I. With respect to school work itself, what relation exists between the early success in elementary subjects and the later success in handling more advanced subject matter? This question is important to all those who may be concerned in advising individuals concerning the desirability and probable profit of continuing their school experience, and of entering occupations in which scholastic abilities may be requisite.
Kelley has recently reported a careful study of the relation between the marks in the fourth, fifth, sixth and seventh grades and the marks received in the first year of high school work. The results, in the case of fifty-nine pupils followed through the six years, were as follows:
Correlation between Marks in the Grades and Marks in First High School Year
| 7th grade | .72 |
| 6th grade | .73 |
| 5th grade | .53 |
| 4th grade | .62 |
His study further seeks to show the relative weight to be attributed to the work of each grade, by applying a formula known in statistics as a "regression equation." He says, "The net conclusion which may be drawn from these coefficients of correlation is that it is possible to estimate a person's general ability in the first year [H. S.] class from the marks he has received in the last four years of elementary school with accuracy represented by a coefficient of correlation of .789, and that individual idiosyncrasies may be estimated, in the case of mathematics and English, with an accuracy represented by a coefficient of correlation of .515.... Indeed, it seems that an estimate of a pupil's ability to carry high school work when the pupil is in the fourth grade may be nearly as accurate as a judgment given when the pupil is in the seventh grade."
Miles finds that the correlation between the average elementary school grade and the high school grade is .71. Dearborn also finds that high school efficiency is closely correlated with success in university work. He studied various groups of high school students, the groups containing from ninety-two to four hundred and seventy-two students each. These were grouped into quartiles on the basis of high school standing, and compared with similar classifications on the basis of university work. Dearborn summarizes his results in the following words:
"We may say then, on the basis of the results secured in this group (472 pupils) which is sufficiently large to be representative, that if a pupil has stood in the first quarter of a large class through high school the chances are four out of five that he will not fall below the first half of his class in the university.... The chances are but about one in five that the student who has done poorly in high school—who has been in the lowest quarter of his class—will rise above the median or average of the freshman class at the university, and the chances that he will prove a superior student at the university are very slim indeed.... The Pearson coefficient of correlation of the standings in the high schools and in the freshman year, for this group of 472 pupils, is .80.... A little over 80 per cent of those who were found in the lowest or the highest quarter of the group in high school are found in their respective halves of the group throughout the university.... Three-fourths of the students who enter the university from these high schools will maintain throughout the university approximately the same rank which they held in high school."
Lowell's investigation, which is discussed in later paragraphs, also bears directly on the question of the relation between college entrance records, college grades, and later work in professional schools. A rather different method of procedure was adopted by Van Denburg, who studied the relation between the first-term marks of high school pupils in New York City and the length of time the pupils continued in school work. The following table gives a general idea of his results: