Annotated Bibliography Project

the_socio-affective_and_academic_impact.pdf

Home >Social Science homework help >Annotated Bibliography Project

ial ~~The Young Gifted Child- i ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~

The Socio-Affective and Academic Impact of Early Entrance to School

Franpoys Gagne -Nadia Gagnier

How well do early school entrants adjust socio-affectively when compared to their regu- larly admitted peers? Despite numerous publi- cations on the subject, much controversy remains, mainly because of methodologically fragile studies. To assess the impact of a new early entrance policy in Quebec, 36 kinder- garten and 42 Grade 2 teachers who had at least one early entrant in their class ranked all their students on four bipolar dimensions (con- duct, social integration, academic maturity, and academic achievement). Data were col- lected for 98 early entrants and 1, 723 regular- ly admitted children. The results revealed no substantial differences between the two groups, but a low correlation between age and adjustment among regularly admitted stu- dents. A semi-qualitative analysis showed that the teachers judged a signiricant percentage of early entrants less than well adjusted; per- haps explaining to a large extent the continu- ing resistance from educators and parents. Still, boys and the youngest among regularly admitted students were the two populations found much more at risk for social-emotional problems than early entrants.

Dr. Fran,oys Gagn6 is an Honorary Profes- sor (retired) in the Department of Psychology at l'Universite du Quebec a Montr6al. He has published extensively in the field of gifted edu- cation over the past 20 years, and is best known in the U.S. and abroad for his Differen- tiated Model of Giftedness and Talent (DMGT), as well as his psychometrc study of peer nominations (Tracking Talents). Dr. Gagne has won major awards in the field, among them publication awards from the American MENSA Association in 1993 and 1998, as well as the 1996 Distinguished Scholar Award from the National Association for Gifted Children. E-mail: [email protected] Dr. Nadia Gagnier obtained her Ph. D. in psychology from l'Universite du Quebec a Montreal in 2000. She now works in prvate practice, specializing in forensic psychology (law-related expertise). She also teaches activ- ities in testing and psychometrcs. E-mail: [email protected]

School districts in the United States began offering early

entrance options for intellectually preco- cious children almost a hundred years ago (Daurio, 1979). Yet, early entrance to school never gained general recogni- tion across North America. Indeed, out of all possible accelerative enrichment' approaches, only one, the Advanced Placement Program, has grown tremen- dously since its introduction in 1952 (DiYanni, 2003). The reason for such a limited dissemination resides in the ambivalent-and often clearly nega- tive-attitudes many educators harbor toward any form of academic accelera- tion. Defenders of accelerative enrich- ment find these attitudes difficult to explain in view of the positive conclu- sions of hundreds of empirical evalua- tions of the academic and socio-affective impact of all accelerative options. Bor- land (1989) summarizes the conundrum as follows: "Acceleration is one of the most curious phenomena in the field of education. I can think of no other issue in which there is such a gulf between what research has revealed and what most practitioners believe" (p. 185). Similar statements abound in the gifted education literature.

Rogers (1991) identified nineteen reviews of accelerative enrichment, cov- ering most of the twentieth century, from Witty and Wilkins (1933) to Van Tassel- Baska (1986). Rogers' literature review and meta-analysis outranks all previous ones as the most comprehensive and definitive, an overwhelming reason to borrow heavily from it. Before that sem- inal doctoral dissertation, reviewers had either lumped together many accelera- tive options or targeted a single one, usu- ally early entrance to school (see Rogers, Appendix A). Not only did Rogers make major efforts to unearth every published study since 1929 (she found 314 of them), but she grouped them by type of acceleration. Twelve categories were thus created: 1) early entrance to school, 2) grade skipping, 3) non-graded classes, 4) curriculum compaction, 5) grade tele-

scoping, 6) concurrent enrollment, 7) subject acceleration, 8) advanced place- ment, 9) mentorship, 1O) credit by exam- ination, 11) early admission to college, and 12) combined options. To examine the academic, social, and psychological impact on accelerants of each individual option, she adopted a special meta-ana- lytic technique called best-evidence syn- thesis, whereby the quantitative and qualitative analysis targets only those studies on a topic that meet clearly stated a priori criteria of methodological soundness (Slavin, 1986). Rogers con- cluded her review with the following general statement.

If nothing else, the synthesis compiled here lays to rest two major misconceptions about acceleration. The first miscon- ception has been that "accelera- tion is acceleration," that is, all forms of acceleration are basi- cally the same. As this study has shown, each of the 12 forms of acceleration has a very different pattern of outcomes for gifted learners. ...The second miscon- ception has been that accelera- tion may have negative or damaging social and emotional consequences for gifted learn- ers. In fact, this study has shown that there are minimal social and emotional effects for the majority of accelerative options. (pp. 200-201) D ogers' specific review of the ILkearly entrance literature covered

68 empirical studies, no less than 22% of her total pool of 314. She selected eight of them as best evidence to assess acade- mic impact. From these, she computed an effect size of .49--considered moderate according to Cohen's (1969) thresh- olds-in favor of the accelerants. In the case of social and psychological out- comes, four studies were highlighted in each category, yielding very small posi- tive effect sizes of .20 and .16 respective- ly. She concluded that early entrance to school ". . . is a reasonably safe decision

128/Roeper Review, Vol. 26, No. 3 Roeper Reviewv Spring 2004, Vol. 26, No. 3,128-138.

Manuscript submitted June 16,2003. Revision acceptedAugust 13,2003.

to make. Across a broad base of short- term and longitudinal studies based pri- marily on school records, academic performance was found to be significant- ly enhanced. Social and psychological adjustment is neither enhanced nor threatened by early entrance to school" (Rogers, 1991, p. 201).

With so much empirical data and positive results, it is tempting to con- clude that no further proofs are needed to support any form of accelerative enrichment, and that Borland's paradox stands once again confirmed. Yet, Rogers' first summative comment strongly qualifies the results from that whole research area.

One would assume that with so great a number of studies as the research base for acceleration that the academic, socialization, and psychological outcomes of at least the most widely prac- ticed forms would have been established. Such has not been the case. For academic out- comes, the picture is fairly clear, but for socialization and psychological adjustment, more is not known than is known. (1991, p. 200)T he last part of the above com- Tment applies especially well to

early entrance to school. Of the 68 stud- ies identified by Rogers, only 30 (44%) examined the social and/or psychological adjustment of early entrants. Within that group, 17 (57%o) did not give enough sta- tistical information to permit the calcula- tion of effect sizes. Concerning that particular subgroup, Rogers notes: "such studies involved a study of school records to determine if accelerates com- pared favorably to the school's popula- tion at large. Some researchers might term this 'quick and dirty' research" (p. 173). She goes even further in question- ing the importance of our knowledge about the psychological impact of early entrance to school. Says she: "Hence, for psychological adjustment outcomes, the evidence for a trivial but positive effect relies primarily on one study" (p. 183). Pennau (1981) conducted that unique study, another outstanding doctoral dis- sertation. Rogers does not address in detail the more common methodological flaws of early entrance evaluations, but others have done so.

For instance, Cornell, Callahan, Bassin, and Ramsay (1991) play the devil's advocate by asking themselves to what extent the available research on

accelerative enrichment can hold up to critical scrutiny.

Unless this question can be answered unequivocally in the affirmative, the case will not be closed on the supposedly harm- ful effects of academic acceler- ation on affective adjustment, and, thus, the persistent con- cerns of educators and parents will not be allayed. (p. 75)

After briefly summarizing both the posi- tive and negative reviews, they formu- late a qualified support:

Early entrance can be consid- ered if the applicants are care- fully screened for social and emotional maturity as wvell as academic aptitude [italics in original]. Early admission should -probably be restricted to children who are extremely advanced relative to their chronological age in both acad- emic and socio-emotional domains. (p. 79)

Unfortunately, their reanalysis does not separate clearly enough the school- readiness literature from the more spe- cific studies on the early entrance to school of intellectually precocious chil- dren. This confusion of the two fields permeates other literature reviews. Yet, the differences are substantial.

The school-readiness issue concerns the impact of age differences within cohorts of children regularly admitted in the same grade, usually the first grade. The question is: do the eldest members of a given cohort, who are almost one year older than their youngest peers, benefit from their maturational advan- tage? The answer is yes, but only mod- estly; most such studies show a significant, if low, positive correlation between within-grade age differences and both academic and socio-affective adjustment (Breznitz & Teltsch, 1989; DiPasquale, Moule, & Flewelling, 1980). Independent of differences in sex, aptitudes, or predispositions, younger entrants within a given cohort are slight- ly more at risk than older ones. That weak maturational effect disappears rapidly during the first years of elemen- tary school. Some educators have used these results as support for their strong objections against early entrance pro- grams (Halliwell, 1966; Hedges, 1977; Maddux, 1983). They did not consider that the samples included all the students and did not control for differences in academic precocity. On the other hand,

the typical early entrance study com- pares carefully selected intellectually gifted students who accelerate with nonaccelerated same age or same grade equally bright peers. The cohort phe- nomenon will be examined separately within the methodological structure of the present study.

ne of the major methodological flaws plaguing early entrance

studies has been their poor operational- ization of the concept of socio-affective adjustment. Although a few dealt with that problem in an exemplary way (e.g., Fund for the Advancement of Education, 1953; 1957; Pennau, 1981), most have used rather crude measures of adjust- ment. That measurement problem is not specific to early entrance studies; it is regularly mentioned in the psychological literature (Anastasi & Urbina, 1997; Busch-Rossnagel & Vance, 1982; Roedell, Jackson, & Robinson,1980). Everyone recognizes that the difficulty lies in large part in the ambiguity sur- rounding the concept itself. Yet, because the strongest objections toward accelera- tive enrichment target socio-affective adjustment, researchers should have given very careful attention to that par- ticular impact. Reviewing the early entrance literature, Jones and Southern (1991) underlined the measurement issue. "Researchers have demonstrated little regard for (1) the technical quality of the instruments used, (2) the natures of the traits being studied, or (3) the the- oretical significance of the traits that they have chosen for evaluation" (p. 63). As we did above, they affined: "Part of the problem in assessing social-emotional adjustment is that it is a nebulous con- cept. It is difficult to describe and mea- sure adequately" (p. 63).

uilding on these global judg- Bments, Lallemand (1996) exam-

ined the methodology and measures used in a sample of 28 evaluation studies of early entrance to school, most of them part of Rogers' pool of 68. The dispari- ties between the two sets lie essentially in Lallemand's decision to exclude (a) empirical studies that did not include clearly identified gifted students (they accounted for 70% [48/68] of Rogers' pool), and (b) studies assessing only aca- demic achievement. Lallemand identified 62 distinct measures of socio-affective adjustment, which she classified into eight categories. They appear here in order of decreasing frequency: (1) factual data (e.g., critical incidents, absenteeism, grade retention); (2) questionnaires pro-

Spring, 2004, Roeper Review/129

posed to teachers, principals, or parents; (3) published tests consisting mostly of behavior rating scales completed by teachers; (4) "homemade" rating scales completed by teachers or students (in the case of long-tern impacts); (5) undocu- mented comments made by monograph writers; (6) sociometric measures, often interspersed with questions about person- ality and temperament; (7) behavior reports in students' files written by teach- ers or principals; (8) classroom observa- tions. Rating each measure on a 5-point scale in terms of its psychometric quali- ties (precision, reliability, validity), Lallemand found that over half had received values of 1 or 2. In the light of her analysis, it could easily be argued that so much measurement error cancels any potential differences between accel- erated and nonaccelerated groups of stu- dents, thus creating the "trivial" effect sizes reported by Rogers (1991). Com- bined with the other criticisms, Lalle- mand's study convinced us that there was ample room for additional evidence, especially if more careful attention was given to methodological design.

A second motive, cultural in ori- .tgin, influenced our decision to

address that question. In 1987, after years of public pressure, the Quebec Ministry of Education introduced a province-wide early entrance policy, locally called "derogation" (to the nor- mal admission regulations). Until then, Quebec had held a strict admission date for kindergarten, namely to have reached 5 years of age by September 30th. Thanks to this new type of derogation, the number of early entrants grew rapid- ly from just a few dozen in 1987 to a yearly average of approximately 1,300 by the mid-1990s (Lallemand, 1996), approximately 8% of the target popula- tion. During that period, the total kinder- garten cohort province-wide comprised just over 90,000 students, approximately 7,500 per month if we assume an equal spread of birth dates over the 12 months. A vast majority of the candidates for early entrance have October or Novem- ber birth dates-87% (85/98) in the case of our sample. Indeed, the parents decide to have them assessed because they appear precocious, usually want to enter school, and risk losing a full year because they are missing by just a few days or weeks the September 30th "guil- lotine." Since very few parents with chil- dren born after December ask for a derogation, the effective population from which the early entrants can be drawn is

reduced to not much more than 15,000. The new policy mandated that a

psychologist assess all candi- dates on three criteria: intellectual pre- cocity, socio-affective maturity, and psychomotor abilities in that order (Gag- nier, 1999). The text of the new regula- tion did not operationalize either of these criteria. But soon, the corporate body of professional psychologists proposed some guidelines (Ordre des Psycho- logues du Quebec, 1993; 1998). Regard- ing the intellectual giftedness criterion, they set a minimum mental age of 5 years and 9 months, almost equivalent to the mental development of the oldest average children in the host group. It cor- responded to an IQ of 117 for children born in October; the required IQ grew as the age decreased. The two other criteria were mainly assessed through the initial interview with the parents, some rating scales, feedback from daycare personnel when available, and direct observations made during the testing situation. The new regulation also required profession- als to "prove" that failure to accelerate the child would create a significant risk of maladjustment; but that condition was interpreted very liberally. The IQ thresh- old and other criteria generated a target population of approximately 12% of the October-November cohort of 15,000, about 2,000 "accelerable" children if we add a small percentage for December births. So, by and large, approximately two thirds of those admissible to the Quebec early entrance program take advantage of that provision.

Many educators voiced their objec- tions to the early entrance government initiative, insisting that educational authorities should collect local data about the academic and socio-affective impact of such a controversial initiative. Bring- ing up the positive results found in the existing scientific literature did little good; Quebec educators, most of them French-speaking, often react with open skepticism when new educational initia- tives are buttressed with results from "foreign" studies. A few years after the initial implementation, a province-wide statistical analysis by the Quebec Min- istry of Education (Beauchesne, 1991) put to rest any fears about the academic achievement of the early entrants. Most of them performed as well and some- times better than their older nonaccelerat- ed peers. Concerning the specific subject of socio-affective adjustment, a few small-scale studies (Bradette, 1989; God- bout & Martineau, 1995) made headlines

in the educational community because they purportedly identified a significant number of children who were adjusting poorly to the school environment. But these small studies suffered from clear methodological flaws. The politically charged situation reinforced our convic- tion that a methodologically sound evalu- ative study of the socio-affective impact of early entrance to school could interest both local and foreign educators. 2

-<uring the planning phase of this " evaluation study, magazines,

newspapers, as well as more scholarly publications in Quebec (e.g., Conseil Superieur de l'Education, 1999) began discussing more regularly the growing problem of sex differences with regard to school achievement and dropping ouL Statistics were showing that the perfor- mance gap was increasing from the lower to the higher levels of the educa- tion system, with boys lagging behind and/or leaving school in increasing num- bers. For instance, in 1998, the province- wide ratio of girls/boys registered in undergraduate programs exceeded 60:40, and was slowly creeping toward 65%, a ratio of almost two to one. Researchers were pointing out that boys began show- ing more achievement and behavior problems in the very first years of their schooling (Tremblay, 2000). A brief sur- vey of the international literature revealed the emergence of a similar awareness in the USA (Hoff Sommers, 2000), Britain (Pickering, 1997), and Australia (Bartlett, 2002; Hawkes, 2001), among other countries. For instance, from 41% in 1970, the propor- tion of women enrolled in U.S. universi- ties had reached 50% by 1980. It now exceeds 55% and, if the trend of the last 20 years continues, women should com- prise about 60% of all higher education students by the end of the present decade (National Center for Educational Statis- tics, 2002). We searched for similar sex differences in the early entrance litera- ture, but found only one discussion of that subject (Fox, 1977); the author affirmed that her survey of a few evalua- tion studies revealed significantly more girls than boys enrolled in early entrance programs. On the question of differential adjustment, Fox stated:

To the extent that girls are developmentally ahead of boys in the early childhood years, we might expect that early admis- sion to kindergarten or first grade would be even more effective for girls than boys.

130/Roeper Review, Vol. 26, No. 3

Research is needed to deter- mine whether or not this is, in fact, true (p. 123).

That comment, as well as the topicality of the question in education circles, jus- tified adding that variable as a causal factor of potential differences in socio- affective adjustment.

The present study was designed to address the following methodological concerns: (1) identify and use the best available measure of socio-affective adjustment in school environments; (2) clearly separate the problem of selected early entrants from that of younger than average students within cohorts of regu- larly admitted children (the school- readiness issue); (3) minimize teacher bias in the assessment process by cam- ouflaging the study's major goal. This last concern ensued from a frequent mention in the literature that teachers harbor negative attitudes toward early entrance to school (Braga, 1971; Jack- son, Famiglietti, & Robinson, 1981; Southern, Jones, & Fiscus, 1989).

Method

Participants Data were collected for (a) early

entrants in kindergarten and their class- mates, and (b) students who had been admitted early in kindergarten and were now in Grade 2, also with their class- mates. A total of 36 kindergarten educa- tors and 42 Grade 2 teachers, all of them having at least one early entrant in their class, agreed to participate in the study. In kindergarten, data were collected for 43 early-admitted children hidden among 695 regular students. In Grade 2, 55 early entrants and their 1,028 class- mates were assessed. Girls represented about two thirds (67/98) of the early entrants identified. The sample included 5.4% of students with foreign cultural background; that percentage was judged too low to warrant a special analytical focus. All the students and teachers were recruited in school districts within the Greater Montreal area.

Instrument Three major goals were pursued in

the design of the instrument: (a) anchor the rating dimensions on solid theoreti- cal ground, (b) keep the task short enough to ensure adequate collaboration from the teachers, and (c) camouflage the real purpose of the study. The only way we found to ensure that camouflage

was to have them rate all their students. In turn; that decision meant creating a rating instrument concise enough to be completed in less than an hour.Rationale. A substantial number

Rof researchers (e.g., Achenbach & Edelbrock, 1986; LaFreniere, Dubeau, Janosz, & Capuano, 1990; Quay, 1979; Rutter, 1967) have tried to identify the basic dimensions of socio-affective adjustment. Through his extensive litera- ture review, Rutter (1967) extracted two major adaptive dimensions: a conduct or externalized dimension, and an anxiety- withdrawal or internalized dimension. According to Rutter, this was the most universal diagnostic distinction in child psychology. A decade later, Quay (1979) reviewed 37 studies on categories of childhood problem behaviors. He judged that four basic dimensions, namely conduct disorders, anxiety-with- drawal, immaturity, and socialized aggressiveness, could account for the 60 distinctly labeled behaviors or items mentioned across these publications. Just like Rutter before him, Quay stated that the first two dimensions were by far the most prevalent in the literature. We adopted Quay's first three dimensions, putting aside the fourth one because the problem it targets rarely manifests itself before early adolescence. To balance these dimensions described as problem behaviors, we created a positive side representing good adjustment. The bipo-

lar dimensions were labeled Conduct, Integration, and Maturity respectively. Academic achievement was added as a fourth bipolar dimension. Table 1 illus- trates the four assessment categories proposed to teachers, just as they appeared on the ranking form.

7valuation task. First, the evalua- L4tors gave basic socio-demo-

graphic information for each student, namely sex, birth date, and ethnic origin; then, they ranked all their students on each of the four dimensions successive- ly. The ranking was done as follows. First, we asked them to choose the five best-adjusted students on the dimension and rank them from 1 to 5. Second, they chose the five least-adjusted students within their group and ranked them from A to E. Third, they chose among the remaining students thliee above average (marked +) and three below average (marked -) students. All remaining stu- dents were rated average (marked = on the answer form). Ranking was judged to be faster than rating (e.g., on a Likert- type 5-point scale), as well as free from "inflation" (commonly called clemency bias), that tendency to use the high/posi- tive values of a scale excessively. The task could be completed in approximate- ly 45 minutes.

Procedure The administrators of 18 school dis-

tricts agreed to participate in the study.

Description tof the Three Bipolar Dimensions of U cGio-Affective Adjustment, plus the Academic Achievement Dimension

POSITIVE POLE NEGATIVE POLE Conduct

Good conduct Bad conduct Negotiates when there is a conflict Hits, attacks Tolerant of frustration Irritable, temper tantrums Obedient, respects authority Disobedient, defies authority

Social Integration Socially integrated and secure Isolated and anxious

Curious, attracted by novelty Anxious, fearful Calm Tense Sociable, comfortable in the group Shy, timid

Maturity toward school tasks Mature Immature

Good level of attention Low level of attention Good level of concentration Low level of concentration Shows perseverance Lacks perseverance

Academic achievement Students showing highest Students showing lowest

academic achievement academic achievement

Note. The original instrument was written in French. The content of this table is an approxi- mate translation for illustrative purposes.

Table 1

Spring, 2004, Roeper Review/131

They were asked to create a list of all kindergarten and Grade 2 teachers who had one or-more early entrants in their class. They were the only ones who would know about the real purpose of the study. They agreed to conceal that purpose to the school principals and the teachers until the end of the data collec- tion. The study would be presented as a test of a new procedure to assess the relationship between socio-affective adjustment and academic performance. The task was explained in person to each teacher who agreed to participate (81% of the 96 invited teachers). Teachers were given 2 weeks to complete their evaluations and mail them back.

Hypotheses Based on the reviewed literature on

socio-affective adjustment in school for regularly admitted students and early entrants, as well as the methodological design adopted, the following hypothe- ses were formulated.

H.1 - Because the selection process focuses on learning potential, the early entrants, in spite of their younger age, will achieve at least as well academically as their regularly admitted peers, both in kindergarten and Grade 2.

H.2 - Because of the faster learning pace of intellectually gifted children, the difference in academic achievement between the two groups will increase noticeably from kindergarten to Grade 2.

H.3 Because the selection process includes a check for potential adjustment problems, the early entrants, in spite of their younger age, will adjust at least as well socio-affectively as the regularly admitted students.

H.4 - The well-documented relation- ship between chronological age and adjustment, including academic achieve- ment, will be observable among regular- ly admitted students, but will not affect accelerated students.

.5 - In line with the literature on sex differences in school adjustment, regularly admitted girls will show better adjustment than their male peers on all four dimensions measured.

H.6 - Because of the selection process, female early entrants will not outperform their male peers on any of the adjustment measures.

Results

Data transformation The teachers' rankings of their stu-

dents were transformed into scores

according to two rules: (a) high ranks (closer to 1) would become high scores, and (b) score differences between adja- cent ranks would increase as we moved toward both extremes. The second rule ensued from past observation that out- liers are more easily differentiated than individuals closer to the mean (Gagne, 1999). Accordingly, the first five ranks were given scores of 29, 25, 22, 20, 19 respectively; the three above-average choices received a score of 17, the aver- age ones a score of 15, and the three below-average choices a score of 13. Finally, the last five ranks received respective scores of 11, 10, 8, 5, and 1. Due to class size differences, the assess- ment procedure produced approximately 5 average children (score of 15) in kindergarten and 10 in Grade 2. Three- month age cohorts were created within the large group of regularly admitted stu- dents. As previously mentioned, under Quebec Ministry of Education regula- tions children must have reached their fifth birthday no later than September 30th to enter kindergarten. Consequently, within each grade level, students born between October 1s' and December 31st became the oldest cohort, followed by the January to March, April to June, and July to September cohorts. The "real" youngest students, namely the early entrants, comprised a separate cohort.

orrelations among the four Cadjustment measures were com-

puted separately for the kindergarten and Grade 2 samples (see Table 2). They revealed a strong relationship between

Maturity and Academic achievement, whereas the two other constructs held much smaller relationships (< .40); cor- relations across pairs of constructs were moderate, in the .45 to .60 range. They were judged high enough to merge them within a MANOVA procedure.

Impact of Mode of Entrance, Sex, and Grade Level

Table 3 presents the descriptive sta- tistics for each of the four adjustment dimensions broken down by Mode of Entrance and Sex. It shows the early entrants' scores to be generally superior to those of regularly admitted children. The means also indicate that teachers judge girls better adjusted than boys; this Sex effect appears stronger than the Entrance effect. A 2 x 2 x 2 MANOVA was con- ducted to deternine the effect of three independent variables-Mode of Entrance (regular vs. accelerated), Sex, and Grade (K vs. 2)-on the four dependent vari- ables combined. The basic assumption of orthogonality was violated. Consequently, SPSS's sequential method was used to calculate sums of squares, which could adjust for unbalanced cells. According to Box's M test, the variance estimates were not homogeneous. Therefore, we used Pil- lais's criterion, the most robust against variance heterogeneity. As shown in Table 4, no main adjustment differences were found between accelerated and regu- larly admitted students. Moreover, the MANOVA produced no significant inter- action effects between Mode of Entrance and other variables. The only significant

Table 2

Mean Scores (SD) Accordinfg to Entrance and Sex: Variable - Level Conduct Integration Maturity Acad. Achiev.

Mode of Reg 15.01 (5.85) 14.90 (S.81J 14.95 (5.83) 14.90 (5.85J Entrance Acc 15.92 (5.33) 16.37 (5.88) 16.19 (5.91) 16.90 (5.39)

Sex Girls 16.84 (5.30) 15.76 (5.78) 16.36 (5.67) 15.94 (5.60) Boys 13.41 (5.81) 14.25 (5.77) 13.76 (5.71) 14.13 (5.93)

Note. The scores range from 29 (1st rank) to 1 (last rank), with 15 as the average score (see text).

Table 3

132/Roeper Review, Vol. 26, No. 3

Correlations Between the Four Adjustment Measures Conduct Integration Maturity Acad. Achiev.

Conduct .35 .56 .41 Integration .37 .55 .61 Maturity .60 .58 .78 Acad. Achiev. .45 .54 .72

Note. The values above the diagonal are from the kindergarten sample, those below from the Grade 2 sample.

effect on the combined dependent vari- ables targeted the adjustment of boys and girls [F (4, 1808) = 7.81,p <.001], a dif- ference favoring girls. However, this effect explains only 1.7 % of the total variance. No other main or interaction effects were significant. A discriminant function analysis (Tabachnick & Fidell, 1989) was conducted to determine which measures best distinguished boys from - girls; the results showed that Conduct, with a standardized discriminant coeffi- cient of .83, was by far the most signifi- cant variable, followed by Maturity with a coefficient of only .32. Moreover, a sepa- rate ANOVA revealed that sex accounted for no less than 8.7% of the total variance of Conduct rankings; the proportion was approximately the same for accelerants and regularly admitted students.

Impact of Maturation: Cohort Comparisons

To verify the existence of a relation- ship between chronological age (in months) and the four adjustment mea- sures, Pearson correlation coefficients were computed separately for the kindergarten and Grade 2 samples; we only used data from regularly admitted, nonretained students. As expected, these coefficients were all statistically signifi- cant and positive (see Table 5); but sig-

nificance was achieved through large samples, the variance explained-remain- ing below 5% in most cases. The Table 5 data suggest that this maturation effect might be already decreasing by Grade 2, although only the difference between the two academic achievement correlation coefficients was statistically significant (z = 2.09; df = 1,617, p < .05).

ow do early entrants fare when H compared with students in the

various cohorts? Table 6 presents descriptive statistics for these five groups on each of the four dependent variables; Figure 1 highlights the results for the Grade 2 sample. The large stan- dard deviations indicate large within- cohort score differences; that observation applies as well to the early entrants. Focusing on the four regular cohorts, the adjustment scores on three measures-Integration, Maturity, Acad- emic Achievement-decrease more or less regularly from the oldest to the youngest cohort; that phenomenon is present in both the Kindergarten and Grade 2 samples. In the case of the Kindergarten sample, the early entrants' adjustment means fit right between those of the second (Jan/March) and third (April/June) cohorts. In the Grade 2 sample, these same four adjustment means exceed even those of the oldest

Results of a MANOVA with Entrance, Sex, and Grade Level as Independent Variables, and the

Four Adaptation 'Measures as Dependent Variable

Source df F partial Eta2 (%) Between subjects 4 Entrance (E) 4 1.84 0.4 Sex (S) 4 7.81 1.7 Grade (G) 4 1.21 0.2 SxE 4 0.65 0.1 GxE 4 1.55 0.3 GxS 4 0.61 0.1 GxSxE 4 0.61 0.1 ***p<.001. Note.N=1,819

Table 4

Correlations Between Age and the Four Dependent Variables. by Grade Level

Grade Conduct Integration Maturity Acad. Achiev. Kindergarten

n= 686 .09 - .20** .26** .26** Grade 2 n= 933 .08 = .13-- .20** .162*

**p<.05 **p<.01 Table 5

cohort. To assess the statistical signifi- cance of some of the above observa- tions, two MANOVAs (Kindergarten and Grade 2) were conducted with chronological age (five groups) as the independent variable'and the four adjust- ment measures as dependent variables. Because of design restrictions, it was not possible to include grade level as a sec- ond independent variable. A priori com- parisons were introduced in the design so that the early entrants would be com- pared with each of the four regular cohorts. The results show that age has a significant impact on-adjustment, both in kindergarten (F (4, 721) = 4.66, p < .001), and in Grade 2 (F (4, 980) = 3.53, p < .001). The a priori tests confirm that the early entrants are judged significant- ly better adjusted than the youngest cohort (July/Sept), except for the Con- duct dimension in kindergarten. The level of adjustment of the early entrants did not statistically differ from that of the three other cohorts, except for Acad- emic Achievement in Grade 2, where the early Entrants' mean was significantly higher than that of all four cohorts of regularly admitted peers.

Profile Analysis The large standard deviations

observed in Table 6 point to a significant number of early entrants with low adjust- ment scores. The examination of individ- ual differences had not been planned at the outset, but a look at some individual results indicated that such an analysis would introduce a distinct and, hopeful- ly, useful perspective to the group com- parson approach. To assess the prevalence of students with perceived adjustment problems, it was necessary to establish a threshold profile below which a student's adjustment would be consid- ered problematic. We created that thresh- old as follows. First, the four adjustment rankings of each early entrant were recoded using the following algorithm: five points for any above average rank- ing (scores of 17 to 31), three points for an average ranking (score of 15), and zero points for any below average rank- ing (scores of 13 or lower). These four new codes gave rise to a series of 4-digit profiles representing various degrees of adjustment. All possible profiles appear in Table 7; they do not take into consid- eration the specific dimension having received a particular ranking. For instance, the "5500" profile means the student was ranked above average on any two dimensionsf and below average on

Spring, 2004, Roeper Review/I 33

Dependent Age cohorts for regularly admiHted students Early variables entrants

Oct/Dec Jan/March ApriVJune July/Sept n (k) = 144 n (k) = 185 n (k) = 192 n (k) = 165 n (k) = 43 n (2) = 214 n (2) = 239 n (2) = 226 n (2) = 254 n (2) = 55

Conduct K 14.94 16.57 14.70 14.22 14.84* (6.30) (6.44) (6.16) (5.69) (6.10)

2 15.40 15.64 15.44 14.33 16.76* (6.13) (5.37) (4.84) (5.92) (4.51)

Integration K 16.47 15.41 14.63 13.19 16.09* (6.25) (6.46) (6.32) (5.25) (5.90)

2 15.93 15.55 14.90 14.43 16.58* (5.68) (5.48) (5.00) (5.56) (5.92)

Maturity K 16.44 16.45 14.916 12.30 15.65* (5.68) (5.91) (6.59) (5.73) (5.93)

2 16.28 15.91 15.35 13.45 16.62* (5.80) (5.49) (5.05) (5.41) (5.57)

Academic K 16.65 16.36 14.56 12.49 15.88* Achievement (6.06) (6.14) (6.02) (5.85) (5.78)

2 16.10 15.74 15.36 13.87 17.69** (5.65) (5.67) (5.07) (5.06) (4.97)

* Mean significantly higher than that of the younger (July/Sept) age cohort. Mean significantly higher than that of the four other age cohorts.

Table 6

Means of the Four Adjustment Scores by Age Cohort for the Grade 2 Sample

19 a Conduct -o- Integration

18 -|r Matunty -Acad. Achiev

cn17- 0 16S

15-

12 Oct-Dec Jan-March April-June July-Sept Early

Entrants

Cohorts Figure 1

the other two. The values 0, 3, and 5 were chosen to give slightly more impor- tance (negatively) to low adjustment than to high adjustment. By adding the four digits of a profile, a new adjustment scale was created, ranging from a minimum of zero to a maximum of 20; the three val- ues chosen produced a different scale score for each profile. Using that scale and the profiles, two new thresholds were

identified (see third column in Table 7): (a) a passage from clearly positive adjustment (+) to some degree of doubt (?), and (b) a passage from doubtful to clear maladjustment (-).3

Table 7 shows the distribution of the early entrants of both sexes, irrespective of grade level, between the three cate- gories thus created. Almost two thirds of the 98 early entrants (63%) were judged

by their teachers to have adjusted very well (+) or relatively well (?) to the school environment. However, if the doubtful cases are put aside, the adjust- ment rate drops just below 50%. Con- cerning the sex effect, this alternative coding approach shows an overrepresen- tation of girls in the 'good' category, and an overrepresentation of boys in the 'bad' category. Indeed, girls obtained a significantly higher average profile score than boys (13.51 vs. 10.26; t(96) = 2.46, p < .01, Cohen's effect size d = .54).

Discussion

The two distinct analyses, quantita- tive and semi-qualitative, described above convey different messages about the impact of early entrance to school. These two perspectives will be addressed separately below, along with a series of methodological caveats. Then we will examine the role of the two other vari- ables introduced in this research design, namely sex and age within-cohort.

Impact of Early Entrance Quantitative approach. The three

hypotheses formulated with respect to the impact of early entrance were in large part verified within the framework of the quantitative data analyses. The lack of a statistically significant main Mode of Entrance effect in the first MANOVA indicates comparable adjust- ment levels for both early entrants and regularly admitted students on all four indices of socio-affective and academic adjustment; that result directly supports our first and third hypotheses. Moreover, because of the different learning pace between the two groups, it was hypothe- sized (1-1.2) that the two groups would differ academically even more in Grade 2. Although the MANOVA did not show a statistically significant Mode by Grade interaction, the cohort comparisons con- firned that tendency. The kindergarten students performed about as well on average as the regularly admitted stu- dents in the middle cohorts; in Grade 2, the early entrants outperformed even those in the October-December cohort. Because this interaction effect manifest- ed itself on only one of the four depen- dent variables, it could not be picked up by the MANOVA.

T hese positive group resultsT acquire even more significance when viewed in the light of two major attenuating factors. First, the threshold used in Quebec to admit students to

134/Roeper Review, Vol. 26, No. 3

: Mean Scores (SD) on the Four Adjustment Measures f for the Early Entrants and

the Four Age Cohorts by Grade Level (K & 2)

Table 7

early entrance is more generous than in most other school districts mentioned in the scientific literature. The minimum IQ of 117, just above +1 SD, creates a target population of almost 15% among children born during the last three months of the year. Second, because only a fraction of the admissible 4-year- olds take advantage of the early entrance program, the regularly admitted cohort includes its fair share of bright and pre- cocious students. Both factors act in tan- dem to reduce the discrepancy between the two groups. Still, our quantitative data globally replicate the positive results found in the past scientific litera- ture (Rogers, 1991); as a group, early entrants show no evidence of being more at risk for adjustment difficulties than their regularly admitted peers.

Semi-qualitative approach. The large standard deviations observed in all tables alerted us to the presence of important individual differences in per- ceived socio-affective and academic adjustment, both among regularly admit- ted students and early entrants. A closer examination of the rankings revealed that the teachers had given below aver- age rankings on two or more dimensions to almost 30% of the early entrants (27/98); adding to them students with a 3330 profile brought the total to 37%. No one will deny that a "failure" rate of 37% justifies serious concerns, especial- ly in view of the avowed goals of early entrance programs. This type of acceler- ative enrichment has been promoted as a way to preempt potential adjustment risks when young intellectually preco-

cious children are denied a learning environment that allows them to progress at their faster learning pace. If many "beneficiaries" encounter signifi- cant adjustment difficulties, the "enrich- ing" measure is clearly not achieving its goal. But, before going any further, does the study's design justify taking the above percentage at face value? Are there any mitigating factors that could make us reinterpret the profiles in a more positive light?

ethodological issues. The first methodological issue concerns

the success of the "camouflage" opera- tion embedded in the research design. Our study was specifically designed to avoid any reference to early entrance, except when the teacher wrote down the student's birth date on the answer sheet. Our efforts to hide the study's goal tar- geted mostly teachers who held strong negative beliefs toward early entrance. For instance, if a teacher had already pinpointed her early entrants because she held definite opinions against that enrichment initiative, there was a clear risk that the label would come to mind as she completed the answer sheet. In kindergarten, temporal proximity increased that risk; the chances were good that the label would have largely disappeared by Grade 2. If a large group of Quebec teachers did in fact hold strong negative attitudes toward early entrance, an unsuccessful camouflage would increase the tendency to give neg- ative rankings to accelerated students. Consequently, the negative stereotype would make early entrants'appear less

adjusted socio:affectively than they real- ly are. Unfortunately, there is no way to tell if the efforts at camouflage were successful and, if so, to what extent.S econd, we chose a ranking pro-S cedure because of its simplicity, clarity, and user friendliness. At the same time, we were aware that such a normative approach automatically cre- ates as many below average rankings as there are above average ones. Because our procedure forced teachers to place eight students in the "below average" area (the five lowest ranks and the next three), that group automatically repre- sents about 25% of an average Grade 2 classroom, and 40% of an average kindergarten classroom. Such percent- ages exceed by far the number of stu- dents who really experience adjustment problems in the average classrooms sampled in this study. In other words, our decision to give a "zero" code to any ranking below average leans heavily toward a conservative approach to the assessment of adjustment. In support of that interpretation, a few participating teachers felt the need to explain in the margin of their answer sheet that their below average rankings did not mean that these students were in need of any therapeutic services. They pointed out they were just following the directions asking them to identify the least adjusted students in their class.

Third, the above discussion has implicitly assumed a causal relationship between early entrance and adjustment. Yet, the correlational nature of the pre- sent study-and most past ones-does not allow an automatic causal linkage between the two phenomena. The etiolo- gy of socio-affective or academic prob- lems is much more complex. For instance, some children could have developed similar behavioral difficulties even if they had not been early entrants. It is even possible in a few cases that the imperfect adjustment observed would have been worse if the child had not been accelerated. As a matter of fact, it is worth noting that among the four adjustment dimensions assessed, only Conduct was almost totally uncorrelated with age (see Table 5). It suggests a reduced developmental influence com- pared to the other dimensions, at least within the short time span of one year covered by our sample of kindergartners and second graders.

C(ummary. The above methodolog- LI ical comments, especially the

second one, indicate that the estimate of 37% of adjustment problems among

Spring, 2004, Roeper Review/1 35

Frequency Distribution of Various Adjustment Profiles by Sex

Profile Value Status n boys n girls Total % cum 5555 20 + 2 17 19 19.4 5553 18 + 3 10 13 32.7 5533 16 + 4 6 10 42.9 5550 15 + 0 2 2 44.9 5333 14 + 2 1 3 48.0 5530 13 ? 2 0 2 50.0 3333 12 ? 2 3 5 55.1 5330 11 ? 1 7 8 63.3 5500 10 - 0 2 2 65.3 3330 9 - 4 5 9 74.5 5300 8 - 0 2 2 76.5 3300 6 - 2 5 7 83.7 5000 5 - 1 3 4 87.8 3000 3 - 6 2 8 95.9 0000 0 - 2 2 4 100.0

Note. Profile components: 5= above average ranking; 3= average; 0= below average. The order is irrelevant. Value is simply the sum of the recoded rankings. Status: + = good adjustment; ? = doubtful; - = bad adjustment.

early entrants most probably exaggerates the reality. How does it compare to the existing literature? Focusing on the Que- bec literature, the studies that looked at the phenomenon with a more clinical approach (e.g., Bradette, 1989) advance percentages between 15% and 20%. This is about half as much as our owvn esti- mate. With regard to the U.S. literature, no clear answer is possible since researchers in most past studies limited themselves to'group comparisons, thus bypassing the whole phenomenon of individual differences in degree of adjustment. We would submit that when the two groups were judged equivalent in terms of their socio-emotional adjust- ment it meant that there were roughly as many cases of maladjustment among the early entrants as among the regularly admitted students.

\5 7hy did we decide to lean vYtoward a more conservative

and cautious measurement approach? First, we believe that an educational ini- tiative designed to "enrich" the school lives of precocious students should be very careful to minimize the failure rate. Early entrance to school is clearly not a life or death decision. Whatever some proponents will argue, the potential risks of postponing school admission are rarely major; they can be easily overcome through later educational decisions (e.g., early entrance to first grade, later grade skipping). Second, we believe it impor- tant to highlight a darker aspect of early entrance that too few past studies have examined. There is a need to show both sides of the coin, and thus tamper the glowing judgments too often given of accelerative enrichment, like the Borland statement quoted at the beginning of this article. Finally, we want to send a mes- sage of caution to educational admninistra- tors who would like to implement an early entrance program in their school districL Those who oppose special educa- tional services for gifted and talented stu- dents watch carefully the impact of every enrichment provision, especially accelera- tive ones. While the vast majority of thriving accelerants will be overlooked, any case of maladjustment will be high- lighted as "proof' that accelerative initia- tives are dangerous for a child's well being. These opponents will refuse to acknowledge that no educational inter- vention has a 100% success rate! Conse- quently, school administrators should be careful to minimize the incidence of wrong decisions, even at the risk of refus- ing early entrance to good, but slightly doubtful candidates. In other words, here

is a situation where Type II errors-creat- ing false negatives or judging wrongly that there is nothing to worry about- need to be avoided because of their politi- cal impact on the survival of such services. Educators who wish to minimize the risks of unsuccessful accelerations should take advantage of a well-struc- tured planning tool, the Iowa Accelera- tion Scale (Assouline, Colangelo, Lupkowski-Shoplik, Lipscomb, & Forstadt, 2003).

Gender and Cohort Effects

Besides studying the impact of early entrance on school adjustment, two other potential explanatory variables were examined as part of the present study: (a) sex differences, and (b) developmen- tal/age differences within same-grade cohorts of students.

Sex differences. In the first MANO- VA, sex emerged as the only significant main effect. It is not, by any standards, a very powerful effect, since it accounts for less than 2% of all individual differ- ences observed on the four adjustment dimensions combined. But, effect sizes vary between the four dimensions: boys are especially disadvantaged in the rank- ings given by the teachers on the Con- duct dimension, namely the expression of aggression, impulsiveness, and dis- obedience. This result should surprise no educator who has observed student behavior in elementary classrooms. As mentioned in the literature review, simi- lar observations appear regularly in sci- entific publications. The fact that this MANOVA is a main effect means that the phenomenon applies regardless of mode of entrance. Indeed, the profile analysis confirms a very significant dif- ference between accelerated boys and girls with regard to profile "positive- ness;" the effect size (d) of 0.54 means that the sex variable accounts for 7% of the total variance of profile scores. These results support our fifth hypothe- sis, namely that regularly admitted boys do not adjust as well as their female counterparts, but they do not support the sixth one, namely that this sex discrep- ancy would not apply to early entrants. Contrary to expectations based on a careful selection process, the sex differ- ence in adjustment applies as strongly to early entrants. Strangely, the gifted edu- cation literature, including the only book dedicated to accelerative enrichment (Southern & Jones, 1991), has never mentioned such a phenomenon.

at seems important to under- V vline is that boys, who account

for about half of the 4-year-old popula- tion, remain underrepresented among early entrants-our sample includes twice as many girls as boys. That ratio duplicates Fox's (1977) own survey of published statistics. How can we inter- pret that difference? It could mean either that the assessing psychologists are more hesitant to recommend boys than girls because they observe behaviors they associate with eventual adjustment prob- lems, or that the parents themselves are more hesitant to send their sons for assessment because they perceive them less mature, thus more susceptible to adaptive difficulties. Unfortunately, we could not access relevant statistics on the sex differences of candidates seen by psychologists nor on those positively recommended. Beyond that potential selection bias, the most worrying ques- tion remains: why are these fewer boys still perceived by their teachers as less well adjusted than their female peers? Are there "real" differences or could they ensue, at least in part, from stereo- typical gender biases disadvantaging boys? That question remains hotly debated, as well as beyond the frame- work of this article. One thing is clear: our results parallel a growing worldwide concern about the problems of boys in the school environment, not only in terms of academic achievement, but also with regard to behavior problems, delin- quency, and dropping out. We hope that future studies will address both the size of the problem and its etiology.D evelopmental effect. The devel-

D opmental effect manifested itself through (a) correlations of the adjustment scores with age, and (b) the comparison of cohort scores among the regularly admitted students in both kindergarten and Grade 2. Whatever the dimension, the youngest (July-Sept.) regularly admitted students fared worse, sometimes far worse, than those in the three older cohorts. That result directly supports the first part of our fourth hypothesis and that maturational handicap remained present even among second graders. Within a time span ofjust one year, enough devel- opmental changes occur to influence sub- stantially the students' school achievement, their maturity toward school tasks and classroom behavior, as well as their ability to interact appropri- ately with their classroom peers. Based on Table 5 correlations, the impact of age on adjustment accounts for 4% to 6% of the variance of adjustment rankings,

136/Roeper Review, Vol. 26, No. 3

which is at the upper limit of what spe- cialists call small effects (Cohen, 1969). Its strength more or less equals that of the sex variable. As mentioned earlier, one dimension, Conduct, appears to resist that maturational process. The youngest cohort still lags behind, but there is no clear linear downward trend among the three older cohorts (see Table 6). What remains, in our view, the highlight of that specific analysis is the size of the adjust- ment difference between the youngest , "normal" cohort (July/Sept) and the even younger early entrants. The upward trend so clear in Figure 1 both confirms the second part of the fourth hypothesis and contradicts positions defended by Halli- well (1966), Hedges (1977), and Maddux (1983). It shows how well the selection process succeeds, on average, in counter- acting the downward trend among regu- larly admitted students. It shows also that the psychologists doing the assessments do take into account more thanjust the child's intellectual precocity; socio-affec- tive maturity is also considered, even with the limited means at their disposal.

Conclusion

The present study lends itself to two major conclusions. First, different analy- sis methods can lead to partly divergent observations. Here, group comparisons revealed that early entrants did not differ much from their regularly admitted peers, a result fully concordant with almost every past study on the subject (Rogers, 1991). On the other hand, a more qualitative examination of the ranking profiles identified a significant percentage of early entrants with per- ceived adjustment problems. One unex- pected benefit from our two-pronged approach is that it could explain to some extent the observed opposition between the globally positive scientific literature on the one hand, and, on the other, the persistently strong resistance to acceler- ative measures among educators. Schol- ars who celebrate the virtues of early entrance focus mainly on the published group comparisons, whereas educators, and especially those who object to accel- erative programs, react strongly to the minority of early entrants (estimated at anywhere between 10% and 20%) who struggle to adjust to the kindergarten environment. For these educators, any failure questions the whole process; they are ready to throw the baby out with the bathwater. On the other hand, propo- nents and researchers who occult the

non-negligible failure rate among early entrants-could be accused of sinning by omission in their efforts to promote the undeniable virtues of a practice designed to enrich the daily school life of intellec- tually gifted students.

ur second conclusion concerns the appropriate priority to

establish among the various sources of maladjustment in school. As our results show clearly, early entrants are definite- ly not the students most at risk for adjustment problems. Two categories of new entrants share about equally high risks: (a) boys, and (b) the youngest among regularly admitted students. In fact, a third category, not directly exam- ined in the present study, should be added, and given absolute priority, namely children with below average intellectual abilities. Hundreds of studies have demonstrated the close relationship between intellectual aptitude tests (or school preparedness tests) and early school achievement (Sattler, 1988). Because these three risk factors are not correlated, they add up. In other words, the new kindergarteners most at risk for adjustment problems are younger than average, intellectually slow boys. That is where our efforts should focus before anything else.

REFERENCES Achenbach, T. M., & Edeibrock, C. S. (1986). Manualfor

the Teacher's Report Form.and teacher version of the Child Behavior Profile. Burlington, VT: University of Vermont, Department of Psychiatry.

Anastasi, A., & Urbina, S. (1997). Psychological testing (74 ed.). Upper Saddle River, NJ: Prentice-Hall.

Assouline, S., Colangelo, N., Lupkowski-Shoplik, A., Lipscomb, J., & Forstadt, L. (2003). Iowa Accelera- tion Scale (2'1 ed.): Manual. Scottsdale, AZ: Great Potential Press.

Bartlett, K. (2002, October). Boys: Getting it right. [Report to the Ilouse of Representatives on the inquiry into the education of boys.] Canberra, Aus. tralia: Parliament of the Commonwealth of Australia.

Beauchesne, C. (1991). La situation scolaire des eleves admis avec derogation en matemelle 5 ans et en pre. miere annde du primaire [rhe school situation of stu- dents adminted early in kindergarten and fist grade]. Quebec, QC: Ministere de l'Aducation, Direction de la recherche.

Borland, J. H. (1989). Planning and implementing pro- gramsfor the gifted. New York: Teachers College Press.

Bradette, S, (1989). tvaluation des comportements et du de'veloppement des enfants admis a la suite d'une derogation a l'dge d'admission. [Evaluating the behavior and development of children admitted through a derogation to normal admission age.] Qudbec, QC' Services educatifs, Commnission scolaire Jeune-Lorette.

Braga, J. L. (1971). Early admission: Opinion versus evi- dence. Elementary School Journal, 72, 35-46.

Breznitz, Z., & Teltsch, T. (1989). The effect of school entrance age on academic achievement and social-edu- cational adjustment of children: Follow-up study of fourth gmders. Psychology in the Schools, 26, 62-68.

Busch-Rossnagel, N. A., & Vance, A. K. (1982). The impact of schools on social and emotional develop- ment. In B. B. Wolman (Ed.), Handbook of develop- mental psychology (pp. 452-467). Englewood Cliffs, NJ: Prentice-Hall. .

Cohen, J. (1969). Statistical power analysisfor the behav- ioral sciences. New York: Academic Press.

Conseil Supdrieur de l'Education (1999, October). Pour une meilleure reussite des garcons et desflller [Toward an improved academic success for boys and girls]. (Avis au ministre de I'Education). Quebec, Canada: Conseil Superieur de l'Education.

Comell, D. G., Callahan, C. M., Bassin, L. E., & Ramsay, S. G. (1991). Affective development in accelerated students. In W. T. Southem and E. D. Jones (Eds.), The academic acceleration of gifted children (pp. 74- 101). New York: Teachers College Press.

Daurio, S. P. (1979). Educational eruichment versus acceleration: A review of the literature. In W. George, S. Cohn, & J. Stanley (Eds.), Educating the gifted: Acceleration and enrichment (pp. 13-63). Baltimore, MD: John Hopkins Uni'ersity Press.

Deschamps, 1. (1999). L'impact developpemental de la derogation il 'dge d'admission d l'ecole primaire [The developmental impact of derogating to the pre- scribed admission age to elementary school]. Montre- al, QC: Unpublished masters' thesis: Faculte des Etudes Supdrieures, Universite de Montreal.

DiPasquale, G. W., Moule, A. D., & Flewelling, R. W. (1980). The birth date effect. Journal ofLearning Disabilities, 13, 234-238:

DiYanni, R. (2003). The story ofAP, the Advanced Place- ment Program. Retrieved from httpJ/1apcentral.col- legeboard.com/prograrnihistoryl.

Fox, L H. (1977). Sex differences: Implications for pro- gram planning for the academically gifted. In J. C. Stanley, W. C. George, & C. H. Solano (Eds.), The gifted and the creative: Afifty-year perspective (pp. 113-138). Baltimore, MD: John Hopkins University Press.

Fox, L. H., &Washington, J. (1985). Prdgrams for the gifted and talented: Past, present, and future. In F. D. Horowitz & M. O'Brien (Eds.), The gifted and talent- ed: Developmentalperspectives (pp. 197-221). Wash- ington, DC: American Psychological Association.

Fund for the Advancement of Education. (1953). Bridging the gap betveen school and college. New York: Ford Foundation.

Fund for the Advancement of Education (1957). They wvent to college early. New York: Ford Foundation.

Gagn6, F. (1999). Tracking talents: Identifying multiple talents through peer, teacher, and self -nominations. Waco, TX: Prufrock.

Gagnier, N. (1999). L'adaptation socio-affective et sco- laire des e'leves admis a une entre'e precoce aupr's- colaire [The socio-affective and academic adjustment of students admitted early in kindergarten]. Unpub- lished doctoral dissertation. Montreal, QC: Departe- ment de psychologie, Universite du Quebec a Montreal.

Godbout, G., & Martineau, P. (1995). Effets des deroga- tions relatives a l'admissibilite exceptionnelle a l'education priscolaire et a 1'enseignement primairea [Effects of derogations related to the exceptional admission to kindergarten and first grade.] Montreal, QC: Service de la formation genmrale, Conumission scolaire de Montreal.

Halliwell, J. W. (1966). Reviewing the reviews on entrance age and school success. Joumal of Educa- tional Research, 59, 395401.

Hawkes, T. (2001). Boy oh boy: Howv to raise and cducate boys. Sydney, Australia: Pearson Education.

Hedges, W. D. (1977). At what age should children enter first grade: A comprehensive reviesv of research. Ann Arbor, Ml: University Microfilms International.

