A Pre-School Character Rating Chart 

Author:: K. M. Banham Bridges :affiliation: Psychologist, McGill University Nursery School

The educational function of the nursery school differs considerably from that of the primary or secondary school. Nursery school education is concerned mainly with the development of character, of social behaviour and of emotional adjustment and only to a lesser degree with the imparting of information and training in manual and linguistic skills. In schools for older children the educational aims are the same but the emphasis is reversed and more attention is given to the development of intellect and skill than to the development of character and social behavior. There is need in both kinds of schools for performance, language, general intelligence and school achievement tests, for the purpose of comparing the abilities of different children and of checking their school progress. But the nursery school wants in addition to performance and language tests some methods for comparing the social behavior, the emotional adjustment and the character development of its pupils. For the past year the writer has been trying out a scheme for measuring roughly the social and emotional development of pre-school children in the McGill University Nursery School. A report of this work will be published at a later date. In addition to this scheme a short rating chart was devised primarily for the use of the nursery school staff in order to help them estimate the children’s characters and follow their character development. This rating chart is described in the following paragraphs together with a few results obtained from its application in the McGill University Nursery School.

The chart was made at the suggestion of the senior teacher, Miss GA Watkins, and was intended as a practical device to serve temporarily in the absence of more exact methods for measuring the components of character. Throughout the period of experimentation with this character rating chart, Miss Watkins cooperated closely, offering suggestions and criticisms based on her long experience with pre-school children thus taking a considerable part in the devising of the chart.

Method of Construction of the Chart 

Since the aim in making the character chart was to create something which would be practical and useful to nursery school teachers who had had little or no training in psychology, there were three facts necessary to bear in mind. First that the chart would need to utilize expressions that would be readily understood by the teachers and would mean the same to all. Secondly, that persons unused to making scientific observations of people tend to form and give opinions rather than to see and describe facts of behavior; and thirdly that the statements of these persons are usually colored by moral and social evaluation. Parents and teachers not trained in scientific methods can scarcely be relied upon to make unbiased statements or judgments with regard to children’s behavior.

In order to meet the above difficulties it was decided to make a rating chart which would utilize the normal tendencies, to give opinions and to evaluate, rather than have them as interfering factors in a more scientifically exact measuring scale. To offset the bias of one individual it was decided that the children should be rated by at least three or four raters who were frequently in contact with the children.1 The teachers’ own wording was used in the chart, as it was found by actual experiment that there was less difference in the meanings they ascribed to the more common though vague terms of popular speech than to the more definite but less familiar terms of the psychologist.

Lists of adjectives and statements describing children’s character and behavior were drawn up by the writer and Miss Watkins. From these items were selected all those which would tell something it was desirable to know about a child’s character, apart from intelligence and skill. Although some of the items indicated aspects of temperament, the more general term character was considered best as a heading for the list. Among the items were adjectives qualifying behavior, such as “affectionate,” “rough,” “trustworthy,” and statements about behavior such as “plays alone,” “shows off intrusively,” which could not be put satisfactorily into adjective form. Synonyms and statements describing similar aspects of character were eliminated, also phrases describing character types rather than traits. Long phrases were shortened so as to fit into a compact form or chart. Finally, fifty concise items were selected which described different aspects of a child’s character of which the teachers desired to have some estimate. These were arranged as alternative statements or characteristics in paired opposites, the more socially desirable characteristics being placed first. The paired items were grouped roughly under three headings, social, individual (which was later changed to personal) and emotional. The first half of the social group contained items referring to a child’s relation with other children and the second half contained items referring to his relation with adults.

Twenty duplicate charts were typed, one for each child in the McGill University Nursery School. The members of the school staff who came into frequent contact with the children then rated each child on the chart according to directions which are given below. No standard of comparison was suggested, neither child to child nor group order-of-merit. The raters were simply required to mark a stroke for each item where the first alternative applied more nearly to the child, a zero where the second alternative applied better, and both a stroke and a zero where both items applied equally well. Dashes were to be placed against doubtful items.

By this means the teachers could see roughly at a glance the distribution of strokes and zeros on the chart apart from numerical scores. Scores, however, were obtained by allowing a point for each stroke, half a point for each stroke and zero, and nothing for a zero. Percentage scores based on the number of items marked (i.e., omitting all doubtful items marked ?) were calculated and the average of the percentage scores for each rater was found. This average was taken as the Character Rating Score for each child. After a few trials it was found necessary to omit from the average the score of any rater who had marked ? against ten or more items, as such scores were too unreliable and were found to vary considerably from the scores of other raters.

