Bacon once wrote, "a prudent question is one-half of wisdom."
problem statements, according to Reinard (1998), set limits on relevant
information and structure inquiry. He asserts that there are two
prerequisites for useful problem statements: (1) questions
must be within the researcher's competence and capabilities;
and (2) questions must be narrow but nontrivial.
questions are the explicit questions researchers
ask about the variables that interest them. Two general questions
researchers ask about communication are "What
is the nature of communication?" and "How
is communication related to other variables?"
researchers feel confident enought to make a prediction, they advance a
Hypothesis, a tentative statement
about the relationship between the independent and dependent variables.
important criterion for selecting a topic
is WHETHER IT INVESTIGATES SOMETHING IMPORTANT. Researchers should
be able to answer SO WHAT? or Who Cares? To do so requires
researchers to develop a clear rationale for why the research is being
Botan, Friedman, & Kreps (1991) identify the research process cycle
model as a five phase process:
order to choose an appropriate research topic, one must first understand
that communication researchers generate ideas which they consider worth
studying from one of two sources:
2: Planning and Desinging Research
3: Methodologies for Conducting Research
4: Analyzing and Interpreting Data
(1998) distinguishes hypotheses
from research questions
when he writes,
need to TEST and REFINE Theory.
need to SOLVE a PRACTICAL PROBLEM
or PURE Research
be DEDUCTIVE (Theory-Driven) or INDUCTIVE (Theory-Generating)
Research seeks to SOLVE PRACTICAL PROBLEMS
research starts with a theory and then seeks to learn whether empirical
(observable) data suport it.
research sets out to contribute to knowledge by answering a real, pragmatic,
social question or by solving a real, pragmatic, social problem.
good hypothesis, according to Bostrom (1998) consists of three primary
represent a formal testing of a theory and a research question represents
a more general exploration of phenomena. . . the choice of hypotheses as
opposed to research questions is determined largely by the state of theory
in the area, the statistical procedure used, and the preferences of the
might also add the following to Bostrom's characteristics:
A good hypotheses is concerned with only one aspect of the research at
a time and is thus, SIMPLE.
A good hypothesis is OBSERVABLE.
It is related to the theoretical phenomena and actually can be observed.
A good hypothesis is TESTABLE.
It is possible to actually make the observations involved.
(1998) argues that "proper phrasing of problems is required to make them
useful." He suggests the following criteria for sound problem
A good good hypothesis investigates something
IMPORTANT--both THEORETICALLY and PRACTICALLY.
A good hypothesis must be of sufficient SCOPE
(1998) suggests that there is utility in understanding what problem
statements are NOT!
Problem statements must be stated unambiguously,
usually as questions.
Except for simple descriptive studies, problem statements must
include at least two variables.
Problem statements must be testable.
Problem statements must not advance value
Problem statements must be clear grammatical
we may find some standards for research questions, it is also a good
idea to know what are NOT useful research questions.
The types of pseudo-problem statements may trick us into spending our time
on them when in fact, they may not be worth our time.
PROBLEM STATEMENTS ARE NOT . . .
Questions asking for obvious yes or no answers.
Questions asking about applying a statistical
tool. These questions confuse the method of
answering the question with the research questions themselves. Problem
questions should ask for relationships, not statistical methods.
Questions proposing personal learning goals.
In fact, personal references have no role in productive research questions.
The following example is inappropriate: "The purpose of this study
is to learn ways advertisers sell products." Purpose statements do
not "learn" anything. The research question should get to the point
and ask about relationships.
Questions that have already been competently
Questions that cannot really be solved.
(Whether because of the lack of resource materials or because the
question deals with broad, philosophical issues that cannot be resolved).
most simple way to phrase hypotheses is to create a statement of simple
example, you could hypothesize that:
There will be a direct
(or positive, or negative, or inverse, or curvilinear) relationship
between (insert favorite first variable of interest)
(insert favorite second variable of interest).
favorite first variable of interest) increases (or decreases,
etc.) the (insert favorite second variable
of interest) decreases (or increases, etc.).
What patterns of deceptive
communication characterize close relationships?
What are the perceived
dimensions of giving feedback in a task-oriented small group?
What types of power
strategies do subordinates perceive themselves using in interactions with
Do men and women report
significantly different levles of satisfaction with conflict interaction?
What is the association
between how patients communicate about their illness, other patient illness-related
behavior, and utilization factors of medical treatment?
How does family mediation
of television affect children's comprehension of specific programs and
of television in general?
What is the relationship
between the amount of television watched by adults and the level of anxiety
about personal safety?
Do managers who use
responsive listening behaviors with subordinates report greater communication
satisfaction than other managers?