Imágenes de páginas

that sex differences might be studied. In the other tables the results have been collated for the entire number of pupils in a class irrespective of sex. The small number of pupils in the classes of the Foxchase School and the unequal distribution of boys and girls rendered it unprofitable to analyze their results in separate tables.

A brief reference to Table II (Appendix, p. 71) will serve to indicate the significance of the arrangement of all the tables. It will be seen that in the third b grade- the lowest grade tested-35 pupils were of the average age 9.9 years; that they wrote 2966 words, or an average of 84 words per pupil; that 94% of the words were spelled correctly; that of all the words written 4 were illegible, and that 25 nonsense words occurred. In the highest grade, the eighth school year, 70 pupils of the average age of 14.7 years wrote 15,211 words, or 217 words per pupil. Of these words 98.9% were spelled correctly, 10 were illegible, and 6 were nonsense words. Further examination of all the tables shows, on the whole, a gradual increase from grade to grade both in the number of words written and in the percentage of correctness.

Comparison of the results of all the tests shows a remarkably small variation in the average spelling accuracy from year to year. The average of the school percentages of Tables I, II, IV, and VI is 95.57%, with an average variation of .62%. If the three tests of the Northwest School be averaged the variation is still less. The averages are 95.9, Table II; 94.9, Table IV; 95.0, Table VI = average of 95.26%, average variation .42%.

The relatively large number of girls represented in the total of Table I (71 girls to 24 boys) may account for the variation of that result from the results of the other tables. Comparing that result with the averages for the girls only of the other tables, we have 96.5, Table I; 96.8, Table III; 96.1, Table V; 96.3, Table VII = average 96.42%, average variation .22% — a very close accordance of results.

The average number of words written, 158 and 152 respectively of Tables I and II, closely approximate. The second and

third tests with the pupils of the Northwest School give 173 and 179 as the average number of words. As was said above, a large majority of the pupils that took the first test participated also in the others. That the experiment was not a new one will account for the increase in the number of words written; for though the successive tests were taken without any reference being made to the preceding ones, the pupils probably tried to outdo their former efforts.

The relations above discussed which obtain between the average results of the several tests when the whole school is taken as the basis of calculation are paralleled by those disclosed when the several school grades are the units of comparison. The average variation is extremely small both for the number of words written and for the percentage of correctness. This is especially remarkable if Tables IV and VI or Tables V and VII be compared in detail. These tables represent results of tests of June, 1897, and June, 1898, taken, as before explained, under more nearly identical conditions than were the others.

In Table VIII (Appendix, p. 74), results of the three successive tests of the Northwest School are collated to show the average results of the three tests and their variations from the average. In Table IX the results of the three tests are treated in the same manner for the boys and girls separately. The median values of the three sets of results are used in these tables in preference to the arithmetical mean.

From Table VIII it may be seen, e.g., that eighth-year pupils wrote on the average 226 words; that the three tests varied. from this result on the average by 18.6 words; that 97.9% of the words were correctly spelled and that in three tests an average variation from this of .6% occurred. The average of all the tests for all the pupils was 173 words, the several tests varying on the average 9 words from this result, while the corresponding percentage of correctness was 95.0, with an average variation of .33%. There is a very regular increase from grade to grade in both rate of writing and accuracy of spelling,

but this is not paralleled by any such regular increase in steadiness of repeated performances of which the values of the average variations may be taken as measures. The lowest school grade, third b, which varied least in number of words (2.6 words), gave the largest variable error in spelling per cent (1.43%). But though no regular increase or decrease in steadiness of result may be seen from grade to grade,

[blocks in formation]

Average number of words: 777 boys, 79 to 245; 780 girls, 81 to 212.

the average variations for each grade are small. This would indicate that the results of successive tests are fairly constant quantities; the results of a subsequent test could therefore be predicted for any one grade or for the school within small limits of error.

Table IX (Appendix, p. 75) discloses some interesting sex differences. The boys wrote on the average more words (boys, 179; girls, 169) and also varied less in the successive trials than did the girls (average variation: boys, 7.6; girls, 10). On the other hand, the girls made fewer errors in spelling, whether the averages by grades or for the entire school be considered

(for the latter girls, 96.3% of correctness; boys, 93.8%), and varied less from test to test (average variation: girls, .23%; boys, .56%).

The accompanying curves present graphically some of the more important relations which are to be found in the above tabulations.

In Fig. 1, the half-yearly and yearly intervals between grades are laid off on the axis of abscissas, the number of words written on the axis of ordinates. (The curves show the general superiority of the boys in rate of movement as measured by number of words written, both in absolute number of words and in relative regularity of gain from grade to grade.

[blocks in formation]

Per cent spelled correctly: 777 boys, 89.6 to 97.3; 780 girls, 95.0 to 98.4.

In Fig. 2, the ordinates represent the per cent correctly spelled. The curves display the marked superiority of the girls in both accuracy of spelling and in regularity of gain from grade to grade.

In Fig. 3, the combined results of the three series of tests (see Table VIII) are given so as to show the rate of increase

in (a) the number of words written, (b) correctness of spelling along with (c) age. The last (age) shows the most regular increase, of course, but the others parallel it very closely, with the exception of a slight falling off in the work of the fifth-year pupils.

The method of plotting Fig. 3 was as follows: the halfyearly and yearly intervals between grades were laid off on the axis of abscissas. An ordinate (at 8) was selected to represent the difference between the minimum and maximum values of each item (age, number of words written, and per cent spelled correctly) to be recorded. The values of each item were then plotted within the limits of this ordinate, so that while the

[blocks in formation]

FIG. 3. Data for these curves in Table VIII. 1557 pupils, boys and girls.
Per cent correct, 93.0 to 97.9. Average number of words, 77 to 226.
Average age, 9.7 to 14.8 years.

curves begin together at the lowest grade tested and end together at the highest grade, they diverge at intermediate points as their values vary from grade to grade. The same method was followed in plotting the curves of Figs. 4 and 5.

Throughout the several grades the words chosen seem to be fairly representative of the working vocabularies of pupils of different ages. Quite difficult proper names and geographical terms were written, and words from their science and other lessons appear. This will be made sufficiently evident in the section which deals in detail with the spelling errors. The rise in the spelling curves therefore indicates not only an increased facility in spelling the shorter words of the lower grades, but also the ability to spell correctly more and more difficult words.

« AnteriorContinuar »