After several ratings had been made at intervals of one or two months conferences were held with the staff and the chart was revised. Some of the items were eliminated as being too ambiguous and difficult to rate, some were re-phrased to make them more definite and one or two new ones were added. Wherever there had been considerable variation in scoring among the staff the items were carefully checked in order to detect ambiguities in meaning. The social desirability of the alternative characteristics in each item was considered by the staff, and the items were again arranged so that, in the opinion of all, the first alternative represented the more desirable characteristic. Tendencies of the staff to be biased in their ratings of different children were noticed, but when the estimates of all four raters were pooled these tendencies were seen to be counterbalanced to some extent. The revised form was used again in the Nursery School, ratings being made on the children at intervals of one or two months.

Below is a sample of the Character Rating Chart in its final form together with the directions for rating and scoring the chart to be given to each rater. The columns at the side of the chart are used for markings by four raters, more may be added if desired.

Directions for Rating 

Which characteristic predominates? From your personal observation, not from someone else’s report, mark (1) if the left hand and (0) if the right hand characteristic applies more nearly in each case. Mark both (1) and (0) if both apply equally well, and mark (-) if neither is applicable. Mark (-) also against the item where no observation has been made. Do not look at the marks of other raters. Cover them with a strip of paper before marking each chart.

Directions for Scoring 

Score one point for each (1) and half a point for each (1 0) marked by a single rater and add the total number of points, disregarding an odd half point.

Find the percentage score by dividing the total number of points (multiplied by 100) by fifty minus the number of items marked (-). -r, .

Total score (items marked (1) X 100
Percentage score =
Items omitted (marked (?))

The Character Rating Score for each child is the average of the percentage scores of all rates. If any rater has marked 10 or more items (?), his or her score is not included in the average.

McGill Pre-School Character Rating Chart 

Name:
Date:

Personal 

Independent or dependent
Self-confident or not self-confident
Energetic or lethargic
Quiet or noisy
Quick in action or slow in action
Persistent or gives up easily
Careful or careless
Purposeful or aimless
Concentrates or is easily distracted
Questioning or incurious
Not destructive or destructive
Original or not original in play
Imaginative or unimaginative

Emotional 

Cheerful or grave
Not variable or variable in mood
Spontaneous or restrained
Not self-conscious or self-conscious
Not nervous or nervous
Not deceptive or deceptive
Not excitable or excitable
Controls tears or cries easily
Venturesome or timid
Little or much disturbed by observation..
Deliberative or impulsive
Does not show temper or shows temper
Does not sulk or sulks
Patient or impatient
Forgiving or revengeful

Score Average % score Number scored % score

Results 

All the children in the McGill Nursery School were rated on the chart seven times between March, 1927, and April, 1928, by four raters: the teachers in charge of each group, the psychologist and the Director’s assistant. No child was rated who had been in the school less than a month. There are twenty children in the school, nine girls and eleven boys, divided into two groups according to age. The seven or eight children between two and a half and three and a half years old play or work together supervised by an assistant teacher, and the remaining twelve or thirteen children between three and a half and five years old play or work in another room supervised by Miss Watkins. These groups are referred to below as Group A and Group B, respectively. For half of each morning the children of both groups play together outdoors. In March, May, and October, 1927, the first form of the chart was used, and in December, 1927, January, February and April, 1928, the final chart, as here presented, was used. As the first form varied only slightly from the second a summary of the results obtained from the use of both forms is given below. In the tables the results obtained when the first form was used are marked I, and when the second form was used they are marked II. All scores here presented are per cent scores brought to the nearest whole number.

The scores of individual children varied considerably, ranging between 31 and 94 per cent, with the majority falling between 60 and 80 per cent. The average scores for each school group and for the whole school are shown below in Table I. There were

- Date
- 1. March, 1927..
- May, 1927…
- October, 1927.
- Group A Av.
- 62
- 64
- 68
- Group B Av.
- 65
- 73
- 77
- School av
- 69
- 73
- 69
- Average…
- 72
- 68
- 70

II. December, 1927. January, 1928.. February, 1928. April, 1928 Average… 66 64 69 68 67 71 65 71 74 70 69 64 70 72 69 I and II Average. 66 71 69

equally high and equally low scores in the two school groups though the majority of scores in Group A were lower than most of the scores in Group B. This difference is shown in the average scores where Group A average is slightly lower than Group B average. The scores may not be read directly down the columns to see the effect of age and school attendance on the scores as different children were in the groups when different ratings were made. Some of the monthly ratings, however, may be compared. Most of the changes, admittance of new children in the place of those leaving and promotions from Group A to Group B, occurred in October and January at the beginning of the school terms. Thus the scores for March and May, for October and December, and for January, February and April may be compared together as they represent ratings for the same children. From these comparisons it will be seen that there is a small increase in the scores with age and school attendance. The slight drop in the scores from May to October and December to January is probably due to the above mentioned factors, admittance of new children and promotions from Group A to Group B.

