EDU738: Educ. Research Across the Curriculum

'... that Truth is not a colored bird to be chased among the rocks and captured by its tail, but a skeptical attitude toward life.' from "Arrowsmith" by Sinclair Lewis

Comparing Research Questions and Hypotheses - So What's the Difference?



As with many concepts, the terms "research question" and "hypothesis" are not entirely unrelated and are sometimes used interchangeably. But for the purposes of our course, I want to emphasize one key difference between them that has ramifications for data collection and data analysis.

You can think of
hypotheses as being multiple choice items with predefined response options such as: "yes, greater than", "no, not greater than", or "no, the same as." For example:

"Girls ask the reference librarian more questions than do boys."

On the other hand, useful
research questions are open-ended items where the questioner is open to a range of response options. For example:

How do boys and girls make use of the school library's resources?

To confuse matters, oftentimes, in published research studies the research hypothesis will be presented as a question, for example, "Do girls check out more books from the library than boys?" But this question is really an implicit research hypothesis: "Girls check out more books from the school library than boys do."

Even a question like "Do boys and girls check out different amounts of books from the school library?" is really an implicit non-directional hypothesis. For example written as a hypothesis this would be: "There is a difference in the number of books checked out of the school library by boys and girls.'"

Given this varying use of the term 'research question', another way to highlight the difference between research questions and hypotheses is that
research questions are generally 'why' and 'how' type questions while, in contrast, hypotheses are 'how much' type questions.

So, if you are pretty clear about the phenomenon you are researching and the range of outcomes you expect to encounter, you probably want to generate a
hypothesis. If the territory is not very well mapped, you probably want to use a more open-ended type of research question.