Breaking Down the Forgotten Half

Abstract

This article presents a research study in Quebec’s adult education centers (AECs), which are now an alternative to general education that allow youths ages 16 to 24 to make up for academic delays or to return to school after dropping out. Data from 386 such students were collected on various personal, psychosocial, and school indicators. Using a latent class analyses, the author identified four different profiles with unique needs. The first two profiles were highly representative of emerging adults who were typical except for their academic delays. Youths in the last two profiles were characterized by victimization and problem behavior, respectively; they demonstrated the need for support that goes beyond the academic and school services usually offered in AECs. The results of this study are discussed in light of new parameters of emerging adulthood and discontinued educational pathways.

Keywords

at-risk students descriptive analysis dropouts identity structural equation modeling

The transition to adulthood has received increased attention over the past 10 years, due in part to the “emerging adulthood” developmental period proposed by Arnett (2000), who suggests that the prolonged transition to adulthood is a period that offers a longer moratorium for identity exploration. One aspect of this newly identified period is the central value attributed to education and, even more so, to postsecondary education. Colleges and universities thus constitute unique settings that provide opportunities for the exploration, skill attainment, self-discovery, and self-mastery on which students can build as they gradually become autonomous (Côté, 2002; Sacker & Schoon, 2007). However, those without such structural opportunities may become more vulnerable than is usual in the transition to adulthood. In fact, the gap between at-risk youths, who lack the personal and social resources to successfully face the challenges of adulthood, and those youths who possess these assets is believed to widen during the critical period of emerging adulthood (Berzin, 2010; Masten et al., 2004). Although this hypothesis is logical, there is little empirical evidence to support it because most of the studies aiming to understand development and adjustment during emerging adulthood have been conducted in “normative” postsecondary settings. Less is known about those youths who either drop out or continue their educational pathways through second-chance settings due to various difficulties, those whom Halperin (1998) named “the forgotten half.”

Dropping Out and the Discontinuity of Educational Pathways

Dropping out and early school failure often pave the way for failure across a wide range of areas; problems that arise include unemployment, involvement in crime, and substance use, as well as unsatisfactory personal relationships (Foster, Flanagan, Osgood, & Ruth, 2005). Quebec has one of the highest dropout rates (25.3%) in Canada (Institut de la statistique du Québec, 2009), which has led to several investments in the areas of prevention and research of the processes of dropping out. Reliable and repeated research studies have revealed that dropping out is a phenomenon attributable to a set of individual, social, family, cultural, socioeconomic, and institutional factors (Archambault, Janosz, Fallu, & Pagani, 2009; Fortin, Marcotte, Potvin, Royer, & Joly, 2006).

In this line of work, Fortin et al. (2006) identified four profiles of at-risk youths who are likely to exit school without a diploma. Students categorized as the “uninterested in school” type (U type) displayed depression symptoms and a serious lack of motivation for school work, although their school performance was around average and teachers perceived them positively. The “school and social adjustment difficulties” type (SSAD type) of student combined poor school performance and high levels of internalizing and externalizing problem behavior. The “depressive” (D type) and “antisocial covert behavior” (ACB type) types also emerged from the work by Fortin et al. (2006). The former is characterized by high-achieving youths who display a high incidence of depression symptoms and come from highly problematic family environments. The ACB type is represented by students who are well adjusted to the school environment and expectations (performance, teacher’s perceptions) but exhibit delinquent behavior and come from chaotic family environments.

Fortin et al.’s (2006) study indicated that poor academic achievement, a powerful predictor of dropping out documented in earlier studies (Battin-Pearson et al., 2000), was less significant than depression (or internalizing problems), which seems to be a central aspect for all the profiles and has varying influence among the categories. The identified profiles comprised students who were at risk of dropping out; however, no data were available to indicate how many of these youths did in fact drop out. Also, given the particularities of Quebec’s educational system and the availablity of second-chance education (discussed in the following section), it is possible that many instances of dropping out could have been categorized more acurately as interruptions because many students choose to reenter school by means of second-chance schools (Bélanger, Carignan-Marcotte, & Staiculescu, 2007). In fact, interruptions are common in traditional postsecondary educational pathways (college and university): Almost 44% of 20- to 22-year-olds put their studies on hold for several reasons (work, travel, etc.) and return to school months or years later (Charbonneau, 2006; Statistique Canada, 2008, 2010).

The Reality of Returning to School at Quebec Adult Education Centers

There are more than 200 adult education centers (AECs) in Quebec’s provincial public educational system. These centers are open to any individual age 16 or older who wishes to complete the secondary level and obtain a first (high school) diploma. When the graduation rate of 24-year-olds among the Canadian population is considered, the AEC system is believed to be responsible for reducing the gap between Quebec’s dropout rate and those of the other Canadian provinces (Statistique Canada, 2010).

In spite of being open to adults of all ages, Quebec’s Ministry of Education has noted a significant decrease in the average age of AEC attendees over the past 10 years (Ministère de l’Éducation, du Loisir et du Sport, 2009). Almost 55% of students attending AECs are under the age of 24. These new parameters represent a drastic change that requires essential adjustments in service delivery by AECs (Ministère de l’Éducation, du Loisir et du Sport, 2009). For half of the 16- to 24-year-olds attending AECs in Quebec, general education in the adult sector is an alternative to general education in the youth sector and allows them to make up for academic delays. These students usually make a direct transition from the youth sector to the adult sector. The other half is composed of returning dropouts who want to obtain their high school diploma. It is believed that the younger adults in AECs—those with academic delays or who dropped out—often share an educational journey punctuated by school difficulties, learning disabilities, and/or adaptation problems. Bearing in mind the number of youths in AECs who drop out for a second time, and the high levels of absenteeism reported in these school settings, it is probable that the difficulties that marked past school experiences and sometimes contributed to the decision to leave still have repercussions on these youths’ adjustment in AECs (Marcotte, 2008, 2009; Marcotte, Cloutier, & Fortin, 2010).