Age Differences 

There was a noticeable correspondence between Character Rating Score and age. Taking the results of all seven ratings, between March, 1927, and April, 1928, together, the average score for children between two years and four months and two years eleven months was 60 per cent. The average score for three year olds, between three years and three years eleven months, was sixtydght; and the average for four year olds, between four and five years, was 76 per cent. This age difference was noticeable even m half year stages as may be seen from Table II.

The scores should be read across the columns and not down them, for in many instances scores in the same age group for the different months represent ratings on different children. The age differences though fairly consistent are, however, small. This is what might be expected, since the children were rated with regard to temperamental and character traits which change relatively slowly as compared with the development of intelligence and skills. One would expect that character ratings would not correlate so highly with age as do intelligence test results. The scores on the Character Rating Chart indicate individual differences rather than age differences. It will be noticed, however, that the average scores for the

Table II. Average Scores in Half-Year Age Groups. Month 2 to 2-5 years 2-6 to 2-11 years 3 to 3-5 years 3-6 to 3-11 years 4 to 4-5 years 4-6 to 5 years I. March, 1927.. May, 1927…. Oct., 1927 Average 54 59 71 61 71 65 64 67 68 71 57 65 72 74 75 74 80 82 71 78 II. Dec., 1927 Jan., 1928 Feb., 1928 Apr., 1928 Average 52 59 56 65 55 62 63 61 75 74 79 70 75 55 59 65 70 62 72 82 71 75 77 77 78 82 79 I and II Average. . 56 61 71 64 74 78

children between three and three and a half years old are higher than the scores for children between three and a half and four years of age. This may be accounted for perhaps in the following ways. It may be in part a result of the school’s being divided into two age groups as previously mentioned. The raters then tended to make comparative judgments with regard to each item in the chart, comparing each child with the others in his school group. Thus a child of three years and five months about to graduate from Group A would be scored high as a result of comparison with the younger children, whereas a child only a month older in Group B would be scored low as a result of comparison with older children in the group.

Moreover, the problem of adjustment is harder for the younger child in the group than for the older child. One kind of character trait may be brought out when a child has to compete with others more able than himself, while quite another set of traits may be manifested when the child knows more and is more able than others in his group. A reason for the specially large difference between scores of these two age groups obtained December, 1927, to April, 1928, is probably the fact that there were two unusually well adjusted children in the three to three and a half year group during this period, while there were two very poorly adjusted children in the three and a half to four year group. Since there was only an average of four children in a half-year age group an extreme in any one child’s score was bound to make a noticeable effect on the group average.

Effect of Promotion in School 

The average of the scores of the eight children who were promoted from Group A to Group B during the year was found for the months before and after promotion. Before promotion the average score was 69 per cent and after promotion the average score of the same children was 66 per cent. This slight drop in the score may be accounted for in the same way as the drop in the average score of the three and a half year age group.

Effect of Length of Attendance at School 

The scores of the children in Group A and B obtained from all the ratings were tabulated according to the number of months since the admission of each child to the school, regardless of the age of the child. The average score was then found for children who had entered school only a month ago, for those who had entered two months ago, for those who had entered three months ago and so on up to fifteen months. In this way it was hoped to show roughly the effect on character rating scores of length of attendance at school, although no allowance was made for absences or vacation periods. The age factor was eliminated in Group A as children were admitted to this group at all ages varying from two years and four months to three years and five months. In Group B the age factor was not quite eliminated. The new; children admitted to this group were mostly under four but their number was almost balanced by those who had graduated from Group A at three and a half years and who had been in the school several months. In addition to these was an equal number of older children between four and five years of age who had been m the school a year or more.

The results for Group B showed an increase in score after six months’ attendance at school. But the results for Group A showed no increase in score with length of school attendance. Even the increase in score for Group B may be accounted for, in whole or in part, by the fact that the children who had been in the school longest were for the most part older than those more recently admitted. This failure of the scores to show increase with time spent at school is probably due to the fact that differences in score due to school attendance were entirely obscured by the large individual differences in scores. As previously mentioned, the scores in Table I show a slight increase due to the combined effect of school attendance and monthly increase in age. Table III below shows the average scores arranged according to school attendance in three month intervals.

