The method of constant stimuli,  like Fechner's other psychophysical methods, can be used todetermine the least intense stimuli that can be perceived (i.e. the absolute threshold) or to determine the smallest difference between two stimuli that can be perceived (i.e. the difference threshold.)

The procedure for using the method of constant stimuli is relatively simple. First, a set of stimuli is selected. Some of these should be above the threshold, some should be below the threshold, but most should be very close to the threshold. The closer the majority of stimuli are to the threshold, the more precise the threshold estimate will be. Thus, it pays to spend a little time initially to get a rough idea of what the threshold is (often via the staircase method), and then select stimuli that are close to it.

Once the stimuli are selected, the trials can begin. The stimuli are presented to the observer one at a time in a random order. For each stimulus, the observer reports whether it was perceived (when determining the absolute threshold) or whether a difference was perceived (when determining the difference threshold). The psychophysicist records the physical intensity of the stimulus and whether it was perceived or not. This process is repeated hundreds of time. While it may be incredibly boring for the participant, the large quantity of data guarantees that the data are reliable.

Determining the Threshold

Determining the threshold from the data can be performed in two ways. One way is to create a plot of the data with stimulus intensity on the X axis and the percent or proportion perceived on the Y axis. When sufficient data is collected from motivated observers the resulting graph almost always takes the shape of an ogive, or S shaped curve. If the data do not approximate an ogive, then it may be very difficult to determine the threshold from the data.

Let's assume that seven stimuli were used. We will arbitrarily label their stimulus intensities as 1 through 7. For the seven stimuli, the psychophysicist has counted the number of times each was presented and the number of times each was perceived. These two numbers are transformed into the percentage of each stimuli perceived:
 
Stimulus Intensity
(arbitrary units)
Number of Times
the Stimulus Was Perceived
Number of Times
the Stimulus Was Presented
Percentage
Perceived
1
1
100
1%
2
5
100
5%
3
24
100
24%
4
50
100
50%
5
78
100
78%
6
93
100
93%
7
100
100
100%

This information is then plotted on a graph:
 
The next step in determining the threshold is to determine what the chance level of performance would be. In this case, the observer responded either that the stimulus (or the difference between two stimuli) was perceived or it was not perceived. Thus, there are two possible responses and chance would be 1 / 2 = 50%. Find the chance level of performance on the Y axis and draw a horizontal line at that value over to the ogive. Then draw a perpendicular line down from the ogive to the X axis. The threshold can be read from the X axis where this line meets its. This is demonstrated below, giving a threshold value of 4: