META-ANALYSES
Meta-analytic studies bring together statistical results from an array of individual studies, with the goal of achieving a broader perspective on the research. As part of A Nation Deceived, Kulik (2004) produced a seminal meta-analysis of social and emotional results from acceleration studies conducted over the previous 40 years. Kulik found that educational plans (e.g., planning for college) benefited from acceleration, while “liking for school,” “participation in school activities,” and “self-acceptance” all showed inconsistent or negligible relationships with acceleration. Importantly, only a “small number of studies” (p. 18) were identified as addressing affective topics, and there were “few studies from recent years” (p. 21). Kulik’s statistical analysis in the report is followed by a comprehensive review of research by Robinson (2004), who suggested that while we still have much to learn, “educators’ worries about harming students by accelerative choices can generally be laid to rest” (p. 64).
Steenbergen-Hu and Moon (2011) supplemented Kulik’s (2004) meta-analysis, increasing the number of studies synthesized and incorporating research published through 2008. The authors focused on factors that might explain the concerns of practitioners in the absence of empirical evidence. Their results echoed those of Kulik, demonstrating no consistent social or emotional differences between accelerated and nonaccelerated students, regardless of whether comparison was made to age peers or grade-level peers. In addition, they found no differential results on the basis of gender or type of acceleration. The authors did detect a significant effect of grade level at acceleration, with students accelerated in elementary grades experiencing social and emotional advantages, while secondary students showed a slight decline in self-esteem upon acceleration. This recommendation for early acceleration supports prior findings and is, in turn, supported by subsequent individual studies.
INDIVIDUAL STUDIES
While meta-analytic studies offer a clear snapshot of the wide landscape of research, they have several limitations. Social and emotional research concerning acceleration is confounded by multiple definitions (e.g., “giftedness,” “self-esteem,” and “acceleration”). Meta-analyses bring multiple studies together to make them statistically comparable, but in doing so, these potentially important differences are lost through consolidation into uniform definitions. This makes it critical to identify studies specific to the purpose and interpret results in the context of the study design. This information is highlighted in the studies presented below, clustered by the acceleration practice addressed.
Grade-based acceleration. Skipping one or more grades is one of the more common forms of acceleration. In fact, Colangelo, Assouline, and Marron (2013) suggest that much of the resistance to acceleration stems from “the misconception that acceleration refers exclusively to grade-based acceleration” (p. 167). Unlike some other forms, grade-based acceleration typically involves the omission of a full year of curriculum, as well as the replacement of an established peer cohort with a new one.
Research on grade skipping is supportive. Hoogeveen, van Hell, and Verhoeven (2009) followed 53 students in the Netherlands as they were accelerated past a single grade level into grades 7 or 8, using the Self-Description Questionnaire II (SDQ-II; Marsh, 1990) to assess differences in self-concept. The comparison group was the newly acquired cohort of grade peers. The authors found no significant differences in total self-concept, but there was a difference in social self-concept regarding same-sex relations, with accelerated students scoring lower than nonaccelerated peers. As students moved into their second year after acceleration, these social differences were exacerbated in boys but disappeared among girls. The authors suggested this may be a result of later onset of puberty in boys, exacerbating physical differences between accelerated and nonaccelerated boys. These results indicate the importance of considering gender when selecting students for acceleration.
In addition to the SDQ-II, Hoogeveen et al. (2009) surveyed the accelerated students and their peers to paint a picture of social standing and reputation among the students. Students were asked to list the three students they liked most, the three they liked least, and which students most often showed a variety of behavioral traits, from conceit to leadership. The authors used the nominations of popularity to calculate social impact (the number of times a student was mentioned, positively or negatively) and social preference (the difference between the numbers of positive and negative nominations). Repeated measures analysis indicated that accelerated students tended slightly toward higher social impact (η2 = 0.013). For reference, this was slightly less pronounced than the difference between boys and girls (η2 = 0.022). The accelerated students received lower standardized scores for social preference (η2 = 0.078), which dwarfed the difference between girls and boys (η2 = 0.014). In summary, the accelerated students were more frequently mentioned as a notable facet in the social experience of their classmates, but typically in a more negative sense.
In terms of trait reputation, conceit was mentioned more often among accelerated students (η2 = 0.121), while cooperation (η2 = 0.057), helpfulness (η2 = 0.052), and humor (η2 = 0.032) were moderately lower among the accelerated sample. Together with the SDQ-II results, these results suggest that while general self-concept seems unrelated to acceleration, the landscape of peer relations is probably still affected.
Hoogeveen, van Hell, and Verhoeven (2012) produced a complement to this study, incorporating a much broader age range and introducing an analysis of environmental factors. The authors assessed 148 accelerated students and 55 nonaccelerated students, all identified as gifted and ranging in age from 4 to 27 years, using three forms of the SDQ calibrated to different age groups. SDQ results indicated no overall differences between accelerated students and their grade peers. There were, however, different results based on the grade at acceleration. The nonaccelerated students showed a decline in both peer relations and general self-esteem as they entered grades 4–6, while accelerated students did not. This interaction with grade level at acceleration is a consistent theme, and the authors draw on these results to recommend early acceleration, suggesting that the passage of time has a positive effect on the self-concept of accelerated students.
The authors also employed a parent questionnaire and student diaries as part of their study. From these sources, the authors discovered that in secondary, nonaccelerated students, the quality of parent interaction had a significant and meaningful impact on total self-concept. This relationship was statistically insignificant for accelerated students, and the authors suggested that the accelerated students were “less susceptible to personal and environmental factors” (Hoogeveen et al., 2012, p. 598). Parents also indicated that risk-avoidance and “underground” behaviors (e.g., denying giftedness or abstaining from gifted programming) were less prevalent in accelerated students.
Early entrance. In the same year that A Nation Deceived was published, Gagné and Gagnier (2004) published a study on students admitted early to kindergarten in Quebec. They assessed 98 early entrants, 43 in kindergarten and 55 in second grade, on four social and emotional constructs: Integration, Maturity, Achievement, and Conduct. To gain greater resolution for comparison, the researchers divided the nonaccelerated classmates into four quarters based on month of birth.
Across all four measures and in both grades, the accelerated students consistently scored higher than their youngest classmates. In most cases, the mean score for the early entrants fell squarely between the second and third cohorts of their classmates (i.e., centering on the age median). In addition, in second grade, the accelerated students exhibited higher achievement scores than all comparison groups. Thus, early entrants did not represent behavioral outliers in either kindergarten or second grade, and their academic achievement recovered and once again exceeded their classmates by second grade.
In an attempt to capture faculty perceptions, the researchers also asked teachers to indicate their five best- and worst-adjusted students on each of the four scales above. The ratings were designed to be blind to the acceleration status of the individual students to avoid bias in the ratings. The results indicated that teachers in both grades found 30% of accelerated students to be “struggling” on two or more scales, in contrast with just below 24% across the entire sample. In the context of the more positive results described above and the authors’ admitted limitations on keeping...