- Months Since Admission to School
- 1-3 mo.
- 4-6 mo.
- 7-9 mo.
- 10-12 mo.
- 13-15 mo.
- Group A
- 68
- 67
- 67
- 67
- 68
- Group B
- 74
- 66
- 80
- 67
- 76

Comparison With Intelligence Test Results 

The character rating scores obtained in May and in December, 1927, were compared with the Stanford-Binet Intelligence Quotients of the children and also with their median scores on the Stutsman performance tests. The character rating scores were found to increase somewhat with Intelligence Quotients but they showed practically no relation to performance test scores. The coefficient of correlation, by the rank difference method, between the character scores and the Stanford-Binet Intelligence Quotients in May was + -39 and in December, -f- .67. The coefficient of correlation between character scores and median performance scores in May was .123 and in December -f- .003.

The reason for the correlation of character ratings with Stanford-Binet Intelligence Quotients will probably be apparent to those who have worked with the Binet tests at the lower age levels. In order to do his best on a Binet test a child must be fairly well adjusted socially and emotionally. He must cooperate with the examiner, do what he is bid and answer questions. These tests, therefore, measure to some extent temperamental and character traits such as are scored on the Character Rating Chart. Since in this rating chart intelligence factors are mostly eliminated, the scores obtained from its use might prove to be a helpful check on intelligence test results. A low score on the Binet tests might be shown to be due to lack of social adjustment rather than to lack of intelligence. It is obvious, however, that both intelligence tests and character ratings are measuring in part the same things, for character and intelligence can never be entirely separated. The performance test situation is somewhat different from the Binet test situation. The child is freer to do as he chooses with the material. The tasks that are set for him to do are more in accordance with his spontaneous interests, and he is not required to repeat so many words. A timid or negativistic child might do fairly well on the performance tests and yet be scored low on the Binet tests and the Character Rating Chart.

Comparison of Sexes 

The average scores of the girls and the boys in Groups A and B were found for all seven ratings. These are presented in Table IV. As may be seen, the scores show almost a negligible difference in favor of the girls. This difference is probably so slight as to be due entirely to chance or individual differences in such a small number of cases.

Table IV. Average Character Scores According to Sex Girls Boys Group A, Form I Form II Average 66 62 69 67 66 Group B, Form I Form II Average 72 71 72 67 72 70

Comparison With Order of Merit Rating by the Staff 

In the absence of any means of testing the reliability of the Character Rating Chart, the scores for March, 1927, and January, 1928, were compared with order-of-merit ratings of the children by the staff. A week before they scored the children on the charts each member of the staff was required to arrange the children in Group A and Group B in order of merit with respect to general character, including social, personal and emotional characteristics. Considerable agreement was shown by the staff in these ratings. By allotting marks according to rank a final order of merit list was drawn up which comprised the pooled ratings of the four members of the staff. This list was compared with the rank order of the children obtained later according to the character rating scores. In both March and January there was very close agreement between these lists. In January the coefficient of correlation, by the rank difference method, for Group A was -f- -99 and for Group B -{- .98. These results indicate very strongly that in marking the children on the Character Rating Chart the raters were wittingly or unwittingly comparing each child with the others in his group in respect to each item, and in this way they arrived at an order of merit rating of the children.

The ratings of the different members of the staff agreed fairly closely. The average coefficient of correlation, according to the rank difference method, between the ratings of different raters throughout the year was -j- .78. More agreement was shown towards the end of the school term than at the beginning. This was probably due to the fact that by the end of the term the raters had had more time to observe the children and therefore knew them better.

Conclusions 

The Character Rating Chart described in this article has shown its usefulness in roughly differentiating pre-school children in a group with regard to general character. The staff of the McGill Nursery School who tried out this chart agreed that it had helped them to clarify and regulate their opinions with regard to the children. It also showed them more clearly than they could guess unaided where the stronger and the weaker points in a child’s character lay; in his social behavior, in his personal characteristics, or in his emotional control.

The numerical scores obtained by the McGill Nursery School staff could be treated in no way as exact measurements, as they were determined by the personal opinions of raters who were comparing each child with his associates. As the children in the school change then the basis for comparison for each child changes. No fair comparisons, therefore, can be made between the results obtained by different raters for children in different schools. The results from the McGill Nursery School are quoted here as an example of what may be done with the chart. In addition the chart might be of use in pre-school clinics. Of course, in all cases in clinics or in nursery schools, care would need to be taken to secure reliable raters who had had the child to be rated under observation for some time. The limit of a month’s observation by the McGill Nursery School staff was found to be fairly satisfactory, though results were more reliable after a longer period.

Disclaimer

The historical material in this project falls into one of three categories for clearances and permissions:

Material currently under copyright, made available with a Creative Commons license chosen by the publisher.

Material that is in the public domain

Material identified by the Welcome Trust as an Orphan Work, made available with a Creative Commons Attribution-NonCommercial 4.0 International License.

While we are in the process of adding metadata to the articles, please check the article at its original source for specific copyrights.

See https://www.ncbi.nlm.nih.gov/pmc/about/scanning/