL AS ALL
CALCULATIONS (30 Points)
? Be sure to answer the following:
o Reframe the question into a research hypothesis (Alternative) and null
hypothesis. Be sure to do this in words as well as symbolically
o What does your statistical tests tell you about the research hypothesis. You
must describe your results.
o What is your eta squared from you statistical analysis. Describe what this
number tell you about your experiment.
o Based on your results what recommendations will you be able to make
about these new antidepressants to the Vice President of Research and
Development for this company
2. The University of the Arts has just hired you to determine if productivity of their
faculty is related to job satisfaction. You decide to give a sample of faculty the Myerson
job satisfaction questionnaire. This is a 25­question test and the higher the score the
more satisfied the person is with their job. You also measure productivity based on the
number of times the person has shown their art in a year in galleries. Create the data set
and then decide what you can tell the Director of personnel. (20 points)
3. Discuss the differences between within group variance and between group variance.
What does a t or F test test? (15 points)
4. I want you to design a study in which you want to predict the number of times that
someone visits a health professional in a year. Note that you must make three
predictions. You also need to tell me about the strength of the relationship between x and
y. Be sure to tell what r squared is from your data and what this tells you about your
data. In this problem you think that the number of mental health symptoms experienced
in the same time period may be an important as a predictor variable. Make up the data
set and carryout the appropriate quantitative test. Describe your findings and tell me how
good are your predictions? (20 Points)
5. Discuss the concept of the standard error of the estimate (10 points)
6. If alpha = .085 can we reject the null hypothesis. Justify your answer. (5 points)