Introduction
Across the country, there are very bright twelve-, thirteen-, and fourteen-year-olds who might, with reasonable effort, successfully initiate college-level studies. Many reasons exist why they generally do not, including the paucity of higher education opportunities for them [6] and, perhaps more fundamentally, widespread, deeply rooted, and frequently reasonable concerns about the hazards of "rushing" development. In individual cases, however, such concerns may be more hampering than helpful. This article suggests that for a constituency of exceptionally talented-but quite young-students, college enrollment, particularly when undertaken in concert with other able and motivated peers, can be both academically enhancing and facilitative of personal and social growth.

The studies about what happens to early entrants in the years after graduation are as reassuring as a relatively small literature can be [4].

Pressey [28], for one, assessed the adult status of 156 participants in the Ford Foundation early entrance program [7, 8], all of whom were to some degree younger than average at entrance to college, and 29 percent of whom had entered under the age of sixteen. Records of a subsetof the group of young scholars were compared with records of regular-age students paired with them on ability. In both the young and the comparison groups, a majority went on to obtain advanced professional training, which the former completed when about two years younger. Otherwise, adult career accomplishments of the two groups appeared comparable. Ten years after graduation, most of the early entrants reported that they had sustained little or no harm from skipping high-school work, that their social difficulties because of youth at entrance had been transitory, and that they had derived gains from beginning their careers early.

Janos [16] extracted long-term follow-up data from Terman's classic longitudinal study of 1528 children identified by IQ scores above 140. The records of the 19 youngest college entrants (mean age at entry = 14.8 years) were compared with those of non-accelerated agemates matched on IQ. The young college students earned higher grades and more academic honors and participated in more extracurricular activities. They graduated from college and entered the professions younger,and were, in 1940, more often rated as high achievers. This qualitativesuperiority had, however, disappeared twenty years later. Once again, no differences between groups in psychological adjustment or social participation were observed at any point during or long after college.

One additional study and a collection of cases complete the roster of efforts to follow the progress of early entrants after college. Stanley and Benbow [36] provided data for, among others, eleven nineteen-year-old (or younger) baccalaureates who had graduated from Johns Hopkins University at least thirty years earlier. The authors concluded. "There are no hints of 'early ripe, early rot.' It is apparent that these early graduates have led or are still leading highly effective adult lives" [36, p. 366]. Finally, Stanley (34) reported the progress of the six unusually young college graduates featured in a 1977 issue of Smithsonian. Five had obtained Ph.D.'s in difficult subjects from top-flight universities and were employed on the faculties of others; the sixth was completing doctoral work in statistics.

Of numerous reports by Stanley (and his associates) on accelerated education, few pertain to markedly young college students. Those that do [27, 35] are indubitably important, but their follow-up term was relatively short, and in any case, most of the subjects were only slightly ahead of agemates in academic-placement. Pollins compared the psychological and social adjustment of 21 males who had, at some point in their higher education, been ahead of agemates by three or more years; comparison subjects were comparably able, non-accelerated agemates. Evaluating data collected at the beginning of the study and five years later, she concluded, "The present study...found no negative effects of acceleration on social and emotional adjustment. In fact, some evidence of positive effects is presented" [27, p. 176].

Stanley [35] reported the progress of 25 students in the Johns Hopkins entering class of 1980 who had entered college at least a year early; 6 of these were age fifteen at entrance, and only 2 were younger. By June 1984, 23 of the 25 had graduated, with a mean grade point average of 3.33 (of a possible 4.0; SD = 0.41), including all of the 8 youngest entrants. Stanley concluded, "The student who begins [college] young after considerable counsel and planning to bridge the academic and social chasm separating high school from college stands an excellent chance of speeding his or her way effectively into high-level professions and a satisfying life" (35, p. 227). Stanley cautioned that special support appeared to be helpful and that females moving into residence halls at fifteen or sixteen had more social and academic problems than males.

Another set of studies whose subjects all participated in one program was conducted at the Early Entrance Program (EEP) at the University of Washington. The studies have considered the status of youngsters who entered college exceptionally young; indeed, the first two of the following citations constitute initial assessments of a subset of subjects of the present longitudinal study. Janos and Robinson [17] found that early entrants liked school more and earned higher grades than regular-age college students; they were also found to be roughly comparable to National Merit finalists in these respects. Using the MMPI and a variety of other self-report indices, Robinson and Janos [32] documented that the psychosocial adjustment of early entrants appeared comparable to that of regular-age college students, of National Merit finalists, and of agemates who had qualified for early entrance but elected to attend high school.

Janos, Robinson, Carter, et al. [19] documented the social function of admitting young students as a group, which appears both to facilitate attachment to age/ability peers (fellow EEP students) and to provide a secure base for the development of relationships with regular-age university students. Finally, Janos, Sanfilipo, and Robinson [20] described a small group of early entrants who were not prospering academically. The females of this group appeared mature in many respects and had consciously chosen to follow a social agenda, whereas the males in this group manifested longstanding family and personal troubles. The females as well as the males, however, appeared to find college far more supportive of their all-around development than available alternatives.

