THE CORRELATION OF SUCCESS IN EXTRA-CURRICULAR

ACTIVITIES AND SUCCESS IN SCHOLARSHIP.

ROBERTA E. YANCEY 1

Howard University.

A question of persistent interest not only to educators and students but to society in general is whether or not any relationship exists between the success of students in those things not generally considered a part of the formal course of study on the one hand and success in scholarship on the other. Are the students who are most active in debating, athletics, school clubs and other organizations also the most active in the work of the class-room? Or, more specifically, is it generally the case that the student who wins the admiration of his mates through his successful participation in extra-curricular activities as indicated by the offices and honors which he receives at their hands also wins the admiration of the faculty as indicated by the marks which make up his scholastic record?

The object of the study reported here is to gather the pertinent facts in an effort to give at least a tentative answer to this question so far as it relates to the college students in Howard University during the past decade. To this end it was decided to make an investigation of the records of students both in extra-curricular activities and in academic scholarship for the ten-year period from 1914 to 1923 inclusive. The former were taken from the class annuals and the latter from the records in the office of the Registrar. The publication of class annuals was suspended during the two war years of 1917 and 1918 so that these years were eliminated, leaving for consideration the remaining eight years of the decade period.

Twenty names were chosen at random from each of the eight classes this number being considered a fair sampling of each class group, constituting, in some cases, as much as fifty per cent of the total class enrollment at graduation. In addition, this number is convenient in view of the method employed. For the purpose in view it seemed advisable to use the Method of Rank Differences for computing the coefficient of correlation. The object sought was an answer to the question "Do students ranking high in one line rank high in the other?" The problem, consequently, is essentially one of rank or relative standing. The

1 Undergraduate student School of Eudcation, Howard University. This investigation was conducted under the supervision of Dean D. O. W. Holmes.

method adopted, therefore, is simple and at the same time appropriate.

In order to make statistical use of the information given in the class annuals it was necessary, first of all, to devise a scale of weights or points based upon the various items of student activity so that the latter might be transformed into numerical scores. It was realized at the outset that such a scale must be arbitrary at best since no authoritative determination of the relative values of such activities has been made. The difficulty is apparent at once. While everyone will agree, for example, that the captaincy of a team is an achievement deserving a higher score than mere membership on the same team, there will doubtless be considerable difference of opinion concerning the exact numerical ratio that will express that difference. Again there is little basis for the just comparison, in figures, of the success represented by election to the captaincy of the football team, for example, and that represented by election to the editorship of the college paper, or to the presidency of the senior class.

These examples make clear the difficulty and at the same time indicate the chief point of weakness in any such study. In the absence of anything better, however, an arbitrary scale had to be created. After a study of such scales in use at Cornell University and the University of Pittsburgh and an informal canvass of campus opinion among faculty and students at Howard University, the scale shown in the accompanying table was adopted as the basis for computing the individual scores for extra-curricular activities.

Table of Weights for EXTRA-CURRICULAR Activities

For brevity the thirty-three descriptive items found in the several year books are indicated in the table by letters with significance as follows:

In this scale the attempt is made to give relative values to all the various items of achievement as recorded in the class annuals except

mere membership in organizations where no selection by students was involved. Since the rank method was to be employed it made little difference how much one activity was rated above another so long as the most important activity was rated first, the next in importance second,

etc.

Examples of the application of the scale to two cases will illustrate the method. In the class annual for 1916 student D is described as follows: President of Athletic Association; Assistant Manager of football; President of Sophomore Class; member of Y. M. C. A. Cabinet. According to the scale the score was computed as follows: 15+ 2 +13+10=40. In the same annual, student E is given this description: President, Deutsche Verein; Vice President of the N. A. A. C. P.; Class Critic in 1913; President of the class of 1915; Vice President of the class of 1915. Applying the scale in this case we have: 10+5+1 +13+2=31. Making allowances for the limitations of the method as already noted it seems fair to assume that the scores thus computed yield a reasonably reliable ranking of the students in those lines of college activity lying outside the work of the classroom.

As stated above the scores for the ranking in scholarship were taken directly from the records in the office of the Registrar. In order that these records might be fairly comparable, only those students whose entire college life was spent at Howard University were included in the list. The passing grades of each student for his entire course in college were averaged and the resulting number taken as the score. Each group of twenty was then ranked in accordance with these scores. Prior to July, 1919, grades were given in numbers. Since that time the letter system has been used. For the purpose of this study these grades were given numerical equivalents as follows: A=95; B=85; C = 75; D = 65. This is in accordance with the practice of the Registrar's office when, for any reason, the numerical equivalents of students' grades are desired.

The eight tables which follow show the scores and the ranks both for activities and scholarship and the computation by Spearman's rank method. In each table the individual scores in extra-curricular activities of the twenty students of a class were tabulated in descending order, and the ranks indicated by the numbers 1, 2, 3, etc. For obvious reasons letters of the alphabet were employed in their usual order instead of the individual names of the students. After this had been done the individual scores in scholarship and the corresponding ranks were appropriately filled in. An example will make the method clear.

Referring to Table I (1914), it was found that the highest score in activities is 80, the next 45, the next 34, and so on as indicated in

the column headed "Activities." The student making the score of 80 is designated as A in the first column and his rank is given as 1 in the third column. In like manner the student making a score of 45 in activities is designated as B and his rank is given as 2. This procedure is followed through the entire set of scores for activities.

It should be noted that when we come to the rank of 7 we find two students with the score of 10. Since these two rank equally, but together must occupy the seventh and eighth places, each is assigned a rank, which is the average of 7 and 8, or 7.5. Since these two take the seventh and eighth positions, the next rank recorded is 9. The same procedure is followed in the same array of figures at ranks 10 and 11; 12, 13 and 14; 15 and 16; 17 and 18, and 19 and 20. This is the common and approved procedure.

The scholarhsip scores are now filled in opposite the appropriate names in column 4 and the ranks in scholarship assigned in column 5. Had the correlation been positive and perfect the ranks in the array of scholarship scores would correspond exactly with the ranks in the activity scores. That is to say, that the student standing highest in activities would also stand highest in scholarship, the second in one array would be second in both and so on throughout the list with the poorest student in scholarship having the lowest score in activities.

On the other hand, if number 1 in activities ranks number 20 in scholarship; number 2 in activities ranks number 19 in scholarship; number 3 in activities ranks number 18 in scholarship and so on, then we would have perfect negative correlation. Mathematically the coefficient of correlation ranges from +1 to-1, the former indicating perfect positive correlation and the latter perfect negative correlation. Between the two is 0 (zero) which means that there is no correspondence at all, either directly or inversely.

Continuing the illustrations from Table I (1914) we pass now to the computation. Theoretically the best method of computing the coefficient of correlation is the product-moment method of Pearson which takes into account not only the ranks but the amount of variation in the scores. For the present purpose we are not interested in the values as represented by the individual scores since these values have served their purposes with the determination of the ranks. The unreliability of the values used in the activity scores emphasizes this point.