Faced with this complex reality, Quebec’s Ministry of Education (2009) has clearly identified the need to improve the complementary educational services so that they are better adapted to the current needs of the young adult population enrolled in AECs. However, even it is believed that these youths are dealing with a lot of psychosocial difficulties, no studies have established the various psychosocial characteristics and needs of the large number of youths who have entered AECs over the past 10 years. In fact, most of the information gathered on these students comes either from youth-sector student records or from psychological or behavioral files that do not always reflect the students’ actual developmental challenges as young adults. This kind of knowledge is of particular value given the role played by adult education in Quebec’s overall graduation rate and the positive alternative it represents to dropping out permanently. On a larger scale, knowledge about emerging adults who do not follow the patterns of mainstream youths is needed to break down information on the “forgotten half” and gain a better understanding of their particularities and needs (Schwartz et al., 2011). In fact, we believe that these youths, who are often considered an at-risk or maladaptive group, may well display distinctive profiles that need recognition and adjusted interventions.

The present study attempts to break down information on the emerging adults attending second-chance schools by exploring groups according to the past and present psychosocial indicators of their pathways. Using statistical procedures similar to those of Fortin et al. (2006), my goal is to identify subtypes of youths attending AECs in Quebec and, when possible, underscore links between the two studies. Using a person-oriented approach rather than a variable approach, this study assumes that the profiles of students in second-chance schools must be viewed in a holistic manner in which past and present psychosocial characteristics actively contribute to their particular educational situations (Berzin, 2010). Moreover, as Tinto (1997) suggests, the influence of time relativizes the importance of some characteristics or emphasizes others. In fact, during the delicate transition to adulthood, I postulate that identity development and its underlying processes are of interest when young adults and late adolescents are considered. Indeed, a wealth of research has highlighted the links between a clear sense of self (commitment) or an active search toward it (exploration) and positive school, social, and personal outcomes in young adults (Marcotte, 2009; Schwartz, et al., 2011; Schwartz, Côté, & Arnett, 2005). Arnett (2000) argues that identity formation is a central developmental task of emerging adulthood, not of adolescence. Studies based on Marcia’s (1966) model of identity statuses suggest that high levels of exploration combined with high (achieved status) or low (moratorium status) levels of commitment are healthier than low levels of exploration combined with high (foreclose status) or low (diffuse status) levels of commitment (Archer, 2008; Berzonsky & Kuk, 2005). Considering the hypothesized central function of identity, I chose to include this variable in the identification of second-chance students’ psychosocial profiles.

The study is guided by the following hypotheses: (1) Multiple profiles with differential degrees of difficulties will emerge from past and present psychosocial factors among 16- to 24-year-olds enrolled in second-chance schools, (2) a number of the profiles identified will share similarities with Fortin et al.’s (2006) dropout risk types, and (3) the profiles will differ significantly in terms of past and actual school adjustment measures.

Method

Sample and Procedure

The sample is composed of 386 youths, 16 to 24 years old (184 male and 202 female), enrolled in the secondary education instructional services of one of the AECs in three regions of Quebec, Canada: Quebec City, Mauricie, and Estrie.¹ Of the youths in the sample, about one third are 16 to 17 years old (n = 105) and two thirds are 18 to 24 years old (n = 281).² The majority of students (87%) are Caucasian and born in Canada. African Americans and Latino Americans accounted for 13% of the sample. This ethnicity portrait³ was representative of the regions that were chosen for this study due to the locations of their urban and rural AECs. Among the 22 AECs solicited, 7 participated in the study. A person from each location publicized the study in the school and invited students to participate. All students 24 years or younger were eligible to participate. However, youths in literacy programs could be included only if their reading skills allowed them to complete the research protocol. A collect day was then scheduled in each school; all interested students were assembled in the same room to fill out the questionnaires. The youths signed an informed consent form in the presence of the research assistant before completing the questionnaires. Completing the questionnaires took between 45 and 75 minutes, depending on the student, and $25 (CAD) was given to each participant as compensation.

Measures

The heterogeneity of the “young adults” who attend the AECs in Quebec can be explained by both the variety of family and personal risk factors that are present during childhood and adolescence and the current psychosocial characteristics that have an effect on the way young emerging adults experience their educational pathways. The variables were deliberately chosen to take into account the past and current psychosocial realities of these youths. The variables were chosen according to their relevance in the literature pertaining to youths in adult education and the developmental assets needed by emerging adults to take on the developmental tasks of this period of life (Schwartz et al., 2005). Finally, I deliberately selected psychosocial indicators over school indicators in the classification procedure, following Fortin et al.’s (2006) results suggesting that psychosocial characteristics bear more significance in differentiating at-risk students than do school indicators, although the former are linked with the latter. As a result, it was decided that psychosocial measures would be used to classify youths and school measures, whereas other indicators would be used to validate the classification (provided that the first procedure yielded significant results). Moreover, this selection of variables is consistent with the need identified by the Quebec Ministry of Education for a better understanding of the psychosocial profiles of the youths attending AECs.

Past traumatic events

A French version of the Child Trauma Questionnaire (Bernstein, Fink, Handelsman, & Foote, 1994) was used to measure the traumas experienced during childhood. Two subscales were retained for the present analysis, that is, Physical and Emotional Abuse and Sexual Abuse. Bernstein et al. (1994) reported internal consistency coefficients between .79 and .96, test–retest reliability between .78 and .86, and good convergent validity with other measures (Paivio & Cramer, 2004).

Services received during childhood and adolescence

This nonparametric measure was created to check the type of services received by the youths during childhood and adolescence. Four categories of services emerged from this measure: (a) youths who were placed under the Youth Protection Act⁴ (YPA) and in a special class⁵ during their educational pathways (special class and out-of-home placement), (b) youths who were placed under the YPA (out-of-home placement), (c) youths who were placed in a special class (special class), and (d) youths who never received any of these services during childhood or adolescence (no services).

Externalizing problems

A French version of Achenbach and McConaughy’s (1997) and Achenbach, Dumenci, and Rescorla’s (2003) Adult Self-Report for adults 18 years or older or Youth Self-Report for youths younger than 18 was used to measure the youths’ externalizing problems. Test–retest reliability studies showed acceptable correlations (rs) between .82 and .93, according to the scales (Achenbach & McConaughy, 1997). The internal consistency shows alphas between .88 and .95 for the externalizing scale.

Psychological distress

Préville, Boyer, Potvin, Perrault, and Légaré’s (1992) Indice de Détresse Psychologique de l’Enquête Santé Québec comprises 14 items that are answered using a 4-point Likert-type scale. Items refer to the intensity of distress felt over the past 6 months. The convergent validity of the psychological distress index and its internal consistency (.88) are considered to be acceptable.

