Printer Friendly

Kindergarten screening predictive inaccuracy: first-grade teacher variability.

ABSTRACT: Screening kindergarten students to predict first-grade achievement is characterized by rather gross errors of both over identification and underidentification. This lack of predictive validity may be the consequence of overly simplistic conceptualizations that fail to take into account differences in first-grade teachers. This study attempted to evaluate the extent to which first-grade teachers differ in their preferences, requirements, and expectancies of students. Twenty-one teachers ranked 86 student descriptors on a continuum ranging from "absolutely contributes to student success" to "absolutely contributes to student failure." Results suggest that teachers vary considerably in the way they rank student descriptors and that these variations may be a factor in the predictive inaccuracy of kindergarten screening.

Screening of kindergarten populations to identify potential first-grade learning problems routinely occurs in most school systems (Lindsay & Wedell, 1982). However optimistic and worthwhile the goals of such screening may be, its predictive accuracy has not been overwhelming (Meisels, 1987). A critical preliminary step toward improving predictive validity would be to hypothesize and explore possible reasons for this inaccuracy.

Although a host of factors and forces interact as young children learn and develop (Fedoruk, 1989), idosyncratic teacher preferences, biases, and expectations may be particularly influential. The teacher has frequently been singled out as an extremely important classroom variable related to student achievement (Dembo & Gibson, 1985; Kauffman, Pullen, & Akers, 1986). The teacher directs instructional activities, selects materials, interprets curriculum, and ultimately evaluates student performance. He or she provides feedback, encourages and supports, disciplines, reinforces, monitors, corrects, and assists. And as Simpson and Galbo (1986) suggest, "The interaction comprising the student/teacher relationship is the primary instrument for school learning" (p. 37).

Teacher attitudes, expectations, biases, and beliefs are thought to influence student achievement (Brophy, 1983; Brophy & Evertson, 1981). However, although "most teachers would agree that some type of expectation, positive or negative, is placed upon every student in every classroom, how one arrives at these expectations varies as much as the effect it has upon the student (Ouzts, 1986, p. 133). As Kedar-Voivodas (1983) stated:

As the single authority figure in the classroom, the elementary schoolteacher determines primarily how much tolerance can exist for deviation from these expectations, what type of deviance can be tolerated, and what the consequences will be for such deviations. (p. 419)

In establishing predictive strength of student cognitive, behavioral, and social variables, however, the conceptual premise has consistently been that student characteristics occur in an environmental vacuum. The assumption appears to be that student characteristics have predictive power regardless of environmental variations. But is it possible that first-grade teachers require, prefer, tolerate, and react differently to student attributes.

Could it be that a certain student would survive in one first-grade classroom while floundering in another? Thelen 1967)suggested, "Everybody seems to realize that some pupils perform better with certain teachers than with others" (p. 18). Thus, a critical first step toward developing more accurate predictive indexes of early school and learning problems may be to examine the extent to which teachers vary in their expectations of students, as well as their tolerance, requirements, and biases. Rather than continuing to focus on specific pupil characteristics that correlate with subsequent school performance, we may need to redirect research focus and modify conceptual assumptions. Abundant previous research has identified student attributes that correlate with academic performance. This study attempts to determine the meaning and relevance of those attributes for first-grade teachers. One explanation for predictive errors of kindergarten screening may rest in the differential nature with which first-grade teachers react to, evaluate, and interpret predictive student characteristics.



A total of 21 first-grade teachers (20 female, 1 male), ranging in age from 29 to 50 years (mean 39.0), were randomly selected from six elementary schools in a large Canadian city. All six schools housed students in Grades K-6. Located in four different areas of the city, the schools included a broad spectrum of middle-class students. The teachers ranged in total years of teaching experience from 6 to 27 (mean 15.8) and in terms of first-grade teaching experience from .5 to 19 years (mean 7.2). In general, first-grade classrooms had approximately 25 pupils and one teacher. Typically, students spent some time (e.g., library or music) with an alternate teacher. Two exceptions were a combined first- and second-grade classroom with team teaching and one first-grade class in which two teachers divided the curriculum.

