[Philip 6/5]
RM: The results of this experiment shows that there is no aspect of the stimulus input to a control system that seems to serve as the cause of the responses that control that input.
PY: Here are the results of my run.
Can you explain these results? Hint: my response was a “movement made from memory”.
[From Rick Marken (2015.06.06.0910)]
[Philip 6/5]
RM: The results of this experiment shows that there is no aspect of the stimulus input to a control system that seems to serve as the cause of the responses that control that input.
PY: Here are the results of my run.
RM: Thanks for doing this Philip! Glad someone’s trying these demos. It also led me to try the demo again and I found that the animation works terribly when run in the Chrome browser on a computer running Windows 7.
PY: Can you explain these results? Hint: my response was a “movement made from memory”.
RM: One thing I have to do is make the symbols clearer because I can’t tell which is the stimulus (cursor) trace and which is the mouse. But my guess is that the square wave is your mouse movement which was repeated from memory. But what I don’t understand is how you knew that the repeated section was starting; in the Javascript version the word “Repeat” is not printed – something else I have to fix. Were you using the Java version?
RM: So my explanation of the results is that there is a high correlation between stimulus and responses on the repeated runs because you were controlling the mouse on the two trials, not the cursor. If you had controlled the cursor – and kept it in a fixed position -- you would have found the results that are expected – high correlation between responses on the repeated trials and no correlation between stimuli (cursor movements) on the two trials.
RM: If, however, you had controlled the cursor relative to a changing reference – that changed the same way on the two trials – you would have found what you found here also: a high correlation between stimuli and responses on the repeated trials. In that case, the only way to tell that the cursor is not the cause of responses is by looking at the correlation between cursor movements and mouse movement; the correlation will be low and, move importantly, the highest correlation between cursor and mouse movements will be found when mouse movements lead cursor movements. If cursor movements were causing movements the best correlation would be found when cursor movements lead mouse movements.
RM: Now off to try to fix this demo.
Best
Rick
–
Richard S. Marken
www.mindreadings.com
Author of Doing Research on Purpose.
Now available from Amazon or Barnes & Noble