|Questions or hypotheses provide further focus
to a study. In qualitative research, the investigation may begin
with a few or no questions, and generate questions from the observations.
Survey research often has a few general questions, each of which are investigated
with several additional more specific questions. Experimental research
may start with questions, but translates them into hypotheses, which state
tentative answers to specific questions of interest, usually postulating
no difference, and then test the hypotheses.
There are four common types of research questions, and the type of question
has critical implications for the appropriate research design.
The first listed types of questions are generally easier to answer than
the latter ones. For descriptive questions, there is only the need
to measure the characteristics of interest in the population of interest.
For associational questions, categorical comparisons or correlations then
are computed for the measured characteristics. To answer causal questions,
there is a need to establish (a) association, (b) temporal ordering (the
alleged causal variables must occur before the outcome variables of interest),
and (c) isolation (ascertaining that other variables may not have caused
the outcomes). For benefit-cost questions, there is also a need to
compare the value of the benefits to the costs of the intervention program.
Descriptive questions (about the characteristics of something).
Example: What is the average math achievement of the nationís fourth graders?
What is the career satisfaction of DC computer industry professional employees?
Associational questions (about categorical differences or correlations).
Example: How does the math achievement of the nationís fourth grade Hispanic
youth compare with that of the nationís fourth grade African-American youth?
Are District of Columbia computer industry professional employees more
or less satisfied with their careers than are DC media industry professional
Causal questions (about what causes observed differences or changes).
Example: What accounts for the wide variation in math achievement among
fourth grade Hispanic youth? Will two hours a week of Math Explosion
exercises over a semester raise the math achievement of low-achieving fourth
grade Hispanic youth?
Benefit-cost questions (about whether the benefits of a given intervention
are worth more than the costs). Example: Do the benefits of Math
Buster exercises exceed the costs of the computer laboratory, software,
and laboratory supervision? Is the benefit/cost ratio of Math Explosion
exercises greater than the benefit/cost ratio of after school tutoring
by the teachers?
or Advance to Lesson A-4
3. What are the questions investigated or hypotheses tests? (Make
sure to distinguish among the four types of questions indicated above.)
Last Update: June 29, 2000