Question

"Hypothesis Testing" Please respond to the following:

• Identify a problem in healthcare that could be illuminated by research. Develop a research question and describe a research design to address it. Be sure to identify your dependent and independent variables, sampling strategy, and statistical tests. Write a null and alternative hypothesis for your research study.
• Assume you rejected a null hypothesis at five percent (5%) level of significance. Give your opinion as to whether you would get the same level of statistical significance if you were to do the experiment over multiple times. Explain your response.

Answer 1

Solution:

Suppose a pharmaceutical company wants to test the effectiveness of a particular drug.

So, in this case the research question is how successfully the drug can help to recover from some disease.

We have to apply the drug for multiple cases, and check the proportion of times it's effectively removing disease.

The independent variable in this case will be the application of the drug and dependent variable is whether the person has recovered or not.

Here the sampling strategy could be to take say 200 cases in which the drug has been applied and to note it down for each cases whether the person is cured or not.

The outcome will be either recovered or not. So the outcome is binary. We have go for one-proportion test in this circumstances. Say the manufacturer's claim is it is effective in 95% cases.

So we have to test H0: p = 0.95 against H1: p < 0.95, where, p is the proportion of cases in which drug was effective.