Student Descriptors

A total of 86 descriptive characteristics were generated from the research literature. We chose descriptors on the basis of an implied or established, direct or indirect relationship to first-grade student achievement. After reviewing the literature, we compiled student attributes that were frequently associated with first-grade students' success and failure. These included a number of linguistic, perceptual-motor, behavioral, and academic readiness characteristics (Barbe, Milone, Lucas, & Humphery, 1980; Brigance, 1982; Gray, 1985; McGlauchlen, 1984; Solan, Mozlin, & Rumpf, 1985). In an attempt to broaden conceptual focus, we also included unconventional student characteristics that have been implicated in first-grade student performance, as follows: ethnicity (Brophy, 1983), socioeconomic status (Mills, 1983), familial structure and interactions (Markman & Jones-Leonard, 1985), gender (Gray, 1985), personality (Fedoruk, 1988a), age (Fedoruk, 1988b), and physical appearance (Brophy & Good, 1974).


Each student descriptor was typed on a 10-cm by 5-cm white card. The deck of 86 descriptors was shuffled and presented to each teacher. A 9-point continuum, large enough to place cards at each point, was placed in front of each teacher. From left to right, the continuum was marked as follows:

1 .Absolutely contributes to student success.

2. Strongly contributes to student success.

3 .Moderately contributes to student success.

4. Mildly contributes to student success.

5. Irrelevant.

6. Mildly contributes to student failure.

7. Moderately contributes to student failure.

8. Strongly contributes to student failure.

9. Absolutely contributes to student failure.

The researcher then explained: On each card is a word or phrase which describes a certain aspect of a student early in first grade. For you as a Grade I teacher, based on your interpretation of curriculum, your management of classroom proceedings, your experience, and your evaluation of students, place each card where you feel it best belongs.

Teachers were free to ask questions and clarify instructions. Some asked many questions and

a l = absolutely contributes to student success; 2 = strongly contributes to student success; 3 = moderately contributes to student success; 4 = mildly contributes to student success; 5 = irrelevant; 6 = mildly contributes to student failure; 7 = moderately contributes to student failure; 8 = strongly contributes to student failure; 9 = absolutely contributes to student failure.

sought elaboration of most student descriptors; others asked few or no questions. The time necessary to complete sorting of the student descriptor cards ranged from 12 to 37 minutes (min) (M = 24 min). Eighty-six percent of the teachers elected to check their rankings of student descriptor cards.


The consistency of teachers' ranking of student descriptors over time was addressed. Four to 6 weeks after their initial sorting of student descriptors, five teachers repeated the procedure. Test-retest consistency of rankings ranged from 98.8% to 81.4% (mean 91.8%) allowing for one position movement on the 9-point continuum.


For each of the 86 student descriptors, range, mean, and standard deviation were calculated. Descriptors were divided into three groups, depending on the extent of teacher-response variability. Minimal disparity across teacher rankings was defined as item range of less than 4 continuum ranking positions. Moderate teacher ranking disparity was defined as a range of 4 positions on the ranking continuum. Acute ranking disparity was conceptualized as a continuum ranking range of more than 4 positions. Table 1 presents student descriptors that evidence minimal teacher ranking variability. The sample of first-grade teachers demonstrated some consistency regarding perceptions of the educational irrelevance of certain ethnic, physical appearance, gender, and academic readiness student attributes. Similarly, there is evidence of substantive teacher agreement regarding the critical role of several linguistic, attitudinal, and attention-organization descriptors. It should be noted, however, that teachers demonstrated some, perhaps critical, ranking differences. For example, many of the descriptors were ranked as irrelevant by some teachers and mildly contributing to success or failure by others. Although this difference represents only one ranking-position movement on the continuum, it may be critical in terms of classroom interactive realities.

