The data we collected seems to suggest a linear relationship between the energy input into the agitation of the dirt particles and the time it takes for those particles to resettle. However, the data is not clear-cut enough to come to any decisive conclusions. In fact, the data reflects the rather poor design of the experiment.
There are a significant number of areas where our procedure shows weaknesses. First of all, our method of agitating the dirt particles was rather crude. Upon striking the wall, the container wobbled erratically, which we did not control. Furthermore, in striking the wall with forward momentum, the particles on the container's bottom were kicked up in a forward or backward direction, and not equally in all directions. A better method may have been to drop the container in a controlled fashion from some height, letting it hit the ground or some other object, upon which it would stop immediately. This would have kicked up the sediment in all directions equally. In fact, any number of other methods would have been more precise than what we chose.
However, any such method would suffer from our next major weakness in procedure. We had no definite and repeatable criteria to judge when the dirt particles had resettled after having been agitated. We relied on subjective judgements, which essentially reduced to guessing. A possible solution would be to use a clear liquid as opposed to milk, and measure when the light from a known source passing through the mixture reached a given intensity. However, this would have added considerably to the complexity of the experiment, the simplicity and short duration of which were two of its most appealing qualities.
In the measurement of the angle, which we shouldn't have needed in the first place, the error was acceptable. On one experiment we extended the string to a right angle for the express purpose of testing our method. We calculated the angle to be 91 degrees, which is close enough that the error from this aspect of the experiment is insignificant compared to the rest.