Null Hypothesis: The new developed drug has no considerable difference from the standard multi-drug regimen used by most cancer patients (µ1 = µ2).
Alternative Hypothesis: The new developed drug is considerably better from the standard multi-drug regimen used by most cancer patients (µ1 > µ2).
The dependent variable in the study is the so-called ‘level of efficacy’ (or in simple term, the level of effectiveness of the drug). This independent variable may be measured by: 1) T-cell counts among AIDS patients, 2) improved blood circulation (measured by blood pumped per ounce per second), and 3) rate of antibody formation. For the sake of simplicity, we shall only consider the first measure of ‘efficacy’ (T-cell counts among AIDS patients). The independent variable in the study is the ‘type of drug’ used to treat patients with AIDS.
For the purpose of theoretical efficiency, we can assume the existence of two groups. Group A is a population sample treated with the ‘new developed drug.’ Group B is a population sample treated with the standard multi-drug regimen. Note that both drugs are assumed to have a general effect on the mitigation of AIDS among patients. A higher population mean (measured by T-cell counts) would indicate a higher ‘level of efficacy.’
Experimenter bias may be exhibited in the study as: 1) error in the specification of experimental maneuver, 2) error in the measurement of outcomes, and 3) faulty interpretation of data. Because of the complexity of the study, it is very likely for the researcher to commit the second error. Selection bias is not present in the study.
There are two pressing ethical issues in the study. First, it is generally unethical to use an untested drug (medical) to a group of AIDS patients (although it may be argued that the drug has been tested many times in the laboratory). Second, it is dangerous to test the efficacy of two sets of drugs to ‘actual’ patients.