The findings above hardly constitute the entire basis of "the case for radical acceleration" [31]. Many believe that "doing nothing--failing to accelerate children who seem ready for more advanced educational placement--is not necessarily the safest course" [31, p. 145]. Expecting youngsters to adapt to situations that fit neither their academic nor social needs may place them at risk for an extended period of exceptional unhappiness; they may become fixed in alienated patterns of adjustment and perhaps even drop out of school. A substantial minority of young people with superb academic talents simply do not function well among typical agemates, although they may continue to grow socially among those with whom they feel a genuine connection [12, I5, 18, 19].

In summary, the literature about early college entrance is small and, in many respects, inadequate. Studies have been limited by relatively small samples, particularly with respect to markedly young subjects, ad-hoc instrumentation, inadequate comparison groups, and short follow-up periods after matriculation. The study reported below represents movement toward replication and extension. Its subjects were, however, still relatively young and in college; its focus is on their attainment and adjustment during the undergraduate years. The investigation was undertaken at the University of Washington (UW), where, since 1977, the Early Entrance Program (EEP) has enrolled and supported a growing concentration of markedly young college students.

The study compared the academic transcripts and psychological questionnaire responses of a very young college group with data from three groups of students varying in age, ability, and academic placement (high school or college).

Before proceeding to the empirical study, an orientation to the EEP [31, 33], may be useful. Its students are carefully selected (see below) and enter the program before age fifteen. Sufficient numbers of students (12-15 per year) are admitted to foster what has proved to be thriving peer interaction. Since 1981 almost all students have completed a year of intensive college preparatory studies in the problematic on-campus Transition School of the EEP, which addresses the almost inevitable academic unevenness a seventh or eighth grader brings to college. Academic counseling is also available, and students are prompted to seek appropriate information from other university personnel and to make autonomous decisions. Emotional support and psychological and academic consultation is offered by the small staff of clinical psychologist and professional educators. A center is provided for social and recreational activities, and young students are invited to feel welcome throughout their undergraduate years and beyond.

Method
Between 1977 and 1983, by maximum age fourteen, 20 males and 23 females had enrolled in the EEP. All had been screened originally with the Washington Pre-College Test [25], which yields equivalent scores on the Scholastic Aptitude Test. Their average SAT-V and SAT-M equivalent scores were well above-the 85th percentile of Washington State college-bound twelfth graders and in the top-scoring 2 percent of the highly select seventh-grade students taking the SAT as part of the Johns Hopkins Talent Search [3]. Students who earn such scores at ages twelve to fourteen tend to earn exceptional scores when reassessed at the usual time as high school juniors and seniors [2]. In addition, among the EEP group, maturity and readiness for college are confirmed by multiple teacher recommendations, as well as by interviews with parents and students, and by observations of students, behavior during visits to the Program's Transition School.

As implied in the literature review, the EEP group must be contrasted with the more widely known accelerates described by Stanley and his co-workers. The latter students typically pursue several years of intensive precollege, academic cultivation in the years after taking the SAT, which is generally in seventh grade. They also tend to be noted for their quantitative gifts. The EEP students tend to be very much younger at college entrance, and a much greater proportion evidence higher verbal than mathematical ability. By any estimate, they constitute a highly gifted population, many of whom, consequently, seem to have previously experienced social disjunction with middle-school classmates.

In order to compare the EEP sample with other able students--those who were the same age and those with whom they were attending class--three comparison groups were chosen. The best available group of comparably bright agemates were 25 males and 19 females (QUALs) who had qualified to enter the EEP but had chosen to attend high school. Because members of the QUAL group were relatively satisfied with conventional, age-graded educational programs, they were expected to represent a stringent standard of healthy adjustment.

Two groups of regular-age UW students--on average about four and a half years older than the EEP group--were also recruited. The first, "regular" (REG) group was considered typical of the select undergraduate population admitted by the UW, the top-ranked public university in Washington state. These 43 students were matched to individuals in the EEP group by gender, year of UW entry, high-school catchment area (as a rough control for social class), and preadmission Washington Pre-College Test (WPCT) Verbal and Quantitative Composite scores. It was expected that the REG group would be comparable to the EEP group in "readiness" for college, although they were of course older and, on average, probably not as academically inclined. A second group of regular-age college students, National Merit finalists (NATs), who were demonstrably of high academic caliber (13, 23), was recruited to provide a better match for ability. The NAT group was also matched to the EEP group on gender and year of matriculation. (The NAT group lacked a match for one EEP female who had enrolled in 1977).

Participating groups' college readiness scores and mean ages at college matriculation are presented in table 1. As noted above, because SAT scores will be more familiar to most readers, but were largely unavailable except for National Merit finalists, table 1 reports SAT estimates derived from WPCT scores according to formulas supplied by the UW's Educational Assessment Center. The NAT group, as expected, scored highest in both verbal and mathematical areas. There were no statistically significant differences between EEP and REG or EEP and QUAL students on the SAT equivalent scores, in the former comparison due to matching procedures. The discrepancies between SAT-V and SAT-M equivalent score are comparable to discrepancies in the test scores of 1986 college-bound seniors, which for SAT-V and SAT-M, respectively, were 426 and 451 for females and 437 and 501 for males [3].

