Question

Exercise Two

(4 pts.) If I wanted to conduct a clinical trial examining the clinical efficacy of a new antidepressant called Quipazine, how would I set up this study to be double blind and why would it be important to do so? Be specific in your description of how the study should be implemented and why it is important to conduct the study double blind. (If you just provide a definition of double blind and a very generic description of the study, you will receive no credit for your answer.)

(2.5 pts.) The answers to the following questionnaire items are based on what scale of measurement? (nominal, ordinal, interval, ratio)

1. Your Macy's account number (do not actually provide me the number!)
2. How many cats live in your household?
3. What is your grade point average in your major?
4. How many hours do you spend on the Internet each week?
5. What temperature is our classroom in the East Academic Building?

(2 pts.) What is the difference between a parametric and nonparametric statistic? How is this related to different scales of measurement?

(1.pt.) Why is it important to be able to differentiate between discrete and continuous data?

Answer 1

(1)

Out of the total number of n volunteers, randomly assign n/2 volunteers to Experimental Group and n/2 volunteers to Placebo Group. This random assignment is to be carried out by a third unconnected person in such a way that both the researcher as well as the volunteer knows who is in which Group. The volunteers in Experimental Group will be given new antidepressant called Quipazine by the third unconnected person. The volunteers in Placebo Group will be given identical looking but without the new antidepressant called Quipazine by the third unconnected person. The final result will be measured from all the n volunteers by the third unconnected person and will be presented to researcher for analysis.

(2)

(i) Account number: nominal

(ii) Number of cats: ratio

(iii) Grade Point Average: ratio

(iv) hours: ratio

(v) temperature: interval

(3)

Parametric statistic makes assumptions about the parameters of the population distribution from which the data are drawn. Non-parametric statistic makes no such assumptions. In the case of parametric statistic, the scale of measurement is interval or ratio. In non-parametric statistic, the scale of measurement is ordinal.

(4)

It is important to be able to differentiate between discrete and continuous data because by the very definition discrete data is defined only for a specific values of the independent variable whereas continuous data is defined for all values in the given interval and due to this difference the statistical treatment of discrete data is entirely different from that of continuous data.