Self-esteem

This measure is a French version of the General Self-Esteem scale from the Self-Description Questionnaire (SDQ-II), which starts off with 11 scales. The SDQ-II has excellent construct validity (.87–.91) and was validated in several countries (Canada, America, and France). This scale includes 10 questions answered on a 6-point Likert-type scale regarding the way people see themselves. Among the SDQ-II scales, Self-Esteem is the most strongly correlated with General Self-Concept (Guérin, Marsh, & Famose, 2004).

Suicide risk

This measure corresponds to a dichotomous response (yes or no) given to the following question: “Over the past 12 months, have you seriously thought of committing suicide?”

Identity

The questionnaire on identity is a French version of Balisteri, Bush-Rossnagel, and Geisinger’s (1995) Ego Identity Process Questionnaire, which includes 32 items that are answered on a 6-point Likert-type scale. The scores are added to obtain a total score for engagement and a total score for exploration. The convergence, internal consistency, and test–retest reliability of the instrument were found to be acceptable (Balisteri et al., 1995).

Analytic Strategy

Latent class analysis

Latent class analysis (LCA) is a statistical procedure that presumes that the observed variables inserted in the analysis are indicators that discriminate between subgroups within the data (Hagenaars & McCuthcheon, 2002, as cited in Keller, Cusick, & Courtney, 2007). This analysis can be thought of as an “improved” cluster analysis that is based on the statistical concept of likelihood. The recent increase in interest in latent class models is due to the awareness that the use of such models can produce powerful improvements over traditional approaches to segmentation, as well as to clustering, regression, and other types of analyses. Latent classes are unobservable (latent) subgroups or segments. Cases within the same latent class are homogeneous on certain criteria, whereas cases in different latent classes are dissimilar from each other in certain significant ways. The LCA method reveals the associations between included variables by dividing the sample into classes, which are also called profiles. Between-class differences emerge in light of the singular probability profiles of the chosen indicators. The M-Plus program (Muthén & Muthén, 2000), used for this specific analysis, proposes an iterative procedure of maximal likelihood to optimize individuals’ classifications within the classes. The maximum likelihood ratio is used to estimate the model fit to the data (log-likelihood [LL] ratio), and the Bayesian information criterion (BIC) is used to determine which model is the most parsimoniously adjusted to the data. Finally, the chosen model and the number of classes retained must have a practical value and theoretical interpretability (Keller et al., 2007). Gender and age were entered as covariates in the analysis.

Validation of the classes

After the classes are identified, it is important to verify that the classes are differentiated in variables that were not used in the classification process. If such differences emerge, they may provide additional information on the nature of the classes. For the purposes of this study and considering the context in which the sample was taken, the variables used for validation were chosen on the basis of their association in the relevant literature (Barnard, 2004; Dickhäuser & Reinhard, 2006) with positive adult adjustment and academic success. These variables are taken from descriptive questionnaires that yielded no composite scores. These measures include questions on past experiences in school and on visions of the future.

Results

Model Selection

A four-class model was retained because it presented the solution that provided the most parsimonious fit to the data within the sample. The BIC was used to determine the model to be used. Table 1 describes the statistics for the one- to six-class models. The BIC related to the four-class model was the lowest: 17,904.56. The probability of attribution to a class was 96.1%, which demonstrates an excellent classification quality. In addition, the four-class model was theoretically relevant and easily interpretable with the data gathered relating to second-chance schools. By contrast, the three-class model incorporated both Classes 1 and 2 in the same category. Because the characteristics of these two classes reflected vital theoretical distinctions in the four-class model, the three-class model was rejected. The five-class model divided Class 4 (in the four-class model) into two subgroups that each contained a very small number of participants. Moreover, only one indicator had a modest contribution to the differentiation of these two groups, resulting in the dismissal of this model as well.

Table 1.

Statistics for the One- to Six-Class Models

Model Number	Number of Classes	LL	BIC	df	Classification Error
1	1	−9,856.57	19,820.35	18	0
2	2	−9,004.24	18,240.75	39	0.0197
3	3	−8,839.78	18,036.92	60	0.0461
4	4	−8,711.07	17,904.56	81	0.0397
5	5	−8,655.10	17,917.69	102	0.047
6	6	−8,609.64	17,951.85	123	0.0392

Note. LL = log-likelihood ratio; BIC = Bayesian information criterion.

Description of the Profiles (Classes)

All the indicators used to identify the classes contributed significantly to the model. Table 2 shows the indicator z scores by class, the Wald statistic, the p value, and R². It should be noted that age was not significantly associated with class identification.

Table 2.

Indicator z Scores by Class, Wald, and R²

	Class 1	Class 2	Class 3	Class 4
Covariates	(54%)	(21%)	(13%)	(12%)	Wald	p value	R ²
Age (ns)	18.98	18.69	19.09	19.28
Sex (% female)***	40	61	89	54
Commitment	1.51	−3.29	1.52	0.26	13.28	.00	.05
Exploration	−2.02	2.15	0.94	−1.08	9.74	.02	.03
Physical and emotional abuse	−19.31	0.08	20.17	−0.94	302.36	.00	.55
Sexual abuse	−2.81	−1.17	6.56	−2.58	225.36	.00	.71
Externalizing problems	−6.20	−1.26	0.03	7.43	99.66	.00	.24
Psychological distress	−6.46	−3.58	2.86	7.18	144.21	.00	.36
Self-esteem	3.90	4.40	−0.71	−7.59	70.28	.00	.30
Services
Special class and out-of- home placement	−0.47	0.06	0.35	0.05	38.21	.00	.05
Out-of-home placement	−0.78	0.54	0.61	−0.37
Special class	0.38	−0.50	−0.27	0.38
No services	0.87	−0.11	−0.69	−0.07
Suicidal ideations	−0.74	−0.88	0.63	0.99	47.30	.00	.26

***

p < .001.

Among the indicators, past sexual and physical abuse, present psychological distress, and self-esteem are, in order, those that contribute the most to the identification of the classes. By contrast, although significant, past services received, commitment, and exploration contribute the least to the distinction of the classes.