After three years of recruitment, the populations of each of the four groups numbered 42-43. Approximately 60 percent had enrolled at the UW, or had been qualified for the EEP, between 1977 and 1981. These students, ranging from freshmen to senior, were recruited during the first year of the study. Because the cohorts added in the second and third years included only entering freshmen (EEP, REG, and NAT and QUALs, smaller numbers were added in 1983 and 1984. Data were collected annually: from the 1982 cohort three times (1982, 1983, 1984), from the 1983 cohort twice (1983, 1984), and from the 1984 cohort only once (1984). Certain data were collected only once; three measures bearing directly on social adjustment and maturity were collected during each year of a subject's participation. Table 2 summarizes these various factors of enrollment and measurement. Slightly varying subject numbers in subsequent tables reflect the fact that not all subjects were willing to complete all measures of the lengthy battery.

Table 1
Age and College Readiness at Matriculation*

The measures included in the test battery focused on several of the issues discussed in the literature review. First, it was thought important to assess aspects of intellectual functioning beyond those reflected in college grades. Although comparability of SAT-equivalent scores existed among the EEP, QUAL, and REG groups at the time of applying to the UW, it was expected that, because of their youth, the EEP and QUAL groups would have developed more rapidly in the average of two and one-half years transpiring between matriculation and the time of participation in this study. To describe changes since matriculation and to provide an up-to-date covariate for intellectual ability in other analyses, verbal ability was assessed with the Concept Mastery Test (CMT) (38). The CMT, composed of a vocabulary test and set of analogy problems, was designed to assess high levels of abstract ability rather than instructional content and has proved useful in other studies of youths identified as gifted [26, 37]. Assessment of students' knowledge about and attitudes toward studying was provided by the four scales (Delay Avoidance, Work Methods, Teacher Approval Education Acceptance) of the Survey of Study Habits and Attitudes (SSHA) [1].

Table 2
Organization of Subjects and Instruments*

Of particular interest was the social and personal adjustment of the EEP group, and for this reason a variety of tests in the battery addressed adjustment and maturity: the California Personality Inventory (CPI) (10); the Tennessee Self-Concept Scale (TSC) [5]; the Inventory of Parent and Peer Attachment (IPPA) [11]; and the Defining Issues Test (DIT) [29]. The CPI and TSC are familiar, widely used, and somewhat overlapping instruments. The Total Positives score from the TSC, a convenient summary score for adjustment, complemented the descriptive richness of the CPI. In addition to the eighteen scales of the CPI, a Social Maturity Index was computed according to procedures set forth by Gough [9]. Gough asserts that this index taps "important determiners of positive and socially constructive behavioral outcomes" [9, p. 189]. The IPPA specifically assessed trust in, communication with, and alienation from parents and peers. The "P" score on the DIT addresses the cognitive underpinnings of mature social interactions. It assesses the degree to which subjects use moral principles to evaluate behavior presented in a series of vignettes. Higher "P" scores reflect greater utilization of humanistic moral principles in evaluations; lower "P" scores reflect greater weighting of consequences and conformity to authority. "P" scores are, however, strongly correlated with intelligence and education [30]. Although the psychological factors indexed by "P" scores probably have some bearing on behavior, such social cognitive measures provide an inadequate basis for reliable prediction [18]. In order to assess possible differences in family relational patterns, the Family Environment Scale (FES) [22] was administered. It provided respondents, assessments of family support and conflict, the directions for growth emphasized by parents, and degrees of family organization and structure.

All analysis programs were selected from the Statistical Package for the Social Sciences [14, 24]. Because conceptually distinct domains were tapped by GPA, Concept Mastery Test Total score, and Defining Issues Test P scores, these indices were subjected to separate Group X Sex ANOVAs. The four scales of the Survey of Study Habits and Attitudes, the scales measuring social and psychological adjustment on the CPI and Inventory of Parent and Peer Attachment, and the Family Environment Scale, in contrast, were submitted to multivariate analyses of variance (MANOVAs) on a measure-by-measure basis. Where significant differences were obtained in MANOVAs, additional MANOVAs separately comparing the EEP with other groups were also conducted. T-tests and, in the case of EEP-REG comparisons, paired t-tests were used for univariate tests. In these post-hoc comparisons, Dunn's procedure was employed to minimize the probability of accepting a statistical aberration as a true difference in population means. Changes over time in DIT P and CPI scores were analyzed using the repeated measures MANOVA procedure.

Results
Response Rates
Of potential subjects initially contacted by letter, 84 percent of the EEP group (20 males and 23 females), 84 percent of the QUAL group (25 males and 19 females), 37 percent of the REG group (25 males and 28 females), and 60 percent of the NAT group (29 males and 30 females) agreed to participate. Because the numbers in the REG and NAT groups who accepted the invitation exceeded the number necessary for pairing, many volunteers were not utilized in this study. In every case, the REG student who most closely matched an EEP student, or in cases of "ties," the REG with higher college readiness scores was retained. NAT students were randomly selected to equal the proportions of EEP male and female students at each class level.

