Page:Popular Science Monthly Volume 67.djvu/546

540

Limitations and Sources of Error.—Before examining the results in detail, it is well to consider the limitations under which they were secured and the sources of error to which they may be thought liable.

Number.—It may be thought that the results lose something of their meaning because only about one blank in three (940 out of 2,700) was returned, but this was considered a very satisfactory showing, because, in the nature of the case, the filling out and return of the blank were not obligatory. Notwithstanding the definite assurances to the contrary contained in the announcement, many students feared that their records would, in some way, be open to the inspection of officials of the university, and therefore balked at making records of their doings in 'black and white,' while more still were either too busy or too lazy and too little interested to make out reports.

It is a familiar principle, though worth mention in this connection, that the habits of relatively homogeneous groups, such as all those under discussion, may be inferred with considerable certainty from data secured from a relatively small percentage of the members of the groups, provided the representation is typical. The more homogeneous the group, the smaller the data essential for safe induction.

That our data are typical is attested by two apparently paradoxical facts: (1) When the results are compiled on the basis of classes and courses, it is found that the averages for a given group are closely similar to those of other groups working under similar conditions, e. g., the freshmen and the sophomores in the College of Medicine give 19.72 and 19.30 hours per week, respectively, to outside study, 10.30 and 10.37 hours per week, respectively, to amusement, while the freshmen and sophomores in arts give 14.10 and 14.94 hours, respectively, to amusement. (2) In every group we note the presence of some extreme cases which vary widely from the average of the group. Thus, 41 male freshmen average 30.04 hours per week in outside study, yet one freshman records 57.00, and still another but nine hours. In a similar manner, we find that every group contains both the studious and those little given to study, both those who overindulge in athletics and those who scarcely take physical exercise at all, both leisurely and over-hasty eaters, etc., etc. In short, the reports are representative.