ThreeCV1: Rules, Please

[From Bruce Abbott (950105.1930 EST)]

Bill Powers (950105.1330 MST) --

ThreeCV1.pas has been received, installed, and run. It has an oddly familiar
look to it...hmmm, where have I seen this before? (;->

You need to clarify something for me before we get into the business of
constructing a "conventional" analysis of the data provided by this demo:
what are the "rules" of the game? That is, what are we to assume the
conventional analyst knows about the situation? Does the analysis have all
the data collected in the demo at his or her disposal (i.e., mouse positions,
cursor positions, disturbance values)? What is the conventional analyst
permitted to do during the data-gathering phase? Simply observe a
participant? Or can the "researcher" actually experiment by, for example,
manipulating the mouse _without_ attempting to control, or run the demo with
and without variable mouse input? [Note: if the researcher can only observe,
then this is not an experiment but a correlational study, with attendant loss
of ability to determine cause-and-effect.] Finally, is the "researcher"
assumed to know that the participant is attempting to keep one (and only one)
cursor on target or are we to assume that this is something the analysis is
supposed to "discover"?

In my suggested analysis of the previous demo, I took the role of experimenter
by manipulating the mouse and noting the status of the screen variables.
However, in so doing I was learning about the _environment_, not about the
participant's performance within that environment. I proposed to learn about
the environment first and then apply that knowledge to the analysis of data
collected during a normal session.

The problem for conventional analysis here is that the "researcher" has no way
to actually manipulate what he or she might take to be the causal variables
(i.e., display variables) and observe the effects of those manipulations on
the participant's behavior. The current version suffers the same defect. The
only other option, if the researcher wishes to learn from experiment, is to
manipulate the mouse.

As an ol' operant guy, I would want to plot the individual data points in
various ways. It would be natural to suppose that the drift (relative to
target) of the cursor chosen for control is the cause and that mouse movement
is the effect, so these variables would be plotted one against the other to
determine their relationship. If the trend appeared to be linear, I would
probably do the regression analysis and report the stats relating to goodness
of fit. If permitted to experiment, I would examine the available variables
when the participant was attempting to keep the cursor on target, when the
participant simply moved the mouse to various positions, and when the mouse
did not move at all. To gather these data would require more than one run of
the demo. Is this permitted?

Despite my total inability to understand the principles of the behavioral
control of perception, which I am told on good authority is conferred only
upon "simple" folk like Rick, Tom, and Bill Leach (;->, I seem to know what
needs to be done to analyze these data correctly, making it difficult to know
how a truly naive but competent researcher would approach this challenge. (In
effect, I've "peeked" at the answer.) So tell me the rules, and I'll try to
get a few of my control-naive colleagues to give it a try. They're not a
random sample of all research psychologists, but they do offer a good cross
section of specialties, ranging from neuroscience to personality.