Hoff Sommers, C (2000). The Ivar against boys: Ho,v misguidedfeminism is harming our young men. New York: Simon & Schuster.

Jackson, N. E., Famiglietti, J., & Robinson, H. B. (1981). Kindergarten and first grade teachers' attitudes toward early entrants, intellectually advanced stu- dents, and average students. Journalfor the Educa- tion of the Gifted, 4, 132-142.

Jones, E. D., & Southern, W. T. (1991). Objections to early entrance and grade skipping. In W. T. Southern and E. D. Jones (Eds.), The academic acceleration of gifted children (pp. 51-73). New York. Teachers Col- lege Press.

Keating, D. P. (1979). Secondary-school programs. In.A. H. Passow (Ed.), The gifted and talented: Their edu- cation and development (pp. 186-198) [78th Year- book of the NSSE]. Chicago: University of Chicago Press.

Spring, 2004, Roeper Review/137

LaFreniere, P. J., Dubeau, D., Janosz, M., & Capuano, F. (1990). Profil socio-affectif de l'enfant d'age presco- laire. [Socio-affective profile of the preschool child.] Revue Canadienne de Psycho-Education, 19, 2341.

Lallemand, J. (1996). Analyse descriptive et critique des instruments employds pour evaluer 1'entree precoce. [Descriptive analysis and critique of the instruments used to evaluate early entrance.] Unpublished Masters Thesis: Universite du Quebec a Montreal.

Maddux, C. D. (1983). Early school entry for the gifted: New evidence and concems. Roeper Revieiv, 5(4), 15-17.

Maker, C. J. (Ed.) (1986). Critical issues in gifted educa- tion. (Vol. 1). Austin, TX: Pro-Ed.

Masse. P., & Gagn6, F. (1983). Observations on enrich- ment and acceleration. In B. M. Shore, F. agne, S. Larivee, R. H. Tali, & R. E. Treniblay (Eds.), Face to face with giftedness (pp.395413). New YorkL Trilli- um Press.

National Center for Educational Statistics (2002). Digest of education statistics, 2002. Retrieved January 9, 2004 from http:J/nces.ed.gov/programs/digest/dO2/tables/dtl74.asp

Ordre des Psychologues du Quebec. (1993). Guide deval. uation pourfins de derogation a l'age d'admission h l'ecole [Assessment guide in the context of deroga- tions to admission age to school]. Montreal, Canada: Author.

Ordre des Psychologues du Quebec. (1998). Guide d'ival- uation en vue d'une dirogation h l'oge d'admission h l'ecole. [Assessment guide in the context of deroga- tions to admission age to school] Montreal, Canada: Author.

Pennau, J. E. (1981). The relationship between early entrance and subsequent educational progress in the elementary school. Dissertation Abstracts Internation- al, 42, 1478 (4A).

Pickering, J. (1997). Raising boys' achievement. Stafford, UIC Network Educational Press.

Quay, H. C. (1979). Classification. In H. C. Quay & J. S. IVefry (Eds.), Psychopathological disorders of child- hiood(2V ed.: pp. 142). New York Iiley.

Roedelt, W. C.. Jackson. N. E.. & Robinson. H. B. (1980). Giftedyoung children. New York. Teachers College Press.

Rogers, K. L. (1991). A best evidence synthesis of the research on types of accelerative programsforgifted students. (Volumes I & 2). Dissertation Abstracts Intemational, no. 9122206. U.M.I. Dissertation Infor- mation Service.

Rutter, M. (1967). A children's behaviour questionnaire for completion by teachers. Journal of Child Psychol- ogy and Psychiatry, 8, 1-Il .

Sattler, J. M. (1988). Assessment of children (2"4 ed.). San Diego, CA: J. M. Sattler Publisher.

Schiever, S. W., & Maker, C. J. (2003). New directions in enrichment and acceleration. In N. Colangelo & G. A. Davis (Eds.), Handbook of gifted education (3.d ed., pp. 163-173). Boston: Allyn and Bacon.

Slavin, R E. (1986). Best-evidence synthesis: An altema- tive to meta-analytic and traditional reviews. Educa- tional Researcher, 9(5), 5-11.

Southem, W. T., & Jones, E. D. (Eds.) (1991). The acade- mic acceleration of gifted children. New York: Teach- ers College Press.

Southem, W. T., Jones, E. D., & Fiscus, E. D. (1989). Practitioners' objections to the academnic acceleration of young gifted children. Gifted Child Quarterly, 33, 29-35.

Tabachnick, B. G., & Fidell, L. S. (1989). Using multi- variate statistics (2,4 ed.). New York: Harper Collins.

Tremblay, Rx E. (2000). The development of aggressive behavior during childhood: What have we learned in the past century? International Journal of Behavioral Development, 24(2), 129-141.

Van Tassel-Baska, J. (1986). Acceleration. In C. J. Maker (Ed.), Critical issues in gifted education: Defensible programsfor the gifted (pp. 179-196). Rockville, MD: Aspen.

Witty, P. A., & Wilkins, L. W. (1933). The status of accel- eration or grade skipping as an administrative prac- tice. Educational Administration and Supervision, 19, 321-346.

The data reported wvere collected as part of a Ph).D. thesis conducted by the second attthor, now in pri- vate practice as a clinical andforensic psychologist.

Notes I Scholars in the field of gifted education common- ly distinguish enrichment and acceleration (Schiever & Maker, 2003), sometimes almost opposing them as the two poles of a continuum (Fox & Washington, 1985; Keating, 1979, Maker, 1986). We believe that the goal of any special provision for academically talented students is to enrich their curriculum by whatever means. By virtue of their condensation (or compaction) of the curriculum in a given unit of time, all accelerative provisions are automatically enriching. Masse and Gagne (1983) labeled it 'enrichment in densi- ty" (p. 409). Consequently, we believe that the expression 'accelerative enrchment" best descrbes the nature and goal of such measures. 2 Unknown to us, a graduate student in another Quebec university conducted during the same period a study on the same subject (Deschamps, 1999). She used a different methodology with a smaller kindergarten sample composed of 25 early entrants, 28 older (Oct-Dec) regularly admit- ted students and 26 young (July-Sept) students. She assessed three dimensions: (a) socio-affec- tive abilities, (b) academic performance, and (c) fine motor control. Out of seven socio-affective measures, the early entrants obtained lower scores than the older controls on only one. Acad- emically, no group differences were observed. In terms of fine motor control, the older group slight- ly outperformed the two other younger groups. Her tables did show large within-group individual differences on all dependent variables, but Deschamps did not examine them. By and large, her conclusions parallel the present ones. 3 No profiles were created for the very large sam- ple of regularly admitted students; they were not judged relevant to the analysis. But, because of generally similar means and standard deviations (see Table 3), the percentages in the three major categories (+ l ? / -) would probably be very simi- lar, maybe with a slightly downward trend.

138/Roeper Review, Vol. 26, No. 3

Roeper Review invites papers for an upcoming special issue.

Conceptualfoundations for the field of gifted education Contributions should accomplish the following.

1. Challenge and/or extend current thinking in the field. 2. Explore and clarify one or more of the following dimensions of the conceptual foundations

for the field: - Theories of giftedness, talent, or creativity. - Philosophical foundations or other tacit thought frameworks that shape our concep-

tions of giftedness, talent, or creativity. - Big-picture trends, issues, and future directions in the field. - Historical perspectives that show how conceptual foundations from the past have

influenced current thinking for better or worse. - Interdisciplinary perspectives with relevance to the field.

Don Ambrose ([email protected]) will serve as guest editor. The deadline for submissions is April 30th, 2004.

TITLE: The Socio-Affective and Academic Impact of Early Entrance to School

SOURCE: Roeper Rev 26 no3 Spr 2004 WN: 0410800833004

The magazine publisher is the copyright holder of this article and it is reproduced with permission. Further reproduction of this article in violation of the copyright is prohibited. To contact the publisher: http://www.roeper.org/