In: Statistics and Probability
A researcher designs an experiment to measure the effectiveness of the new ointment in treating shingles. Singles is a very serious, painful disease. A medical doctor examines the sixteen voluntary subjects. She finds four of subjects each with a moderate case of shingles and of these two are females and two are males. The other twelve subjects each have a severe case of shingles and of these eight are females and 4 are males. All subjects are aged 50 to 60, and except for shingles are in good health. For the standard ointment treatment, the doctor knows that females tend to respond better to the treatment than males do.
The statistician in charge of designing the experiment decides to conduct a double blind experiment using two treatment groups of subjects. Neither the doctor nor any of the subjects will know who receives which ointment. The control group will receive the standard ointment treatment for which the effectiveness is well known. The experimental group will receive new ointment. The two ointments are in identical tubes and both ointments appear identical. The subjects will receive detailed instructions on the application of the ointment. Each subject will apply the same amount of ointment three times a day, which is the recommended dosage of the standard ointment. They will have weekly follow up visits over two months after which the experiment will end. During weekly follow-up visits the doctor will assess if patients are correctly applying the ointment and use a Likert scale from 0, 1, 2, 3, to 4 to judge the severity of the rash with 0 indicating no rash and 4 a very severe rash.
Because males and females tend to respond differently to the standard treatment, the researcher must block males and females into a block of all males and another block of all females. For the male block, he randomly assigns one male with moderate case of shingles to new ointment treatment and the other male to the standard ointment treatment. Likewise, he randomly randomly assigns two of males with a severe case to new ointment treatment and the other two to the standard ointment treatment. Similarly, for the female block, he randomly assigns one moderate case to each of the two treatments. In addition, he randomly assigns four of females with a severe case to new ointment treatment and the other four to the standard ointment treatment. He runs the experiment for two months. For both males and females within each block and between each block, he compares how well and how fast the new ointment and the standard ointment work.
A.Why did the design not include a placebo?
B. Does the lack of a placebo put the results into question?
C. Do you see any design flaws in the original design?
Assume individuals with other serious health problems like cancer have a much harder time curing shingles than otherwise healthy individuals and that generally people older than 70 are also harder to cure.
4) When the experiment is replicated, should this additional information be taken into account? Describe how this would affect the design of the experiment.
A. The design is an experimental design where the main purpose is to find out the effect of Standard Ointment and New Ointment on the clearly defined subjects. in other words which ointment offers better curing effect than other. In case of placebo that could add an an psychological effect with no measurabe parameters. Hence it is not included
B. Lack of placebo will not put the results into question as the result obtained will be free from any bias
C& 4. The focus group chosen here are in the age group of 50 - 60 only. This can be considered a stepping stone for future experiments. If we find there is a difference between Standard Ointment and New Ointment, we can extend the experiemtn in new set of subjects like 70+ patients or patients who have other major health concerns. If there is no difference exists between the two ointments then we can design a different experiements on the effect of age or other disorders.