Dr. Hannley said that once the project is determined, then the researcher needs to define the population to be studied, define the period of time for the study, and select the variables to be measured.
Explore This IssueApril 2007
She said that writing the hypothesis for the study is another basic part of the process. I particularly believe you should use the null hypothesis: For example, ‘There is no difference in symptom resolution or disease-specific quality of life in children with recurrent acute sinusitis treated with endoscopic sinus surgery and those treated medically.’
Another form of hypothesis is the alternative hypothesis, such as Children with recurrent acute sinusitis treated with endoscopic sinus surgery will have significantly better symptom resolution and better disease-specific quality of life than those treated medically.
Dr. Hannley said she favors the null hypothesis because it tends to avoid bias. In the null hypothesis, the researcher is stating that he or she is looking at the two treatments as being equal, not that the researcher expects to find that one treatment is better than the other.
State your hypothesis in a clear, concise sentence, she said. It should be simple specific, and stated in advance.
She cautioned researchers to be careful in determining what will be measured in the study. Be parsimonious: If it won’t add to your answer, don’t do it. Each additional variable complicates your statistics and increases your sample size requirement, she said.
Among the variables to consider in the study are the ages of the participants, the involved site, the disease outcome, the tissue type, and whether the variables scrutinized are dependent, independent, confounding, or are background information.
She also suggested that before embarking on the project, researchers should consult a statistician to help design the type of statistical treatment and the required sample size estimations.
Dr. Hannley said that is also essential that the researcher discuss and involve a sponsor or mentor in the planning of the study and perform a careful, comprehensive literature review.
The researcher also has to determine the study strategy, Dr. Hannley said. That means choosing between a basic science and a clinical investigation, between a prospective or retrospective study, the duration of the study, and other factors.
When a researcher selects a the sample, he or she needs to describe the characteristics of the subjects who will be eligible for participation in the study and also describe the population outside the sample selected to which the researcher wishes to generalize the conclusions.