+ The Experiment:

- We seek to determine how good folks are at distinguishing between different masses. Accordingly, we set out 5 - 10 masses in the 50 - 150 g range and a known or reference mass of 100 grams. We label the rocks in random order, 1, 2, 3, . . . , 9 say. (Do not label them in increasing mass order!) And we identify the known mass and its actual mass for the participants.

- We ask participants to indicate whether the sample rocks are of greater mass than the known mass. We provide a slip of paper in which we could ask them some information for later references, e.g., their major, their position (student - undergraduate or graduate, faculty and discipline, administrative, staff, etc.) Do not ask for names unless, say, you wish to have a contest for "best" guesses.

The real information we want is their opinion on whether the mass of the rocks, numbered 1, 2, . . . , 9, are more than the sample/known mass. So we provide a form such as the following:

Sample Rock Number 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9
Check if mass is more | | | | | | | |
than 100 g

- We set the collection table out for whatever time we believe reasonable to get sample data - one week will be sufficient especially if there is a sign, say with words like "Rock-A-Rama," "Let's Rock," "Science Experiment," or "Which Major is More Perceptive," and some catchy graphics.

- Once we have collected the data, we compute the percent of respondents who say that a given rock has mass greater than the 100 g standard mass. We plot this percent vs. the actual mass of the given rock.

- If participants had perfect knowledge, e.g. a scale - which we do not provide(!), the plot of the returned data would look like this:

- If all or most participants just guessed then the plot of the returned data could look something like this:

- A reasonable population with 9 sample rocks 10, 40, 70, 90, 100, 120, 125, 150, 175 g might yield the following plot.