Information Technology Reference
In-Depth Information
Other examples of variables can be mentioned to provide you with a
sense of what needs to be considered in designing and analyzing the data
from a research study:
If you wish to compare females and males on some performance
measure, both genders must be represented in the study.
If you wish to compare psychodynamic, cognitive-behavioral, and
client-centered psychotherapy on improvement in general function-
ing, then all three types of therapy must be included in the study.
If you wish to evaluate the differences in standardized math test
performance in school districts of different sizes, then different sized
school districts must be represented in the study.
The same argument holds for dependent or outcome variables even
though we have no control in advance of the study over what values
these measures may take. The key is to assure ourselves in advance of the
data collection that the values we could have obtained on the dependent
variable were capable of being different and that there was nothing about
the way we designed the data collection process that may have biased
or limited these values in any reasonable way. In the mood study, for
example, we would not want to expose the students to stimuli that might
make them especially joyous or sad; that could certainly bias their feelings
and corrupt any conclusions we might otherwise draw from the results of
the experiment.
1.5 EXPLORATORY RESEARCH AND HYPOTHESIS TESTING
Scientists do not collect data at random. Instead, we have a rather specific
idea about the particular subject matter on which we intend to focus in a
particular study. These ideas can be more specific or less specific. Up to a
point, the mood study we have been describing could have resulted from
either strategy.
The focus of the research could be somewhat broad and less specific.
For example, researchers may wish to study the variety of factors that affect
mood, and select the color of the surroundings as one of the many variables
they will use. Based on this, they design a study in which students complete
a mood inventory in either a blue or red room. This sort of research is
considered exploratory, and the researchers may have few expectations
but instead intend to simply observe what happens.
The research focus could also be relatively specific. For example, the
researchers may know that color of the surroundings can affect mood,
and that blue tones create relative calm whereas red tones create more
excitement. Now they wish to put their expectations to a specific test.
Based on this, they design a study in which students complete a mood
inventory in either a blue or red room. In conjunction with this design, they
formulate a hypothesis - specific predictions derived from their model or
understanding of how external stimuli affect mood - that students in the