Class 1: Youths with no major problems (NMP)

The majority of the sample (54%) falls within this class (212 youths: 60% male and 40% female). This group is characterized by the absence of serious internalizing and externalizing problems (psychological distress and externalizing problem behavior). There are no evident past traumas (abuse) marking the histories of these youths, and they are underrepresented in the system services during adolescence and childhood. Also noteworthy, the low level of identity exploration combined with average levels of commitment is an indicator of high prevalence of the foreclose identity⁶ status within this subgroup. Finally, the high levels of self-esteem displayed by these youths seem to concur with the overall adaptive profile of this class.

Class 2: Open/resilient (O/R) youths

Of the sample, 21% fall into this subgroup (82 youths: 40% male and 60% female). This class is similar to the previous one (Class 1) in terms of current absence of major problems. However, youths within this class are more prone to identity exploration and are significantly less committed to school than are the other youths in the sample. The self-esteem score for this class is the highest in the sample, although 25% of the students were placed outside of their homes during childhood or adolescence, and one third received specialized services in school. The combination of social or educational services, which are usually associated with risks (Berzin, 2010), and high levels of self-esteem and identity exploration contributed to identifying resiliency within this group.

Class 3: Victimized and distressed (VD) youths

Of the sample, 13% fall into the third profile (47 youths: 89% female and 11% male). The youths in this class clearly have the most discernible histories of past trauma, as indicated by high levels of abuse (emotional, physical, and sexual). Also, these students currently have to deal with major internalizing problems such as psychological distress and suicidal ideation (31%). In terms of services received during childhood or adolescence, 36% were placed outside their homes (protection services) and 47% received specialized educational services.

Class 4: Problem behavior and highly distressed (DPB) youths

Of the sample, 12% fall into the last profile (45 youths: 46% male and 54% female). In spite of the fact that this class has the fewest members, it encompasses youths with the heaviest portrait of current difficulties. In fact, these youths combine the highest levels of externalizing problem behavior and psychological distress of the sample. In addition, they have the lowest self-esteem scores and are more inclined to suicidal ideations (49%) than are youths in other profiles. Among this subgroup, only educational services are noticeable, with almost half of the sample having received this type of service in the past.

Validation of the Classes

The first analysis (not shown) determined if there were differences between classes regarding the level of the courses given at the AECs. More precisely, I wanted to know if the classes were different according to the academic levels of the students. This preliminary analysis, F(3, 6) = 3.27, ns, was not significant, indicating that the classes were not different from one another regarding the academic levels of students as assessed by the AECs at their admission. This absence of differences shows that students within each class are enrolled in various educational levels⁷ and that each class is not, in that respect, different from another. Validation of the classes was then conducted using variables that are known to be relevant to the student’s experience in school, academic success or failure, and perseverance. More precisely, descriptive questionnaires of past schooling experience, involvement in school work, and views of the future were used to corroborate and further analyze the profiles. Table 3 presents all the variables used for validation purposes, along with the contrasts between classes. Table 3 also presents F or chi-square values for independence tests and the results of Tukey’s post hoc procedure to assess statistically significant pairwise comparisons.

Table 3.

Variables Used for Validation Purposes, F or χ² Statistic, and Significant Class Differences

	Class 1	Class 2	Class 3	Class 4	Total	Significance Test (F or χ²)	Class Differences
Children
Yes (%)	4.8	8.8	23.4	4.7	7.9	χ² = 19.09***	3 > 1, 2, 4
Dropped out before entering AEC
Yes (%)	45.5	55.6	84.8	70.5	55.3	χ² = 28.46***	3, 4 > 1, 2
Years before entering AEC (%)						χ² = 13.76*
Less than 1	26.5	35	5.6	23.3	23.8		3 < 1, 2, 4
1 or 2	33.7	32.5	27.8	36.7	32.8
3 or more	39.8	32.5	66.7	40	43.4		3 > 1, 2, 4
Past school experiences (% agree)
It was hard to make friends at school	17.3	13.4	42.6	37.8	22	χ² = 24.29***	3 > 1, 2, 4; 4 > 1, 2
I liked to participate in activities at school	57.1	60.5	44.7	37.8	54	χ² = 8.62*	1, 2 > 3, 4
Most teachers didn’t care about me	18.2	23.1	47.8	47.7	26.3	χ² = 28.97***	1, 2 < 3, 4
There were adults to whom I could talk	80.7	72.8	57.4	62.2	73.9	χ² = 14.77**	1, 2 > 3, 4
School was often a waste of time	23.4	30.5	55.3	51.1	32.1	χ² = 26.36***	1, 2 < 3, 4
I had friends to whom I could talk	85.1	80.2	59.6	75.6	79.8	χ² = 16.05**	1 > 2, 3, 4; 2, 4 > 3
Teachers helped me when I needed it	78.7	67.1	47.8	53.3	69.5	χ² = 24.30***	1 > 2, 3, 4; 2, 4 > 3
School was a place where I felt put aside	16.6	22	52.2	53.3	26.3	χ² = 43.92***	3, 4 > 1, 2; 1 < 2
Involvement in school work	27.11	27.53	28.75	24.34	27.08	F = 9.90	4 < 1, 2, 3
Visions of the future (% agree)
Getting a job depends on my work in school now	83.5	86.6	82.2	91.1	84.9	ns
I know enough work fields to make a choice	77.9	84	88.6	46.7	76.7	χ² = 28.73***	4 < 1, 2, 3
I will always have a boring job	7.1	6.1	6.7	20.5	8.4	χ² = 9.53*	4 > 1, 2, 3
I would like to go to college or university	71.8	75.6	70.5	64.4	71.6	ns
Regardless of my academic achievements, I will have a low salary	6.7	9.8	11.4	26.7	10.3	χ² = 16.00***	4 > 1, 2, 3; 1 < 2, 3
I know myself well enough to make a choice about my future job	84.2	89	88.9	71.1	84.3	χ² = 7.99*	4 < 1, 2, 3
What happens to me in the future depends on me	99.1	98.8	100	95.6	98.7	ns
I want to have children	77.5	78.8	82.2	75	78	ns
There are few things I can do to change important aspects of my life	14.4	15.9	22.2	33.3	17.8	χ² = 9.90*	4 > 1, 2, 3
I often feel helpless regarding life problems	24.9	22.2	56.8	75	33.9	χ² = 56.03***	3, 4 > 1, 2
I have little control over what happens to me	23.6	24.7	37.8	51.1	28.8	χ² = 16.16**	3, 4 > 1, 2
I can do almost anything when I put my mind to it	92.9	91.5	93	77.8	90.8	χ² = 10.55**	< 1, 2, 3
I have no solutions for some of my problems	14.7	14.6	26.7	54.5	20.7	χ² = 38.19***	4 > 3, 2, 1; 3 > 2, 1
I want to get married one day	44.2	40	68.2	35.6	45.1	χ² = 44.92***	3, 4 > 1, 2; 2 > 1

Note. AEC = adult education center.

p < .05. **p < .01.***p < .001.

