Stimulus-Response Compatibility and the Simon Effect

In the figure below, you see a setup that can be used for a SRC experiment. There is a screen and a keyboard with just two buttons, one on the left and one on the right. When working with a regular PC keyboard, you can just choose a left and right positioned button. For example the left and right shift buttons, or the A and L button (because the A is on the left side of the keyboard, and the L is on the right side of the keyboard).

The fingers should be positioned on the buttons, so that they are ready to press one of the two buttons when needed.

Stimuli are presented on a screen. There are three stimuli:

Below you see two example trials. You need to read the three panels of each trial from left to right. First, the participant is just waiting and concentrating. Second, the participant sees a stimulus and needs to analyze it, that is, match it with a response. Third, the response must make a decision and respond. Typically, these three steps will take less than 4 seconds.

In both trials, the participant has to press the left button, because the stimulus says "left". In the example on the top, the experimental condition is compatible, because the stimulus position (left) corresponds to the response (left). But in bottom example, the experimental condition is incompatible, because the stimulus position (right) does not correspond to the word and the required response (left).

In this demonstration, you will see the words "left" or "right" appear on the screen as described above, and you need to press the A or L button. The word "left" always requires the A button; the word "right" always requires the L button. There are compatible and incompatible trials. More detailed instructions appear in the demo.

At the end of the 100 trials, you will see how fast you were in the compatible and incompatible conditions.

When looking at the data, it is clear that it is much easier to respond to compatible than to incompatible stimuli. But why exactly is this? Could it be that the left hand responds more quickly to all stimuli that appear left, and the right hand to all stimuli that appear right? If that is the case, one could argue that there is a direct anatomical explanation underlying SRC. You can test whether or not this is the case yourself, just by crossing your arms. Your hypothesis is that SRC is caused by an anatomical link between stimuli and responses, whereby left stimuli are responded to quickest by the left hand and right stimuli by the right hand. Your prediction is that the left hand would still respond quickest to left-positioned stimuli. If it is only the relative location of the hand that matters (e.g., the left hand resting on the right side of the keyboard, as is the case with crossed arms), then the crossing of the arms should make no difference at all. It should be easy enough to test this yourself!

The following part of this lesson is only for advanced use. Not difficult, yet not necessary for understanding the basics.

You can download all the files and run the trials of the experiment. Then, make sure you set the data file to the name of your data file and analyze. You will likely find that you respond on average in less than 500 ms, and that you are around 30 ms faster in response compatible than in response incompatible trials

In the figure below, you see an analysis of the example data with 200 trials. In the left panel, you see the average response time and standard errors for the four conditions of this experiment. In the right panel, you see the same for the error rates. You see that the red lines (incompatible data) are higher for both the response times and the error data, meaning that the participant was slower and less accurate in this condition. You also see that the participant performed different with the left and right hand.

Apart from these averages and standard errors, you might have other questions about performance. For example, it could have been the case that the participant was very motivated in the first half of the experiment, but then got tired and made many mistakes. This sort of problem is difficult to see if you just look at the averages of conditions. What you really should do when analyzing the data of any experiment is to plot all the data points.

Below, you see the response times of all trials in the order the trials were performed. Red dots indicate trials on which an error was made. The black horizontal line indicates the grant average, and the blue line a local average (using a Lowess filter).

These data are now not split up by condition. The focus is now to see whether the participant did not suddenly change his or her speed or level of accuracy. You can see that RTs go up and down quite a lot, but that there are no dramatic changes over time. This is normal random variation, common to any biological variable.