The data available to test for differences between responders and non-responders were SAT-equivalent test scores. No differences were obtained between EEP students who accepted or declined to participate; the same was true of QUALs. REG students who accepted the invitation to respond scored higher on the SAT-M equivalent (p = 0.05). Data permitting this kind of comparison were not collected for the NAT group.

Intellectual and Academic Achievement Variables
Table 3 presents data for CMT total scores and various academic performance measures obtained from university transcripts. Basically, on the CMT, the EEP, QUAL, and NAT groups scored higher than the REG group. EEP students also scored much higher on the CMT than a sample of 117 normal college juniors from a large midwestern university [21]. Apparently, then, the two groups which tested high on college readiness tests at an early age (EEP and QUAL), and the one which entered college already identified as having unusually effective ability (NAT) were indeed psychometrically superior to more typical college students. The NAT group, in addition, outscored both the EEP and QUAL samples (p<0.05). It is not surprising that the very brightest of college-age students pursuing education in an age-graded fashion were able to maintain an edge over the bright, but much younger EEP group on the verbal ability test. With respect to college performance, the CMT data were reassuring. In short, to the extent that university-level work employs reasoning abilities such as those tapped by the CMT, the EEP group appeared better qualified than typical college-age students (p < 0.001).

No significant differences among the groups were obtained on the four SSHA scales. This indicated that EEP students' attitudes toward study and knowledge about study skills were comparable to those of the other students, including the academically oriented National Merit finalists. Tables provided with the 1984 manual supplement confirmed that EEP students' scores on the four scales were, on average, well above the 50th percentile of the normative sample of 4579 freshmen at eight four-year colleges. Females in all groups scored higher than males on the SSHA (p = 0.004). The percentile scores of EEP males on the four scales of the SSHA were between 50 and 60; among EEP females, the respective percentile scores were between 60 and 75.

Table 3
Selected Ability and Academic Achievement*

In the analyses of variables extracted from academic transcripts in June 1984, four EEP students were not included because they had transferred to other universities. The picture that emerged was similar to that already presented for the CMT. EEP students earned a mean cumulative GPA comparable to that of the NAT group, confirming preliminary results published by Janos and Robinson [17]. This similarity obtained in humanities, social sciences, natural sciences, and honors courses, and at the freshman, sophomore, junior, and senior levels. On the other hand, EEP students earned higher GPAs than REG students (p< 0.001), a difference replicated in every distribution area and at every level except in senior level courses. Once again, the REG sample appeared representative. The mean GPA for EEP students was also consistently higher than the undergraduate UW average. For example, for the 24,752 undergraduates enrolled at the UW during the Autumn Quarter 1984, the mean GPA was 3.01 on a 4-point scale (SD=O.58).

The differences in GPAs reported above can be most meaningfully considered in the context of generally similar patterns of course enrollment followed by the three college groups. EEP students were not taking fewer courses. There were no differences among the groups in total number of UW undergraduate credits earned by June 1984. EEP and NAT students had completed more credits at the freshman level than the REG students (p = 0.001). One reason may be that the EEP students had, with staff encouragement, sampled a greater variety of introductory level courses; NAT students, presumably, had completed more courses at the 100-level because that is how UW Honors Program "core curriculum" seminars are numbered.

In counseling, staff have been cautious in advising EEP students not to "get in over their heads," and entrance into the honors options is encouraged only after careful evidence of need and desire for more demanding courses have been manifested. Honors courses often demand a maturity of intellectual functions that mayor may not be truly mastered by the very young, however bright they may be. When careful consideration precedes involvement in honors courses, EEP students have tended to do well. In any case, it is a minority of even the NAT students whose transcripts reported either type of honors options.

Patterns of enrollment reflected similar gender differences among all three groups. Among EEP students, females took more social science courses than males, who made up the difference by taking more engineering and computer science courses (including computer programming). Among the EEP group as a whole, however, majors for both sexes evidenced a fair degree of balance among a wide variety of departments, including the following partial lists: Males majored in accounting, anthropology, ethnomusicology, history, philosophy, political science, and psychology, as well as chemistry, computer science, engineering, mathematics, and physics. Females majored in physics, microbiology, mathematics, electrical engineering, chemistry, and biology, as well as English, political science, psychology, religious studies, and Southeast Asian studies.

Psychosocial Variables
We would like to consider the Total Positives scale of the Tennessee Self-Concept Scale first, briefly, because it provides the best overview of this section and because the greater detail supplied by the CPL seems worth elaborating. All groups scored comparably on the Total Positives scale of the Tennessee Self-Concept Scale, close to the mean for T-scores. In short, the TSC provides evidence that all groups are best characterized as normal and healthy in psychological adjustment.