Results of the validation procedure reveal significant differences on almost all of the variables, but mainly between Classes 1 and 2 and Classes 3 and 4. Results regarding past school experiences indicate that youths in Class 3 show longer periods away from school than do other youths. Almost all past experiences in school are significantly more negative for the less adaptive profiles (Classes 3 and 4) than for youths presenting more adaptive profiles (e.g., Classes 1 and 2). However, “having friends to talk to at school” and “receiving help from teachers when needed” differentiated Class 1 from Class 2; youths without problems had more positive experiences regarding these aspects than did O/R youths and VD youths, who had the worst past experiences of all the classes in those areas. Finally, the two adaptive profiles also differ significantly regarding the question “School was a place where I felt put aside,” with Class 1 faring better than Class 2. This item also differentiated Classes 1 and 2 from Classes 3 and 4 in the expected direction. By contrast, the measure of involvement in school work showed no significant differences between the profiles. Regarding “views of the future,” Class 4 shows differences on all items when compared with all other classes. In fact, DPB youths have a significantly darker vision of their futures than the others do. Regarding the items associated with the internal locus of control, Classes 3 and 4 seem to fare worse than Classes 1 and 2. Also the locus-of-control items “There are few things I can do to change important aspects of my life” and “I can do almost anything when I put my mind to it” differentiated Class 4 from the others.

Discussion

Second-chance schools are important contributors to Quebec’s graduation rate as well as for society’s growth (Looker & Thiessen, 2008; Ross & Gray, 2005; te Riele, 2000). However, little is known in Quebec about the youths enrolling in these schools except that they are believed to have followed a pathway punctuated by risks, academic difficulties, and failures and that they are different from their counterparts who enroll mainly in postsecondary education (Entwisle, Alexander, & Olson, 2004; Tyler & Lofstrom, 2009). The present study aimed to learn more about youths enrolling in such second-chance settings in Quebec. The general hypothesis guiding the study was that among this heterogeneous group, several distinctive psychosocial profiles reflecting unique pathways and needs would emerge. To test this hypothesis, an LCA was conducted.

Results of the LCA clearly demonstrated the heterogeneity of the 16- to 24-year-old students enrolling in AECs, thus supporting the first hypothesis. The most prevalent of the four distinct profiles identified appeared rather normal, with students presenting no significant problems other than an extended educational pathway. This profile (Class 1, NMP youths) seems to reflect the discontinuity and instability associated with the current longer transition to adulthood proposed by Arnett (2000, 2007). Furthermore, this profile can be conceptually related to the most prevalent profile identified by Fortin et al. (2006): the U type. Indeed, U-type students had the fewest personal, family, and school problems and very closely resembled students in the control group. Among all youths in the current study, this group has the highest proportion of students (55%) making a direct transition from the youth sector to the adult sector. This characteristic tends to demonstrate that many of these students were never genuine dropouts, even if they may have been considered at risk. Results from Raymond (2008) and from Bushnik, Barr-Telford, and Bussière (2002) also highlighted that younger students who reenter educational settings rapidly or make a direct transition to second-chance schools have better psychosocial profiles than do other students.

The second profile (Class 2: O/R youths) seems to suggest that some youths, especially women, display positive psychosocial adjustment, as demonstrated by high self-esteem and low psychological distress, in spite of the fact that some of them have a history of involvement with child protection services and specialized education services. Moreover, the O/R profile was characterized by the moratorium identity status, showing low commitment and high exploration, which is believed to be a normal and positive identity process for emerging adults (Berzonsky, Nurmi, Kinney, & Tammi, 1999). However, in the present study, the O/R profile quite resembles the NMP profile even though the former is composed of youths with high levels of commitment and low exploration. Because the overall actual psychosocial adjustment of both of these profiles is positive (although to various degrees), it is impossible to rule on which is the healthier identity process or status. Even if both groups carry more resemblances than differences (a three-class model included these two groups in one class), we chose to maintain these groups separately because identity exploration is a colossal task of emerging adulthood that is associated with well-being and self-discovery and with the prevention of antisocial activities and health- compromising behaviors (Schwartz et al., 2011).

The third profile (Class 3: VD) is composed of youths who reported episodes of trauma during childhood and still bear the scars of this abuse. This profile can be viewed through the lens of posttraumatic stress syndrome where past events still vividly affect the present and distort the person’s view of himself or herself (Marcotte, 2009). This particular profile comprises young women (89% of the profile) who are more likely to be parents and to have postponed returning to school longer than members of other classes. The increased risk of early pregnancy and parenthood in abused (especially sexually abused) women has been widely demonstrated in research conducted on youths exiting child protection services (Chase, Maxwell, Knight, & Aggleton, 2006; Reilly, 2003). This profile can be associated with the Fortin et al.’s (2006) D type because both profiles show levels of internalizing problems that reach the clinical threshold, good academic adjustment (as measured in this study by involvement in school work), and low levels of externalizing problem behaviors. Fortin et al. (2006) demonstrated that D types have the most conflict-ridden families; the VD youths in the current study reported abusive and chaotic familial experiences.

