From Page 85
(3) Very good. ROC or TLR is the one factor of utmost concern, because it
states the minimum essential operational capabilities, and technical,
logistical, and cost information needed to initiate full scale development.
Well, we've got two questions to go, and still need to talk about
statistics, experimental design, and psychophysical methods.
Let's take
these topics in order of appearance in this course, which means Lesson 30
comes first--psychophysical methods. This was the first of the five lessons
to deal with experimental methodology and statistical techniques. If you'll
remember, psychophysical methods are concerned with the relationship between
stimulus intensity and sensation, or the perceived level of stimulation.
And the purpose of these methods is to provide procedures for determining
thresholds.
The two lessons that followed (Lessons 31 and 32) were concerned with
experimental design, and we touched on things such as different ways to
classify variables and different types of research methods. A good deal of
time was spent in Lesson 31 examining the various types of research methods,
be they theoretical or empirical.
The majority of the lesson was spent
discussing this second type of research, breaking it down into observation,
correlation, or experimentation.
Lesson 32 deals more directly with specific problems found in
empirical research and how to overcome these problems by controlling your
relevant variables (be they subject, sequence, or situational RVs). Then,
some time was spent dealing with internal and external validity as it
relates to experimental design. When we talk about the internal validity of
an experiment, we are addressing ourselves to the issue of whether our
experimental results can be attributed confidently to the independent
variables manipulated by the experimenter. Now, when moving from internal
validity to external validity, one is moving from the mechanics of design to
what?
(1) The mechanics of face validity. Turn to Page 27.
(2) Presenting a good external idea to your peers of what your experiment's
trying to evaluate. Turn to Page 34.
(3) The ability to generalize from your results. Turn to Page 55.
67
Previous Page