Scores obtained on the first administration of the CPI are listed in table 4. The groups differed significantly on scales labeled Social Presence, Responsibility, Socialization, Self-Control, Achievement via Conformance, Achievement via Independence, and Intellectual Efficiency. The differences did not arise, however, in comparisons between EEP students and the other groups taken singly. It was necessary to use a stringent probability level because of the large number of statistical comparisons executed. Consistent with findings on the CMT, NAT students would have seemed to score higher on the Intellectual Efficiency scale without using Dunn's correction, but in no other instance did differences between the EEP and NAT or EEP and QUAL groups approach even a lenient criterion of statistical significance. In comparisons with the REG group, however, EEP students would have appeared to score higher on Responsibility, Good Impression, and Achievement via Independence, and lower on Social Presence and Self-Acceptance, but these differences disappeared with Dunn's correction. If these last differences are not, in fact, due to chance, they would imply that EEP students were somewhat less assertive than typical college students, corroborating earlier findings [32].

Of particular interest is a Social Maturity Index derived and validated by Gough [9]. The index involves positive weightings for Socialization, Responsibility, Flexibility, and Dominance, and negative weightings for Good Impression and Communality. High scorers tend to be described as dependable, foresighted, and capable. In this study, females scored higher on this index (p = 0.007). Group differences were also significant (p< 0.05). As observed on the cognitive and academic measures, the NAT group, notably the females, scored higher than the other groups; QUAL females also tended to score high. EEP students scored about as well as others outside those two highest-scoring female groups.

Ninety-six students completed the CPI more than once. Overall CPI scores for all groups increased across two testings separated by nearly a year (p = 0.008), which appeared attributable to rises in mean scores on Responsibility, Self-Control, Tolerance, and Good Impression. One might expect that similar rises would be observed on the Social Maturity Index, which, as mentioned above, has been regarded as a particularly good index of the adequacy of social functioning. However, scores were found to be relatively stable from first to second testing in this study.

Because the Inventory of Parent and Peer Attachment specifically tapped the social domains in which concerns for the development of early college entrants have most often been voiced, the scores from its repeated administrations were analyzed to compare the EEP group with each of the three comparison groups. Even using this method, which maximized the probability of finding differences, not a single significant one was obtained. EEP students' responses indicated that they trusted and communicated with parents and peers much as did comparably bright agemates, bright older college students, and typical college students. There was no evidence of their being more alienated from parents or peers than the other groups. Attachment patterns such as these have been positively related to psychological well-being, particularly to enhanced self-esteem and life satisfaction and to an increased ability to cope constructively with stress [11].

Table 4
California Personality Inventory Scores (First Administration)*

Scores on the Defining Issues Test, like overall CPI scores, rose for all groups, Second "P" scores, taken about a year after the first, were significantly higher (p < 0.001), but none of the groups changed more than the others. "P" scores were, as one might expect, correlated with CMT scores, although the correlation was low (r = 0.18; p- 0.006) in this restricted range. With CMT scores entered as a covariate, no significant differences among groups on either first or second DIT "P" scores were observed. Females scored higher than males (p < 0.02). EEP students scored at almost exactly the same levels as comparably bright agemates, who were, presumably, undergoing the socially and cognitively maturing experiences of high school. That EEP students also scored comparably to NAT and REG students suggested that they were developmentally "in sync" at college. The EEP male mean on the first administration of the DIT was equivalent to the mean reported for 2479 college students by Rest [30], and the EEP female mean on the first administration was equivalent to the mean Rest reported for 183 graduate students. EEP students' scores at the first testing were also comparable to those reported by Keating [21] for 117 juniors at a large midwestern university. At second testing, EEP students of both sexes scored at the level reported for a normative sample of graduate students [30].

The groups differed significantly on the Achievement Orientation and Moral/Religious Orientation scales of the Family Environment Scale (p = 0.03). EEP students scored lower than REG students on Achievement Orientation (p = 0.002), and they also tended to score lower on Moral/Religious Orientation (p = 0.09). Although the first difference might seem counterintuitive, Moos (personal communication) has emphasized that the FES Achievement Orientation scale measures conventional attitudes which would probably not be valued within families pursuing nontraditional educational options. In short, the FES yielded no indication that EEP families were more controlling, pressuring, or conflictful than families of comparison students.

Clinical Impressions
Although we made strenuous efforts to identify outcome variables tapping a wide range of psychological functioning, it might also be useful to share briefly some composite impressions formed in working closely with these EEP students in the course of more than ten years. In the first place, we have been struck by the enormous variability these young people have presented. Seventeen of the 43 impressed us with their outstanding brilliance, clarity of purpose, self-discipline, inner harmony, interpersonal adequacy, and apparently smooth and easy development. At the other end, seven appear to have struggled strenuously--for at least a while--in all of these domains, and to have been compromised by marked personal and familial vulnerabilities. The balance of EEP students appear to have experienced the normal wear and tear of adolescence, bringing to their adaptation normal assets and vulnerabilities in non-intellectual domains. They have freely reported difficulties associated with their EEP experiences, typically with respect to a lack of preparation for science (a problem that the Program continues to address in various ways), and a frustration with having had to make career decisions on an accelerated timetable, although it is a rare few who said they wished they had gone to high school. Many entered the EEP with serious problems of adjustment in their previous school programs, and it appears likely that, for them, the supportive aspects of the Program have buffered what might have been a more problematic development elsewhere.