Finally, the last profile (Class 4: DPB) displays the greatest current difficulties, exhibiting both internalizing and externalizing problems. Also, almost half of these youths report having thought of suicide during the past year and have the lowest levels of self-esteem. Half of these youths received specialized education services during their educational journeys. They also currently display high levels of delinquency, aggressive behavior, and psychological distress, and they appear to be the least functional and adapted subgroup of all. Earlier studies have shown a continuity of problem behaviors from infancy through adulthood (Fergusson & Lynskey, 1998) and their expansion into internalizing spheres in the transition to adulthood (Marcotte, Fortin, Royer, Marcotte, & Potvin, 2008). Links between these youths and the SSAD type (Fortin et al. 2006) can easily be established. In fact, the negative past experiences described in our sample’s profile concur with the overall adjustment difficulties reported by the SSAD type, which is characterized by academic difficulties, negative teacher– student relationships, and a lack of social skills. In addition, in our study, DPB youths are the ones who envision their futures most negatively and report a low level of control over their lives. This profile represents the most vulnerable youths enrolled in second-chance schools and, consequently, merits the highest attention and consideration.

Fortin’s ACB type did not find any correspondence in our study. This particular type of dropout risk, apart from its delinquent tendencies, was characterized by an overall positive adjustment in school. It is possible that delinquency in a context of satisfactory academic performance could be interpreted as a normative and temporary feature of adolescence (Moffitt, 1993) that is less influential in emerging adulthood. On the other hand, the O/R profile did not have any corresponding profile in Fortin’s work. This is probably because this particular profile is mainly characterized by identity exploration, a developmental task under way in adolescence yet the epicenter of which is now believed to be in emerging adulthood (Schwartz et al., 2005). This aspect was not measured by Fortin et al. (2006). Nonetheless, three out of four profiles shared similarities with those of Fortin thus validating our second hypothesis.

Finally, the third hypothesis was only partially supported by the results of the study. In fact, the variables used to validate our profiles differentiated two main types of profiles: adaptive (Classes 1 and 2) and maladaptive (Classes 3 and 4). As such, the variables measuring past experiences in school showed overall better experiences in school for the NMP and O/P youths than for the others. This outcome is consistent with Fortin et al.’s (2006) study, which suggested that school variables were less accurate than psychosocial variables in discerning at-risk youths. Likewise, youths in Classes 1 and 2 believed that they have more control and power over what happens to them in comparison to youths in Classes 3 and 4. By contrast, the psychosocial profiles of Classes 1 through 4 were represented equally among students who enter AECs to reach postsecondary education in order to obtain minimal diploma or professional qualifications. Also, involvement in school work seems to be the same among profiles. These results suggest that the obstacles met by vulnerable students (Classes 3 and 4) in second-chance education settings may be primarily influenced by their perceptions of their past and present personal, school, and familial difficulties, which directly affect their optimism about their futures and their current adjustment. Knowing that, improving the psychological and social services dedicated to helping these youths, the interventions and programs aimed at increasing students’ empowerment and self-knowledge, and the teacher–student relationships that foster positive reconstructions of school experiences could be the first goal of complementary services.

To conclude, even if the young adults reentering educational settings after dropping out or the youths directly accessing AECs from the youth sector are believed to be an at-risk group of students, it is imperative to break down this group to better understand the distinctive psychosocial profiles of youths engaged in this nonlinear path in order to adapt our interventions and efforts to their respective needs. To that end, an LCA procedure on a sample of 386 youths in Quebec’s second-chance schools was conducted to achieve that goal. Results showed that four distinctive profiles, or classes, of youths attend AECs. The youths in two profiles, including the majority of youths of our sample, do not display major difficulties and are not different in substance from emerging adults enrolled in colleges or universities, except for their academic delays. However, the two other profiles, representing a quarter of the sample, reveal youths in highly vulnerable situations. Their risks emanate from past and present realities and merit our full attention. The identification of the four profiles is of particular value to the Quebec Ministry of Education, which clearly delineated the need to improve the supply of complementary educational services in order to better adapt them to the needs of the young adult population enrolled in AECs. In fact, this first assessment of the psychosocial profiles of youths in AECs may serve as a basis to improve our knowledge of this particular student population. More precisely, research mandated by the Ministry of Education and supported by corresponding financial means could include a representative sample of all AEC youths, which could lead to a clearer demonstration of profiles and needs.

This study is a first step in that direction, although it has a number of limitations that should be addressed. First, the sample is not representative of all schools in the province. Also, the length of the questionnaires limited the sampling by excluding all youths enrolled in literacy programs, which often include highly vulnerable youths. Contrary to Fortin et al.’s (2006) study, we did not use a control group; it is thus impossible to know to what extent the results obtained in AECs are different from those that would be obtained in postsecondary settings. Finally, results from LCAs should always be interpreted with caution in light of the intraclass variations. Accordingly, one youth belonging to one specific class does not necessarily present all the attributes that characterize that profile. Such analyses should be used to direct and inspire prevention, identification, and intervention but not to label youths entering second-chance schools because, for some of them, the “difficulty” tags attributed in childhood or adolescence are often still a source of devaluation and contribute to their negative self-images in adulthood.

Footnotes

Notes

Author

Julie Marcotte is a professor-researcher at the Université du Québec à Trois-Riviéres, Département de psychoéducation, 1605, chemin Sainte-Foy, entrée 8, Québec, G1S 2P1, Canada; Julie.Marcotte1@uqtr.ca . Her research focuses on the transition to adulthood for vulnerable youths, identity formation, and impact on psychosocial adjustment, as well as second-chance education.

References

Achenbach

T. M.

Dumenci

Rescorla

L. A.

(2003). DSM-oriented and empirically based approaches to constructing scales from the same item pools. Journal of Clinical Child and Adolescent Psychology, 32(3), 328–340. doi:10.1207/s15374424jccp3203_0210.1207/s15374424jccp3203_02

Achenbach

T. M.

McConaughy

S. H.

(1997). Empirically based assessment of child and adolescent psychopathology: Practical applications (2nd ed.). Thousand Oaks, CA: Sage.

Archambault

Janosz

Fallu

J.-S.

Pagani

L. S.

(2009). Student engagement and its relationship with early high school dropout. Journal of Adolescence, 32(3), 651–670. doi:10.1016/j.adolescence.2008.06.00710.1016/j.adolescence.2008.06.007

Archer

S. L.

(2008). Identity and interventions: An introduction. Indentity: An International Journal of Theory and Research, 8(2), 89–94.

