Fundamentals of Social Research
Adam J. McKee, Ph.D.
This content is released as a draft version for comment by the scholarly community. Please do not distribute.
Your research hypothesis is your statement of how you think things will turn out. What follows are some suggestions for writing good research hypotheses.
Suggestion 1: A good hypothesis should name two or more variables and explain the expected relationship between them.
Example: There is a negative relationship between annual household income and crime.
In the above example, “household income” and “crime” are the variables of interest. The researcher expects that as household income goes up, involvement in criminal activity will go down.
Suggestion 2: A good hypothesis is declarative and not phrased as a question.
Suggestion 3: If a relationship is expected among only certain types of individuals, then that population should be explicitly stated in the hypothesis.
Example: Among juvenile females, Therapy X will effectively reduce drug use. (The researcher expects that the treatment (Therapy X) will only have the predicted impact on juvenile females).
Suggestion 4: State your hypothesis so that it is as specific as possible while not exceeding a single sentence in length.
Suggestion 5: If a comparison is to be made, then explicitly state what elements are going to be compared.
Examples: Bad: Females are more likely to respond positively to Therapy X. Better: Females are more likely to respond positively to Therapy X than males.
Suggestion 6: Omit needless words.
A good hypothesis will be free of words and phrases that do not add meaning.
Suggestion 7: Limit the scope of your hypothesis to what you will actually study.
That is, do not include implications or value judgments about your predicted outcomes.
Suggestion 8: Avoid the term “significant” in all of its forms.
Remember that in research “significant” has a special meaning. Avoid this term to prevent confusing your reader.
Suggestion 9: Never use the word prove. As we discussed in previous chapters, scientists never prove anything.
Suggestion 10: Give your variables a single name and stick with it.
This suggestion directly contradicts what English professors tell you. Conventional wisdom says that repetition is dull and that you should vary your words. In the sciences, however, precision mandates that variables should always be called by the same name.
Suggestion 11: Merge more than one hypotheses into a “statement of the hypotheses” if clarity can be maintained.
Example: Individuals who undergo Treatment X will experience higher success at remaining drug-free and lower incidents of recidivism than those receiving Treatment Y.
In the above example, there are really two hypotheses—one having to do with “remaining drug fee” and the other dealing with recidivism. Since both of these dependent variables are closely related to the treatment variable, it makes sense to merge them into a single statement.
Suggestion 12: In a large study that will test many hypotheses, present each hypothesis in a numbered list.
Suggestion 13: State your hypothesis explicitly before you begin discussing your methods.
It is much easier for your readers to evaluate the adequacy of your methodology when they understand exactly what you are trying to accomplish.
Suggestion 14: State your research hypothesis. It is generally inappropriate to just state your statistical null hypothesis. If the research hypothesis is known, then the informed reader will assume the null hypothesis.
Suggestion 15: A hypothesis should be testable. This will usually mean that your variables are defined in operational terms.
Modification History File Created: 07/25/2018 Last Modified: 07/25/2018
This work is licensed under an Open Educational Resource-Quality Master Source (OER-QMS) License.