By 1987, 27 of the 43 EEP students in this study had graduated, 24 proceeding to selective graduate or professional schools. One was simultaneously employed in a high-paying job in the defense industry, and another had elected to pursue a second undergraduate degree in Japan. Two were working and developing plans to attend graduate school. Among the others, the great majority were continuing to progress with college studies; only one student had dropped out because he had found studies unfulfilling and because he needed to support his young family. By August 1987, 2 EEP students had earned Ph.D.'s, one at age eighteen (in mathematics), and one at age twenty-three (in geophysics); both are postdoctoral fellows at first-rate universities.

It is our impression that QUAL students were more likely to attend colleges or universities more prestigious than the UW, with possible associated advantages, including eligibility for scholarships. Enrollment in the EEP had not, however, precluded 10 of the 43 EEP students from transferring, at some point during their undergraduate careers, to one of the following universities: Brown University; California Institute of Technology; Duke University; Milliken University; Princeton University; Reed College; the University of California, Berkeley; University of Maryland; and University of Minnesota. Seven of the 22 Graduates, in fact, obtained their degrees other than at the UW. It was also our impression that EEP students were more likely to transfer to other institutions than were REG or NAT students.

In the course of ongoing evaluation of the Program, we have systematically sought feedback from students, parents, UW faculty, as well as therapists of a half dozen students who have sought outside counseling. Unfortunately, these studies have lacked comparison groups or suffered from other inadequacies of method, and so do not appear to warrant extensive description. In brief, the average EEP student communicated feeling satisfied with the choice to enter college early. In response to questionnaires, an overwhelming majority of parents (about 95 percent) have reported satisfaction with their child's progress, not only academically, but also with respect to personal and social development and quality of relationships within the family. The therapists have unanimously regarded college as an asset to their clients' coping, despite reservations about the concept of "early entrance" in general. Faculty typically have had a hard time distinguishing early entrants from regular-age college students, and they have almost always been greatly surprised to learn EEP students' ages. Assertiveness appears to be one of the few respects in which the faculty have suggested that EEP students differ from regular-age college students, in that they see younger students as asking fewer questions in class, but more often making contact out of class. In the wave of data collection reported here, the topics of dating and living situations were virtually ignored. This defect will be remedied in studies currently underway. However, early entrants face unusual challenges in both of these domains. The young men, not unexpectedly, have had difficulty attracting the romantic interest of college-age females, whereas several of the young women have become so involved in such activities that their studying suffered. EEP youngsters of both sexes, many more than their high-school peers, tend to move out of family settings early, typically between the ages of fifteen and seventeen. Problems of various kinds, such as transitory dropping of grades, have sometimes been associated with such moves, but again, the consensus of students, parents, and program staff has generally been that the gains have outweighed the losses.

Discussion
This article began by reviewing follow-up studies of early college entrants. Certainly, in historical terms, early entrance has been under-taken relatively rarely, and then after more careful and continued evaluation than attends most other educational decision making. Perhaps because of such extreme caution, follow-up reports on early entrants, although few in number, are in virtual consensus that academic and career gains far outweigh the transitory distress associated with making social adjustments to college. Naturally, more and better studies of these complex issues are in order. Nevertheless, the studies suggest that it is time for secondary and higher education administrators to place a higher priority on solving the problems of which highly able learners bitterly complain, such as the inadequacy of "chronologically appropriate" instruction, and their profoundly painful sense of alienation from so-called peers. Those highly capable youths who experience such problems in severe degree for the duration of high school may well be at greater risk for impaired development than those who find at least partial fulfillment in college at an early age. The current study addresses one approach--certainly not the only one--to providing options for such students.

The current study depicted patterns consistent with the overview: Youngsters who had entered college markedly early exhibited higher ability test scores and grades than more average college students in a large state university, although they did not attain quite the levels of performance, or complete as many honors courses, as did National Merit finalists. The study found no association between early entrance and psychological or social impairment. Indeed, in every comparison, the early entrants were virtually indistinguishable from comparably bright agemates who had elected to attend high school. There is a suggestion that they may be somewhat less assertive than average college students, but this characteristic may derive as much from an introversive tendency observed among academically focused individuals [18] as from their youth. A modest fraction of early entrants exhibited compromised adjustment, but so did subjects in each of the other groups. Furthermore, even for these troubled early entrants, college appeared to be a constructively integrating experience, both from our perspective and that of therapists outside the Program who were working with the students.

From its inception the Early Entrance Program has attempted to address the presumed needs of most of its participants for affiliation with others like themselves-needs for others not merely the same age, but also bright, exploratory, and willing to work hard to learn. By providing conditions under which such youngsters work and play together for extended periods of time, the EEP has been able to foster social contact and release the natural gregariousness of teens, many of whom had been previously alienated and unhappy. Year in and year out, we have observed the flourishing of mid-adolescent peer culture and its branching out to include older university students as well. With each other, the early entrants appear to find a communality permitting genuinely mutual exchange. About half of them had experienced such friendships before, but in so many cases, the unique opportunity of enrolling in college along with other thirteen-or fourteen-year-olds as academically inclined as they were, opened up, for the first time, the world of intimate friendships.

Such emotional and social growth demands time, and for many students it takes energy away from study. For that reason, a significant number of early entrants have earned somewhat lower grades than they might, settling, for example, for "3.0s" when they might easily earn "3.5s." Of course, the same phenomenon is observed among typical college students. Is this "immaturity" or a "normal" process? The answer, of course, is that it is probably both. But we do not view it as the kind of immaturity that should preclude a college experience for very young people; like everyone else, they have to make important decisions and live with the consequences. Over the years, we have concluded that, as a rule, they are better equipped to do so for themselves than are others to do it for them. When battles are avoided, and support is readily available, they seem to seek and use it frequently.

There can be no doubt that this study raises more questions than it answers. It cannot illuminate how these students will perform as adults, personally and in their careers. However confidently one may predict later functioning from established patterns of constructive adaptation and growth, follow-up studies are clearly in order. Furthermore, there may be systematic differences in adaptation patterns, not only between males and females, but also between those talented in different cognitive domains (for example, verbal and quantitative), about which we would like to learn mole. Research on individual differences in motivational characteristics and work styles also appears fascinating and potentially fruitful. From a scientific standpoint, the area is rife with virgin terrain for exploration.

This study provokes not only additional questions, but stem cautions. Our data were supplied by carefully selected students participating in a comprehensive program of support. In the first place, before entry, these young students evidenced academic readiness, defined by objective measures such as college aptitude tests; intense personal commitment, often signaled by overcoming major obstacles to college admission; and the signs of maturity that led us to believe they might keep themselves effective, on-task, and socially at ease while in college. Second, support was available from the moment a student entered the EEP. Each year, a sizable group of full-fledged peers was congregated to furnish a social support; students also were offered ongoing psychological, academic, and career counseling. Such extensive support, of course, may not be essential for any given student; nevertheless, some such provisions should probably accompany the institution of any large-scale policies to facilitate markedly early college entrance.

Any worth while evaluation of markedly early college entrance must acknowledge the unique needs of the individuals who seek to exercise the option. Their quality of life during adolescence, as well as their productivity and satisfaction as adults, is at stake. For a substantial proportion, adequate intellectual challenges simply cannot practically be provided within public school systems, except in perhaps the largest metropolitan areas. Furthermore, some individuals may be further harmed by keeping them in regular schools, under the assumption that social development will be fostered merely by propinquity with agemates, despite evidence that, for them, constructive social interactions are generally lacking. For this group, viable alternatives to early entrance are few. Although many very bright youngsters adapt to less-than-optimal academic fare without suffering profound limitation and may find satisfying peer relations in high school, pronounced wishes for college studies by highly motivated, well organized, and academically ready young people argue convincingly for programs designed to facilitate this option.

*Please refer to original version for all tables.

It is thanks to the funding of the William H. Donner Foundation that this research was conducted and that the Early Entrance Program at the University of Washington survived the death, in 1981, of its founder. Halbert B. Robinson. Thanks are also due to the unbelievably faithful former staff of the Center for the Study of Capable Youth, especially Greg Sullivan, Beth Manger, and Janis Short, who saw to it that no administrative snafu long delayed this project: and to Chris Cassel, Dan Schluter, Mark Heinzig, and Alicia Wise, who enthusiastically saw to the more tedious aspects of data management. Many thanks to Redmond Reams for his ongoing input regarding methodology and statistical analyses.

References

Brown, W. F., and W. H. Holtzinan. Survey of Study Habits and Attitudes. New York: The Psychological Corporation, 1965.

Brody, Linda E., and C. P. Benbow. "Accelerative Strategies: How Effective Are They for the Gifted." Gifted Child Quarterly, 3 (1987),105-10.

Center for the Advancement of Academically Talented Youth (CTY), Johns Hopkins University, The 1987 CTY Talent Search Report. Baltimore, MD., CTY, 1987.

Daurio, S. P. "Educational Enrichment Versus Acceleration." In Acceleration and Enrichment: Strategies for Educational Change, edited by W. George, S. J. Cohn, and J. C. Stanley. Baltimore, MD: Johns Hopkins University Press, 1979.

Fitt, W. H. Tennessee Self-Concept Scale. Nashville: Counselor Recordings and Tests, 1965.

Fluitt, J., and S. Strickland. "A Survey of Early Admissions Policies," College and University, 59 (1984), 129-35.

Fund for the Advancement of Education of the Ford Foundation. Bridging the Gap Between School and College. New York: Research Division of the Fund, 1953.

Fund for the Advancement of Education of the Ford Foundation. They Went to College Early. New York: Research Division of the Fund, 1957.

Gough, H. G. "Appraisal of Social Maturity by Means of the CPI." Journal of Abnormal Psychology, 71 (1966), 189-93.

______Manual for the California Psychological Inventory, Palo Alto, Calif.: Consulting Psychology Press, 1969.

Greenberg, M., J. Siegel, and C. Leitch. "The Nature and Importance of Attachment Relationships to Parents and Peers during Adolescence." Journal of Youth and Adolescence, 12 (1983), 373-86.