Arnett

J. J.

(2000). Emerging adulthood: A theory of development from the late teens through the twenties. American Psychologist, 55(5), 469–480. doi:10.1037/0003-066X.55.5.46910.1037/0003-066X.55.5.469

Arnett

J. J.

(2007). Suffering, selfish, slackers? Myths and reality about emerging adults. Journal of Youth and Adolescence, 36(1), 23–29. doi:10.1007/s10964-006-9157-z10.1007/s10964-006-9157-z

Balistreri

Busch-Rossnagel

N. A.

Geisinger

K. F.

(1995). Development and preliminary validation of the Ego Identity Process Questionnaire. Journal of Adolescence, 18(2), 179–192. doi:10.1006/jado.1995.101210.1006/jado.1995.1012

Barnard

W. M.

(2004). Parent involvement in elementary school and educational attainment. Children and Youth Services Review, 26(1), 39–62. doi:10.1016/j.childyouth.2003.11.00210.1016/j.childyouth.2003.11.002

Battin-Pearson

Newcomb

M. D.

Abbott

R. D.

Hill

K. G.

Catalano

R. F.

Hawkins

J. D.

(2000). Predictors of early high school dropout: A test of five theories. Journal of Educational Psychology, 92(3), 568–582. doi:10.1037/0022-0663.92.3.56810.1037/0022-0663.92.3.568

10.

Bélanger

Carignan-Marcotte

Staiculescu

(2007). La diversité des trajectoires et la réussite éducative des adultes en formation de base [Diversity of trajectories and the educational success of adult basic education]. Montreal: Centre interdisciplinaire de recherche/développement sur l’éducation permanente–Université du Québec à Montréal.

11.

Bernstein

D. P.

Fink

Handelsman

Foote

(1994). Initial reliability and validity of a new retrospective measure of child abuse and neglect. American Journal of Psychiatry, 151(8), 1132–1136.

12.

Berzin

S. C.

(2010). Vulnerability in the transition to adulthood: Defining risk based on youth profiles. Children and Youth Services Review, 32(4), 487–495. doi:10.1016/j.childyouth.2009.11.00110.1016/j.childyouth.2009.11.001

13.

Berzonsky

M. D.

Kuk

L. S.

(2005). Identity style, psychosocial maturity, and academic performance. Personality and Individual Differences, 39(1), 235–247. doi:10.1016/j.paid.2005.01.01010.1016/j.paid.2005.01.010

14.

Berzonsky

M. D.

Nurmi

J.-E.

Kinney

Tammi

(1999). Identity processing style and cognitive attributional strategies: Similarities and difference across different contexts. European Journal of Personality, 13(2), 105–120. doi:10.1002/(sici)1099-098410.1002/(sici)1099-0984 (199903/04)13:2<105::aid-per325>3.0.co;2-h

15.

Bushnik

Barr-Telford

Bussière

(2002). À l’école secondaire ou non: Premiers résultats du deuxième cycle de l’Enquête auprès des jeunes en transition [In high school or not: Results from the second cycle of the Youth in Transition Survey]. Ottawa: Statistique Canada, Ressources humaines et Développement des compétences Canada.

16.

Charbonneau

(2006). Réversibilités et parcours scolaires au Québec [Reversibility and educational pathways in Quebec]. Presses Universitaires de France, 160(1), 111–131.

17.

Chase

Maxwell

Knight

Aggleton

(2006). Pregnancy and parenthood among young people in and leaving care: What are the influencing factors, and what makes a difference in providing support? Journal of Adolescence, 29(3), 437–451.

18.

Côté

J. E.

(2002). The role of identity capital in the transition to adulthood: The individualization thesis examined. Journal of Youth Studies, 5(2), 117–134. doi:10.1080/1367626022013440310.1080/13676260220134403

19.

Dickhäuser

Reinhard

M.-A.

(2006). Factors underlying expectancies of success and achievement: The influential roles of need for cognition and general or specific self-concepts. Journal of Personality and Social Psychology, 90(3), 490–500. doi:10.1037/0022-3514.90.3.49010.1037/0022-3514.90.3.490

20.

Entwisle

D. R.

Alexander

K. L.

Olson

L. S.

(2004). Temporary as compared to permanent high school dropout. Social Forces, 82(3), 1181–1205. doi:10.1353/sof.2004.003610.1353/sof.2004.0036

21.

Fergusson

D. M.

Lynskey

M. T.

(1998). Conduct problems in childhood and psychosocial outcomes in young adulthood: A prospective study. Journal of Emotional and Behavioral Disorders, 6(1), 2–18. doi:10.1177/10634266980060010110.1177/106342669800600101

22.

Fortin

Marcotte

Potvin

Royer

É.

Joly

(2006). Typology of students at risk of dropping out of school: Description by personal, family and school factors. European Journal of Psychology of Education, 21(4), 363–383. doi:10.1007/bf0317350810.1007/bf03173508

23.

Foster

E. M.

Flanagan

Osgood

D. W.

Ruth

G. R.

(2005). The transition to adulthood for vulnerable youth and families: Common themes and future directions. In Osgood

D. W.

Foster

E. M.

Flanagan

Ruth

G. R.

(Eds.), On your own without a net: The transition to adulthood for vulnerable populations (pp. 375–390). Chicago: University of Chicago Press.

24.

Guérin

Marsh

H. W.

Famose

J.-P.

(2004). Generalizability of the PSDQ and its relationship to physical fitness: The European French connection. Journal of Sport and Exercise Psychology, 26(1), 19-38.

25.

Halperin

(1998). The forgotten half revisited: American youth and young families, 1988–2008. Washington, DC: American Youth Policy Forum.

26.

Institut de la statistique du Québec. (2009, 24 September). Population selon le groupe d’âge, régions administratives, Québec, 2006 [Population by age group, administrative regions, Quebec, 2006]. Retrieved June 23, 2010, from http://www.stat.gouv.qc.ca/regions/lequebec_20/population_20/poptot20.htm

27.

Keller

T. E.

Cusick

G. R.

Courtney

M. E.

(2007). Approaching the transition to adulthood: Distinctive profiles of adolescents aging out of the child welfare system. Social Service Review, 81(3), 453–484. doi:10.1086/51953610.1086/519536

28.

Looker

D. E.

Thiessen

(2008). Le système de la seconde chance: Résultats des trois cycles de l’Enquête auprès des jeunes en transition [The second chance system: Results from the three cycles of the Youth in Transition Survey]. Gatineau, Quebec, Canada: Politique stratégique et recherche. Retrieved from http://publications.gc.ca/collections/ collection_2008/hrsdc-rhdsc/HS28-142-2008F.pdf

29.