Table 2 presents student descriptors for which there was moderate teacher ranking disparities. Teachers varied moderately in their perceptions of the educational significance of many linguistic, perceptual-motor, ethnic, physical appearance, socioeconomic, academic readiness, personality, and familial descriptors. For example, although the student descriptor "Anglo-Saxon" had a mean ranking very close to irrelevant, it was ranked as strongly contributes to student success" by at least one teacher. In addition, although "Disturbs others" had a mean ranking of moderately contributing to failure, this descriptor was ranked by at least one teacher as irrelevant and by at least one teacher as strongly contributing to student failure. One cannot help wondering if such disparity in teacher rankings differentially affect teacher-student interactions and teacher evaluations.

Table 3 presents student descriptors that demonstrated the most acute ranking disparity across the sample of first-grade teachers. In promotion of our suspicions, there are more descriptors characterized by acute disparity than there art; characterized by moderate or minimal disparity. As would be expected, Table 3 standard deviations are generally large, reflecting a range of 5 or more continuum ranking positions for each item.

Interestingly, we cannot conclude that teachers vary only in their evaluations of certain types of descriptive student characteristics. As Table 3 shows, acute teacher ranking variability occurred within a number of linguistic, perceptual, ethnic, physical health and appearance, socioeconomic, academic readiness, personality, behavioral, age, gender, and familial items. For example, at least one teacher ranked the student descriptor "thin and pale" as irrelevant in terms of student functioning, whereas at least one other teacher rated the same item as absolutely contributing to student failure.

In fact, Table 3 contains several descriptors with such large ranges one begins to wonder what teachers were thinking about during the sorting of certain items. The student descriptor "Chinese," for example, ranged in ranking position from "strongly contributes to student success" to "moderately contributes to student failure"; a range of 7 points on a 9-point scale. Casual debriefing with selective sample teachers suggested that for this descriptor, some teachers focused on linguistic differences that may place Oriental children at a disadvantage, whereas others focused on the strength of Chinese family units and intense motivation, which often translate into academic advantage. Similarly, the descriptor "disruptive" ranged from "moderately contributes to student success" to "absolutely contributes to student failure." When asked how being disruptive in class could contribute to student success, the sorting teacher proclaimed, " It's the squeaky wheel that gets the oil." Although unusual in their exaggerated variability, the rankings of these descriptors vividly illustrate how teachers react differently to certain student characteristics.


Traditionally, when we examine kindergarten populations to identify those children who we suspect may be at risk of encountering difficulty in first grade, we take measures of specific student skills or characteristics that we believe have predictive value. Because we consistently and exclusively assess only child characteristics, the assumption appears to be that the predictive indexes employed are environmentally absolute. That is, they are skills and characteristics required of all first-grade pupils by all first-grade teachers. But how reasonable is this assumption? If we can infer that the modified q-sort procedure employed here accurately reflects teacher biases, preferences, requirements, and beliefs, and if that in turn translates into differences, intolerance, interactions, and evaluations of students, the results of this study seem to suggest that first grade may not be an absolute experience for all young children. The prerequisite competencies required in one classroom by one first-grade teacher may be vastly different than those required in another classroom by another teacher. It may be, then, that the strength of predictive indexes is relative to the ecological or situational meaning of those indexes.

A critical next step, a study we have now embarked on, involves profile comparisons of high-risk first graders and their teachers, along the dimensions and descriptors used here, to determine whether predictive accuracy can be improved.


Barbe, W. B., Milone, M. N., Lucas, V. H., & Humphery. J. W.(1980). Basic skills kindergarten: Foundations for formal learning. Columbus, OH: Zaner-Bloser.

Brigance, A. H. 1982). Brigance K & I Screen for Kindergarten and First Grade. North Billerica, MA: Curriculum Associates.

Brophy, J. E. 1983). Research on the self-fulfilling prophecy and teacher expectations. Jornal of Educational Psychology, 75, 631-661.

Brophy, J. E., & Evertson, C. M. 1981). Student characteristics and teaching. New York: Longman.

Brophy, J. E., & Good, T. L. (1974). Student-teacher relationships: Causes and consequences. New York: Holt. Rinehart, and Winston.

Dembo, M. H., & Gibson, S. 1985). Teachers' sense of efficacy: An important factor in school improvement.

