Variability

[olafsdot]

Hello,

I am doing the following study (in theory)
Subjects have in front of them 2 targets, x inches a part and the task is to tap between them as fast as they can.
The target to the left is defined as "start point" and the target on the right is defined as "end point".

I measure the variability of the cordinates of the "end point"

In this case the "start point" location varies because the subjects tap back and forth.

Can anybody clarify for me how a variable "start point" affects the variability of the "end point" vs. a case where the "start point" was always the same (variability 0)?

Thank you!

[Junes]

I'm not quite sure I follow you, but this is what I understand (correct me if I'm wrong):

IV: Location start point (fixed/variable)
DV: difference between end point (as presented) and location where subjects actually end up, in inches

Did I summarize your DV correctly? You said you measure end point, but since that seems to be controlled by the software I don't see how that can be your DV. So is your DV instead the difference (inaccuracy) between target end point and subject end point?

And your question is how start point being fixed/variable affects this difference (or rather, the variability, since the mean will presumably be 0 if there's no bias towards either side)?

That's not really a stats question I think. It's rather the topic of your research, so it's something you can empirically determine and/or find in the literature.

[olafsdot]

Thank you for the reply but you completely misunderstand the task.
I attached an illustration that may explain it better.

I forgot to mention that the instructions subjects get is to tap within the targets.

The DV is the x-coordinate of the points in the "end target" (the task is performed on a digitized tablet)
The IV is in this case the "Start target" or "Start point", that is if it is a "target" of certain width (thus has a non-zero variability) or a "line" (thus approx. 0 variability).

My question, again is: statistically speaking, shouldn´t there be a difference in the "amount" of variability of the DV between those two cases? How would you describe those two cases from statistical point of view if you had to model them?

Thank you again!

[olafsdot]

Hello,

I'm bumping this question...
I've asked around but nobody has been able to clarify this for me....all input would be appreciated!!

[Junes]

Hi olafsdot, I did have a short look at it and I would still like to reply. Might be tonight (GMT+1) or later, though.

[olafsdot]

Thank you Junes, it would be a great help if you can give me some insight into the solution of this problem!

[Junes]

Looking at it in more detail, I'd still say this more a theoretical/empirical question regarding motor psychology than a statistical question.

So, to summarize:

* IV: start position, DV: x-position
* in condition Line subjects start at a line.
* in condition Target subjects start at a rectangular target
* in both conditions subjects tap back and forth
* presumably, then, the DV is actually a vector of horizontal positions (since the subjects reaches the End target multiple times). Or a mean or something.

Now, your question regards the variability of the DV. It does seem plausible to me that the end position will be less variable if subjects start at a fixed position. However, this is entirely an empirical/theoretical question, not something you can derive through statistical means.

What will you do if the variance does turn out be different for both conditions? I'm not sure. While t-tests (assuming you plan on using something like that) require the variance to be constant, they are relatively resistant to violations of this assumption. Otherwise you could use tests that don't require this assumption.