Pitch scaling does not quite conform to the power law as predicted by Stevens; the exponent of the power law differs across frequency ranges. [The line would be straight if a doubled frequency was perceived as a doubled pitch.]
This curve shows the data points collected by Stevens in his first three experiments, where subjects were asked to divide a musical range into four equal ranges. [The dots are the data points; the different colors reflect the three different ranges which he tested.] The red "+"s in the graph show the anticipated data for the experiment you will be doing, though there is individual variation. as the pitch doubled, the line would be straight.
In this experiment, you will be given a set of "keys", with the first one set at a certain tone. You are to adjust the frequency of the second key [using the "<" and ">" keys] until its pitch sounds twice as high as the first. When satisfied with that key, click "Next" to shift to the next higher key, which will be preset in the neighborhood of the expected doubling. If you decide a previous key should be adjusted, use "Previous" to back up to that point and make the change, then continue forward [a key at a time] as before.