The Elementary School Journal, 86, 173-184. Fedoruk, G. M. ( 1988a). School entrance age: A matter of unselected clientele. Comment on Education, 18, 9-12. Fedoruk, G. M. 1988b). Student temperament and school achievement: A broadening of conceptual focus. Early Childhood Education, 21, 21-27.

Fedoruk, G. M. (1989). Kindergarten screening for first grade learning problems: The conceptual inadequacy of a child-deficit model. Childhood Education, 60, 40-42.

Gray, R. (1985). Criteria to determine entry into school: A review of the research. Springfield: Department of Planning, Research and Evaluation, Illinois State Board of Education. (ERIC Document Reproduction Service No. ED 260 826)

Kauffman, J. M., Pullen, P. L., & Akers, E. (1986). Classroom management: Teacher-child-peer relationships. Focus on Exceptional Children 19, 1-10.

Kedar-Voivodas, G. (1983). The impact of elementary children's school roles and sex roles on teacher attitudes; An interactional analysis. Review of Educational Research, 53, 415-437.

Lindsay, G. A., & Wedell, K. (1982). The early identification of educationally at risk children revisited. Journal of 'Learning Disabilities, 15, 212-217.

Markman, H. J., & Jones-Leonard, D. (1985). Marital discord and children at risk: Implications for research and prevention. In W. K. Frankenburg, R. N. Emde, & J. W. Sullivan (Eds.), Early indentification of children at risk: An international perspective (pp. 59-77). New York: Plenum.

McGlauchlen, P. L. (1984). An annotated bibliography reviewing recent research dealing with factors and innovations of school readiness. South Bend: Indiana University. Exit Project. (ERIC Document Reproduction Service No. ED 252 312)

Meisels, S. J. (1987). Uses and abuses of developmental screening and school readiness testing. Young Children, 42, 4-9.

Mills, B. C. (1983). The effects of socioeconomic status on young children's readiness for school. Early Child Development and Care, 11, 267-274.

Ouzts, D. T. (1986). Teacher expectations: Implications for achievement. Reading Horizons, 26, 133-139. Simpson., R. J., & Galbo, J. J. 1986). Interaction and learning: Theorizing on the art of teaching. Interchange,17, 37-5 1.

Solan, H. A., Mozlin, R., & Rumpf, D. A. (1985). The relationship of perceptual-motor development to learning readiness in kindergarten: A multivariate analysis. Journal of Learning Disabilities, 18, 337-344.

Thelen, H. A. (1967). Matching teachers and pupils. National Education Association of the United States Journal, 56, 18-20.


GENEVIEVE M. FEDORUK is a Social Sciences and Humanities Research Council of Canada Post Doctoral Fellow in the Department of Educational Psychology, and CHARLES A. NORMAN (CEC Chapter #43) is a Professor in the Department of Educational Psychology at the University of Alberta, Edmonton, Alberta, Canada.

Manuscript received February 1989; revision accepted July 1989.

Exceptional Children, Vol. 57, No. 3, pp. 258-263. (c)1990 The Council for Exceptional Children.
COPYRIGHT 1990 Council for Exceptional Children
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Fedoruk, Genevieve M.; Norman, Charles A.
Publication:Exceptional Children
Date:Dec 1, 1990
Previous Article:Curriculum-based measurement and two models of follow-up consultation.
Next Article:Making the transition to postsecondary education and training.

Related Articles
Who becomes an "at-risk" student? The predictive value of a kindergarten screening battery.
The RISK screening test: using kindergarten teachers' ratings to predict future placement in resource classrooms.
All-Day Kindergarten.
Building blocks: it's kindergarten prep taken to a whole new level, where getting the teachers and parents ready is just as important as educating...
A longitudinal study of early identification markers for children at-risk for reading disabilities: the Matthew effect and the challenge of...
Group won't duck school reform.
Identifying a multivariate screening model to predict reading difficulties at the onset of kindergarten: a longitudinal analysis.
Kindergarten to 1st grade: classroom characteristics and the stability and change of children's classroom experiences.
The timing of early reading assessment in kindergarten.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters