There is no significant difference between the means of the t-cell counts in AIDS patients taking the standard multi – drug regimen generally prescribed to most AIDS patients and the means of the t – cell counts of AIDS patients taking the newly developed drug. ALTERNATIVE HYPOTHESIS : There is a significant difference between the means of the t – cell counts in AIDS patients taking the standard multi-drug regimen generally prescribed to most AIDS patients and the means of the t – cell counts of AIDS patients
taking the newly developed drug. METHODS OF SELECTION OF SUBJECTS FOR THE STUDY: In selecting subjects for the study , we make use of the “simple random sample”. We select two groups of samples of 25 each. We use the “lottery method “ of choosing samples. We first consider the two largest hospitals in the city as source of subjects. From the first hospital chosen, we get all the names of AIDS patients in that particular hospital. From tis list, using lottery method we will select 25 names.
This is the CONTROL GROUP, which will be subjected to medication using the standard regimen. From the second hospital chosen, we again select 25 names using lottery method. This is the TREATMENT GROUP, which will be subjected to newly developed drug. We will make use of a statistical test known as the T- test in testing our Hypothesis. In making tests such as T – test, certain assumptions are to be made. First, it must be assumed that the two samples are random samples independently Drawn from distributions that are normal.
Secondly, it must be assumed that the Variances of the populations from which the samples are drawn are the same. The second condition is referred to as “homogeneity of variance”. Through the use of a table at random numbers , the research worker can be assured that the sample or samples are random ones and are correctly drawn.
The problem of homogeneity of variance is as not easily taken care of. There has been a considerable research related this problem over the years, and today definite statements cannot be made about it to the satisfaction of all 1 statisticians. PROBLEM OF SELECTION BIAS : The potential problem I see in this area, is the differences in stages or number of years a certain AIDS patient has been afflicted by the HIV Virus. The t-cell counts of of those patients afflicted with aids for longer years tend to be less than the t – cell counts of those patients just afflicted by the HIV virus just recently. Selection of subjects for an active control trial can affect outcome.
The population studied should be carefully considered in evaluating what the trial has shown. For example, if many subjects on the trial have previously failed to respond to the control treatment, there would be bias in favor of the new treatment. The results of such a trial could not be generalized to the entire 2 population.
EXPERIMENTER BIAS : The choice of significance level between 0,01 or 0. 05: The researcher may be biased towards choosing between alpha level 0. 01 or 0.05. In the given study, if the given null hypothesis is satisfied or tested true, it Implies that the newly developed drug is a potential subject. The researcher will Favor the choice of 0. 01 significance level, since this gives a greater probability of accepting the null hypothesis. As an illustration, consider the study using T-test, with a degree of freedom = 60, at 0. 05 level Tabular T-Value = 2. 000 at 0. 01 level Tabular Value = 2. 650. If observed T = 2.35, the null hypothesis will be rejected at alpha level 0. 05 While still be accepted at 0. 01 significance level.
RECOMMENDATION : Finally, I propose the use of statistical test using T- test using Two-Sample Tests between population means. Why T- test : A major study was conducted by Borneau (1960). He concluded that the use of T – test will often result in probability statements that are extremely accurate despite the problems in homogeneity of variance and normality of the underlying
distributions. Edwards ( 1973) notes that if the T – test is applied to two independent random samples of size 25 or more, the T – test is relativlely unaffected by rather severe violations of the assumptions of the homogeneity of variance and normality of the distribution of the population. ETHICAL CONCERNS: When a new treatment is tested for a condition for which no effective treatment is known, there is usually no ethical problem with the use of TREATMENT Group.
However ethical problems may arise in this case study, since there is a known standard regimen for AIDS patients to take in order to retard the advancement of AIDS disease. In cases such as this, when there is an available treatment to retard advancement of HIV virus, the research study may experience problems of acceptability.
NOTES : 1. Downie, N. M. and Heath, Robert W. , 1983. Basic Statistical Methods.. New York : Harper and Row Publishers, Inc. 2. http://www. fda. gov/cder/guidance