Marcia

J. E.

(1966). Development and validation of ego-identity status. Journal of Personality and Social Psychology, 3(5), 551–558. doi:10.1037/h002328110.1037/h0023281

30.

Marcotte

(2008). Quebec’s adult educational settings: Potential turning points for emerging adults? Identifying barriers to evidence-based interventions. Canadian Journal of Educational Administration and Policy, 77. Retrieved from http://www.umanitoba.ca/publications/cjeap/pdf_files/marcotte.pdf

31.

Marcotte

(2009). Identity development and exploration and their psychosocial correlates in emerging adulthood: A portrait of youths attending adult education centers in Quebec. Vulnerable Children and Youth Studies, 4(4), 279–287. doi:10.1080/1745012090292319710.1080/17450120902923197

32.

Marcotte

Cloutier

Fortin

(2010). Portrait personnel, familial et scolaire des jeunes adultes émergents (16–24 ans) accédant aux secteurs adultes du secondaire: identification des facteurs associés à la persévérance et à l’abandon au sein de ces milieux scolaires [Personal, family, social and school characteristics of emerging adults accessing adult secondary sectors: Identification of factors associated with persistence and abandonment within these school communities]. In Action concertée Persévérance et Réussite Scolaires [Concerted actions toward school success and perseverance]. Quebec, Canada: Fonds québécois de la recherche sur la société et la culture.

33.

Marcotte

Fortin

Royer

É.

Marcotte

Potvin

(2008). La situation scolaire et sociale et l’ajustement des jeunes adultes émergents ayant présenté des problèmes de comportement extériorisés au secondaire [The school situation and social adjustment of young emerging adults with presented externalizing behavior problems at school]. Revue des sciences de l’éducation, 34(1), 141–161. Retrieved from http://id.erudit.org/iderudit/018994ar

34.

Masten

A. S.

Burt

K. B.

Roisman

G. I.

Obradovic

Long

J. D.

Tellegen

(2004). Resources and resilience in the transition to adulthood: Continuity and change. Development and Psychopathology, 16(4), 1071–1094. doi:10.1017/S095457940404014310.1017/S0954579404040143

35.

Ministère de l’Éducation du Loisir et du Sport. (2009). Les indicateurs de l’éducation, Édition 2009: Gouvernement du Québec [Education indicators, 2009 edition: Government of Quebec]. Retrieved from http://www.mels.gouv.qc.ca/sections/publications/publications/SICA/DRSI/IndicateursEducation2009.pdf

36.

Moffitt

T. E.

(1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100(4), 674–701. doi:10.1037/0033-295x.100.4.67410.1037/0033-295x.100.4.674

37.

Muthén

(2000). Integrating person-centered and variable-centered analyses: Growth mixture modeling with latent trajectory classes. Alcoholism: Clinical and Experimental Research, 24, 882–891.

38.

Paivio

S. C.

Cramer

K. M.

(2004). Factor structure and reliability of the Childhood Trauma Questionnaire in a Canadian undergraduate student sample. Child Abuse and Neglect, 28(8), 889–904.

39.

Préville

Boyer

Potvin

Perrault

Légaré

(1992). La détresse psychologique: détermination de la fiabilité et de la validité de la mesure utilisée dans l’enquête Santé Québec 1987 [Psychological distress: Determining the reliability and validity of the measure used in the 1987 Quebec Health Survey] (Research paper). Quebec, Canada: Ministère de la Santé et des Services sociaux, gouvernement du Québec.

40.

Raymond

(2008). Décrocheurs du secondaire retournant à l’école [High school dropouts returning to school]. Ottawa: Statistique Canada. Retrieved from http://www.statcan.gc.ca/pub/81-595-m/81-595-m2008055-fra.pdf

41.

Reilly

(2003). Transition from care: Status and outcomes of youth who age out of foster care. Child Welfare: Journal of Policy, Practice, and Program, 82(6), 727–746.

42.

Ross

Gray

(2005). Transitions and re-engagement through second chance education. Australian Educational Researcher, 32(3), 103–140.

43.

Sacker

Schoon

(2007). Educational resilience in later life: Resources and assets in adolescence and return to education after leaving school at age 16. Social Science Research, 36(3), 873–896. doi:10.1016/j.ssresearch.2006.06.00210.1016/j.ssresearch.2006.06.002

44.

Schwartz

Beyers

Luyckx

Soenens

Zamboanga

Forthun

. (2011). Examining the light and dark sides of emerging adults’ identity: A study of identity status differences in positive and negative psychosocial functioning. Journal of Youth and Adolescence, 40(7), 839–859. doi:10.1007/s10964-010-9606-610.1007/s10964-010-9606-6

45.

Schwartz

S. J.

Côté

J. E.

Arnett

J. J.

(2005). Identity and agency in emerging adulthood: Two developmental routes in the individualization process. Youth and Society, 37(2), 201–229. doi:10.1177/0044118X0527596510.1177/0044118X05275965

46.

Statistique Canada. (2008). Les études postsecondaires—participation et décrochage : différences entre l’université, le collège et les autres types d’établissements postsecondaires [Postsecondary education—Participation and dropping out: Differences across university, college and other postsecondary institutions] (No. 81-595-M, Catalog No. 070). Retrieved from http://www.statcan.gc.ca/pub/81-595-m/81-595-m2008070-fra.pdf

47.

Statistique Canada. (2010). Note sur l’obtention du diplôme d’études secondaires et la fréquentation scolaire, selon l’âge et la province, 2009–2010 [Note on graduation from high school and school attendance by age and province, 2009–2010]. Retrieved from http://www.statcan.gc.ca/pub/81-004-x/2010004/article/11360-fra.htm

48.

te Riele

(2000). “The best thing I’ve ever done”: Second chance education for early school leavers. Retrieved from http://www.eric.ed.gov/PDFS/ED457513.pdf

49.

Tinto

(1997). Classrooms as communities: Exploring the educational character of student persistence. Journal of Higher Education, 68(6), 599–623.

50.

Tyler

J. H.

Lofstrom

(2009). Finishing high school: Alternative pathways and dropout recovery. The Future of Children, 19(1), 77–103. doi:10.1353/foc.0.001910.1353/foc.0.0019