Joint control tasks

CSGnet Archive CSG2007
1

I just had an idea for a tracking experiment that could be carried out by anyone with Bill's mouse-tracking demos, and two mice. Perhaps someone already has?

Plug both mice into the same computer (I've tried this -- their outputs are added together) and have two people attempt to perform the same tracking task simultaneously. Will they do better, the same, or worse than either one acting alone? If worse, will they come to blows over it?

How about one person operating both mice, one in each hand? Do they control better or worse than with one?

There are various measures of what someone is controlling, e.g. correlation of mouse with disturbance. What will those measures look like when two people are performing the task?

I don't have any particular hypothesis to test here, it just looks like something whose result, whatever it is, might be interesting.

For a more sophisticated version of the experiment (that I have no idea how to program) use two computers in different rooms, connected over a network. Program them so that only one tracking task is being performed, both screens showing the same thing, with the effects of the two mice being added together. The participants may or may not be told of each others' existence. Will one of them end up slacking off and leaving it to the other to do the controlling?

What happens if the two screens do not show exactly the same thing, but one has the reference point or the cursor displaced a small or large amount from the other?

I got to thinking about this in considering the idea of "cooperative conflict": two people trying to control more or less the same perception with more or less the same reference, their efforts being added together, yet failing. It's easy to help someone push-start a car by pushing alongside them, but you can't help someone thread a needle that way.

···

--
Richard Kennaway, jrk@cmp.uea.ac.uk, http://www.cmp.uea.ac.uk/~jrk/
School of Computing Sciences,
University of East Anglia, Norwich NR4 7TJ, U.K.

2

I just had an idea for a
tracking experiment that could be carried out by anyone with Bill’s
mouse-tracking demos, and two mice. Perhaps someone already
has?
[From Bill Powers (2007.06.27.1245 MDT)]

Richard Kennaway (2007.06.27) –

Tom Bourbon did this 15 over 20 years ago. I fixed him up with a game
port with two potentiometers patched into the two axes where the joystick
goes. He wrote the programs in Turbo Pascal. He had two people
controlling a cursor, sometimes together and sometimes by disturbing the
other person’s cursor. I don’t have a reference to that, but I suspect
that Dag Forssell can come up

with one.

Your suggestions for further explorations are all excellent. I do wish
there were a pool of graduate students looking for thesis topics – this
would be a great one. Warren Mansell is about the only one at present,
and he is starting to move into simulations. You will see him in
Manchester next Fall, I presume.

Best,

Bill P.

3

[From Rick Marken (2007.06.27.1220)]

I just had an idea for a tracking experiment that could be carried
out by anyone with Bill's mouse-tracking demos, and two mice.
Perhaps someone already has?

Yes. I've done stuff like this and Tom Bourbon as well. I've done the
stuff where both mice are used by the same person to control for a
particular result; Tom (at my suggestion;-)) did a _lot_ of work on
what happens when two different people use each mouse to control the
result. Oh, and one can do this with my tracking demo at:

http://www.mindreadings.com/ControlDemo/BasicTrack.html

Plug both mice into the same computer (I've tried this -- their
outputs are added together)

That is so cool! Yes, it does work.

and have two people attempt to perform
the same tracking task simultaneously. Will they do better, the
same, or worse than either one acting alone? If worse, will they
come to blows over it?

That would be a great demo. They would probably do worse, possibly
very much worse if both stay high gain about keeping it on target and
the parallax difference results in a difference in what constitutes on
target for the two.

For a more sophisticated version of the experiment (that I have no
idea how to program) use two computers in different rooms, connected
over a network. Program them so that only one tracking task is being
performed, both screens showing the same thing, with the effects of
the two mice being added together. The participants may or may not
be told of each others' existence. Will one of them end up slacking
off and leaving it to the other to do the controlling?

That is a great idea!! I will try to get back into Java programming
and do that as a project.

What happens if the two screens do not show exactly the same thing,
but one has the reference point or the cursor displaced a small or
large amount from the other?

Conflict, big time!

I got to thinking about this in considering the idea of "cooperative
conflict": two people trying to control more or less the same
perception with more or less the same reference, their efforts being
added together, yet failing. It's easy to help someone push-start a
car by pushing alongside them, but you can't help someone thread a
needle that way.

Yes. I think Bill has spoken to this in some context. Sometimes
cooperation works -- like when people cooperate to build a skyscraper
-- and sometimes it doesn't -- like when you try to have a team of
doctors do brain surgery. The trick is to know when it will work and
when it won't.

Best

Rick

···

On 6/27/07, Richard Kennaway <jrk@cmp.uea.ac.uk> wrote:
--
Richard S. Marken PhD
rsmarken@gmail.com

4

[from Tracy Harms 12:29 PM 6/27/2007]

A well-known example of this, contrived for social
amusement, is the "three-legged race."

--- Richard Kennaway wrote:

···

...

I got to thinking about this in considering the idea
of "cooperative
conflict": two people trying to control more or less
the same
perception with more or less the same reference,
their efforts being
added together, yet failing. It's easy to help
someone push-start a
car by pushing alongside them, but you can't help
someone thread a
needle that way.

____________________________________________________________________________________
Sucker-punch spam with award-winning protection.
Try the free Yahoo! Mail Beta.
http://advision.webevents.yahoo.com/mailbeta/features_spam.html

5

[From Rick Marken (2007.06.27.1235)]

Tracy Harms(12:29 PM 6/27/2007) to Richard Kennaway

A well-known example of this, contrived for social
amusement, is the "three-legged race."

A nifty example of a task where cooperation is _required_. It's not
quite the same as Richard's two-mouse situation, though, where the
best thing for one party to do (if both parties want the cursor in the
same place) is to stop moving their mouse. If one person in a three
legged race stopped running, neither would get to the finish line.

Best

Rick

···

--
Richard S. Marken PhD
rsmarken@gmail.com

6

[From Richard Kennaway (2007.06.27 2037 BST)]

[from Tracy Harms 12:29 PM 6/27/2007]

A well-known example of this, contrived for social
amusement, is the "three-legged race."

Excellent!

···

--
Richard Kennaway, jrk@cmp.uea.ac.uk, Richard Kennaway
School of Computing Sciences,
University of East Anglia, Norwich NR4 7TJ, U.K.

7

[From Richard Kennaway (2007.06.27 2037 BST)]

[From Rick Marken (2007.06.27.1235)]

Tracy Harms(12:29 PM 6/27/2007) to Richard Kennaway

A well-known example of this, contrived for social
amusement, is the "three-legged race."

A nifty example of a task where cooperation is _required_. It's not
quite the same as Richard's two-mouse situation, though, where the
best thing for one party to do (if both parties want the cursor in the
same place) is to stop moving their mouse. If one person in a three
legged race stopped running, neither would get to the finish line.

Another example of this sort would be a version of the rubber-band demo, in which the two people are asked to move the knot along a path drawn on the board.

···

--
Richard Kennaway, jrk@cmp.uea.ac.uk, Richard Kennaway
School of Computing Sciences,
University of East Anglia, Norwich NR4 7TJ, U.K.