Hollingworth, L. Children above 180 IQ Stanford Binet: Origin and Development. Yonkers: World Book, 1942.

Holland, J. L. "The Undergraduate Achievement of National Merit Scholars and Certificate of Merit Winners." Bulletin of the National Association of Secondary School Principals, 44 (1959), 140-42.

Hull, C., and N. Nie. SPSS Update 7-9. New York: McGraw-Hill, 1981.

Janos, P. "The Self-perceptions of Uncommonly Bright Youngsters: Clinical Inquiries Guided by Informal Social Comparison Theory." In Perceptions of Competence and Incompetence across the Lifespan, edited by J. Kolligian and R. Sternberg. New Haven, Conn.; Yale University Press, forthcoming.

______. ”A Fifty-Year Follow-up of Terman's Youngest College Students and IQ-matched Agemates.” Gifted Child Quarterly, 31 (1987), 55-58.

Janos, P., and N. Robinson. "The Academic Performance of Students in a Program of Extraordinary Academic Acceleration." Gifted Child Quarterly, 29 (1985). 175-80.

______."Social and Personality Development of the Gifted and Talented." In The Gifted and Talented: Developmental Perspectives, edited by F. Horowitz and M. O’Brien. Washington, D.C.: American Psychological Association, 1985.

Janos, P., et al. "A Cross-sectional Developmental Study of the Social Relations of Students Who Enter College Early," Gifted Child Quarterly, 32 (1988), 210-16.

Janos, P., S. Sanfilipo, and N. Robinson. "Patterns of 'Underachievement' in Markedly Accelerated University Students." Journal of Youth and Adolescence, 15 (1986), 303-13.

Keating, D. P. "A Search for Social Intelligence." Journal of Educational Psychology, 70 (1978), 218-23.

Moos, R. H., and B.S. Moos. Family Environment Scale. Palo Alto: Consulting Psychologists Press, 1981.

Nichols, R. C. and A. W. Astin. "Progress of the Merit Scholars: An 8-year Follow-up." Personnel and Guidance Journal, 44 (1966), 673-86.

Nie, N., et al. Statistical Package for the Social Sciences. New York: McGraw-Hill, 1975.

Noeth, R. J. Washington Pre-College Counselor's Guide. Seattle: Washington Pre-College Testing Program, 1978.

Ogburn, M., and N. Colangelo. "Test Performances of Superior Students on the Concept Mastery Test: 1965-1976." Gifted Child Quarterly, 23 (1978), 79-84.

Pollins, L. "The Effects of Acceleration on the Social and Emotional Development of Gifted Students." In Academic Precocity: Aspects of its Development, edited by C. P. Benbow and J. C Stanley. Baltimore: Johns Hopkins University Press, 1983.

Pressey, S. "'Fordling' Accelerates Ten Years After." Journal of Counseling Psychology, 14 (1967), 73-80.

Rest, J. Manual for the Defining Issues Test. Minneapolis: University of Minnesota Press, 1979.

______. Development in Judging Moral Issues. MinneapoIis: University of Minnesota Press, 1979.

Robinson, H. B. "A Case for Radical Acceleration: Programs of the Johns Hopkins University and the University of Washington." In Academic Precocity: Aspect of it's Development, edited by C. P. Benbow and J. C. Stanley. Baltimore: Johns Hopkins University Press, 1983.

Robinson, N. M., and P. M. Janos. "The Psychosocial Adjustment of Youngsters in a Program of Marked Academic Acceleration." Journal of Youth and Adolescence, 15 (1986), 51-60.

Robinson, N. M., and H. B. Robinson. "The Optimal Match: Devising the Best Compromises for the Highly Gifted Student." In New Direction for Child Development: Developmental Approaches to Giftedness and Creativity, edited by D. FeIdman, (No.17), San Francisco: Jossey-Bass, 1982.

Stanley, J. "How Did Six Highly Accelerated Gifted Students Fare in Graduate School." Gifted Child Quarterly, 29 (1985), 180.

_____."Young Entrants to College: How Did They Fare?" College and University, 60 (1985), 219-28.

Stanley, J. and C. Benbow. "Extremely Young College Graduates: Evidence of Their Success." College and University, 58 (1983), 361-71.

Stanley. J., D. P. Keating and L. Fox. Mathematical Talent. Baltimore: Johns Hopkins University Press, 1974.

Terman, L. M. The Concept Mastery Test Manual, Form T. New York: Psychology Corporation, 1956.

Terman, L. and M. Oden. Generic Studies of Genius: Volume 4. The Gifted Child Grows Up. Stanford: Stanford University Press, 1947.

Permission Statement

Permission to reprint this article has been granted to The Davidson Institute for Talent Development by the Ohio State University Press, copyright, 1989.

This article is provided as a service of the Davidson Institute for Talent Development, a 501(c)3 nonprofit dedicated to supporting profoundly gifted young people under 18. To learn more about the Davidson Institute’s programs, please visit www.DavidsonGifted.org.