14 May 2009

Distribution of my reaction times

One of the many tests at cognitive fun! is the Visual Reaction Time test. It is a simple test: position the cursor within the four arrows close to the red dot and then click the mouse as soon as the large green dot appears. The aim is to click in less than 250 milliseconds, although I don’t know the significance of that particular number.

I did 100 sequential tries this morning, while writing down the time for each. I started the test assuming that the distribution of my times would be normal, but soon it dawned on me that my performance had a lower limit, which appeared to be 187 ms. After I had scored several of those, it became obvious that the distribution wasn’t going to be normal*. Here is the histogram for my 100 time scores.


The 4 longest times were the results of distractions; I had looked away for a moment or something like that.

Here is the histogram for all the values at or below 300 ms. It is a peculiarly skewed distribution. There is a cluster of very short times (187 and 188 ms), then a gap followed by another cluster and finally, a few long values.


My average for all 100 tries is 229 ms. If the 4 longest times (all above 297 ms) are deleted, the mean becomes 218 ms.

Give it a try and see how fast you can react to green.

*Theoretically, any and all normal distributions extend from negative to positive infinity, the only differences between the individual distributions being the mean and the standard deviation values.


Duane Smith said...

My lower limit was 173ms. I had several between 173 and 178ms. While I was not as diligent as you in writing the numbers down, I think my average was about the same as yours. I had a few long ones also. But I think there was a correlation between my reaction time and the length of time between the start of the previous trial and the when the green appeared in that previous trail. The closer the delay in trial n+1 to the delay in trial n the faster I was in trial n+1 or so I think. Shorter than the previous delay, I was surprised and not quite ready and longer than it, I began to be distracted wondering if it would ever come. Of course, I can't verify this but it sure seemed that way.

Kevin Bonham said...

My reaction time is generally slow. In this case my minimum was at 203 which I had scored about seven times out of 40, but then I got 188.

Anonymous said...

In answer to the comments, I just wanted to say a couple of things. Since this is not a matter of training for the Olympics, one's shortest reaction times are not that significant, and neither are one's longest reaction times; what counts of course is where the majority of one's reaction times are clustered.

As for the histogram of the distribution of times at or below 300 ms, I think the oddly disjunct distribution is probably an artifact due to the limitations of the measuring device. No doubt if the measurements were much more finely divided, and you did a thousand reactions rather than 100, you would get a curve rather than this disjunct spread.

I found that I got better with practice (not surprisingly) and that I did better when seated a little further back from the monitor than I often sit.


Susan J. Hewitt