Abstract
The purpose of this study is to investigate the relationship between motivation in high school and postsecondary enrollment among 10th-grade students whose parents did not go to college. Specifically, this study (1) identified distinct groups of students’ self-reported reasons for attending schools among 10th graders, (2) examined whether these groups were differentially associated with indicators of college preparation and enrollment, and (3) investigated whether the time to postsecondary enrollment differed across groups. A latent class analysis was conducted to classify students into different motivation orientations. Using data from the Education Longitudinal Study of 2002, the current study found three distinct classes of school motivation, with different reasons for attending school. The largest class (53%) was characterized by high intrinsic, identified/introjected, and external motivations for attending school. Patterns of college preparation and enrollment outcomes varied across motivation orientations. Implications for school professionals and supporting programs are discussed.
Keywords
Going to college is an important life decision and has a long-term impact on individuals’ career options and advancement. Several studies reinforce widely accepted understandings that college graduates fare better in the workforce than those without a college degree (e.g., Kena et al., 2015; Oreopoulos & Petronijevic, 2013). It is also firmly established that higher lifetime earnings are associated with more years of education. Lifetime earnings impact health and mortality, retirement income security, and various aspects of social mobility. Those with a college degree are more likely to be employed and to hold higher status jobs (Kim et al., 2015).
However, not all high school graduates choose to go to college. While many factors affect a high school graduate’s decision to go to college, previous literature has found that the likelihood of high school graduates enrolling in postsecondary education is strongly correlated with parents’ education levels (Choy, 2001). Cataldi et al. (2018) found that the majority of high school students completed their high school education and were awarded either a diploma, general equivalency diploma certificate, or another high school education completion equivalency. However, the college attendance rate was 78% for students whose parents earned a postsecondary degree compared with 63% for students whose parents attended some college and 58% for those whose parents have not attended college. In addition to college attendance, previous studies have found that students whose parents have not earned a bachelor’s degree are less prepared academically and are less likely to have the expectation of earning a bachelor’s degree themselves (Cataldi et al., 2018; Chen & Carroll, 2005). These undergraduates whose parents never enrolled in postsecondary education are often referred to as “first-generation college students (FGCS)” in the literature (Cataldi et al., 2018). These characteristics are identified in the literature as potential barriers to reaching postsecondary education for many FGCS. However, it is important to acknowledge that some FGCS manage to excel and demonstrate academic resilience despite these obstacles (Morales, 2014).
Previous studies have focused on college motivation among FGCS once they are enrolled in college, but relatively fewer studies have studied this population while still in high school. Students’ school motivation in high school may play a role in the college preparation process and may affect their decisions to attend college or not after high school graduation. However, very few studies have examined this relationship, especially among high school students whose parents did not go to college. More studies are needed in this area to identify early indicators for postsecondary enrollment and implications for related support services.
Literature Review
Individual Characteristics of FGCS
Previous studies have found that FGCS are more likely to be Latinx or Black and come from low-income families as compared to students whose parents have attended college (K. V. T. Bui, 2002; Ishitani, 2003; Terenzini et al., 1996). These individual characteristics of FGCS were also observed to be related to their college success and persistence (Ishitani, 2003). Using data from the National Education Longitudinal Study (NELS: 88), Chen and Carroll (2005) found that compared to non-FGCS, many FGCS struggled academically after they enrolled in college, needed remedial courses during college years (55%), had not decided on an undergraduate major (33%), earned fewer credits in their first year, and had a lower first-year grade point average (GPA). In addition to academic adjustment, other studies found that FGCS were more likely to report social and adjustment issues (K. V. T. Bui, 2002; Pascarella et al., 2004).
Although differences in college success were observed between FGCS and non-FGCS, these differences often reflect small effects (Aspelmeier et al., 2012). Strayhorn (2006) applied a three-stage college impact model to explore how differing student background characteristics, precollege experiences, and college experiences are related to college cumulative GPA. The study found that after controlling for students’ demographic characteristics and precollege and college academic experience, no significant difference in achievement was observed between FGCS and non-FGCS. Strayhorn’s (2006) study highlights the importance of precollege and college experiences associated with being FGCS. Because the impact of the FGCS status seems to be largely confounded with low socioeconomic status (SES) and minority status, issues related to the FGCS status have often been discussed from a social/cultural capital perspective (Aspelmeier et al., 2012).
Family Characteristics of FGCS
Students whose parents did not go to college may be disadvantaged in terms of “cultural capital,” which is defined as, “the value students gain from their parents that supports and assists them as they navigate the college experience and seek a higher social status and greater social mobility” (Ward et al., 2012, p. 6). Consequently, family guidance for college preparation and expectations for students to obtain a college degree may be lacking (Pascarella et al., 2004). For example, taking rigorous courses that help students prepare for attending college is crucial for students’ academic readiness. DeFreitas and Rinn (2013) found that early opportunities in advanced math and science classes are associated with higher math self-concept scores (as well as verbal self-concept scores) and higher academic achievement (DeFreitas & Rinn, 2013). Often, encouragement to enroll in these higher level courses comes from parents, and parents without college experience may view the benefit of such courses differently than parents with college experience (Horn & Bobbitt, 2000).
Previous studies have also suggested that parents’ knowledge of the college campus environment, access to human and financial resources, and familiarity with the terminology and general functioning of a college setting are also part of the social and cultural capital and are key to college enrollment and achievement (McCarron & Inkelas, 2006; Perna & Titus, 2005). Students whose parents have not attended college are less likely to have college information and college-related parental involvement (K. Bui & Rush, 2016; Thayer, 2000). These students are less likely to aspire to attend college and may be less academically prepared; they are less likely to complete the required college entrance exams, the application process, or the arrangements for college enrollment as compared to students whose parents have earned a bachelor’s degree (Choy, 2001; Wilbur & Roscigno, 2016). Wilbur and Roscigno (2016) also found that such disadvantages due to the social–cultural capital of the FGCS were not eliminated even as students’ SES increased. These findings pointed out the unique challenges faced by FGCS in the college-going process.
Precollege School Experience of FGCS
FGCS often receive limited guidance for college preparation from their high schools as well as from their family. Literature has consistently suggested that FGCS are less academically prepared when they enter college (Thayer, 2000). One of the reasons cited in the literature is that FGCS are often placed in vocational, technical, and/or remedial programs, which were not designed for preparing students for college (Striplin, 1999). Redford and Hoyer (2017) found that, in 10th grade, a lower proportion of FGCS expected to obtain a bachelor’s degree or advanced degree compared to their peers. A large proportion of FGCS had not thought about taking the American College Testing (ACT)/Scholastic Assessment Test (SAT) in 10th grade and had a lower GPA in high school (Redford & Hoyer, 2017). Engle and Tinto (2008) suggested strategies that schools can use to improve academic preparation for low-income FGCS, such as providing more information and counseling on course-taking, support to complete challenging coursework, access to rigorous college preparatory courses, and facilitating a college-going culture in school.
School Motivation and Academic Success of FGCS
The aforementioned characteristics are identified in the literature as potential barriers to reaching postsecondary education for many FGCS. However, some FGCS achieve academic success despite these obstacles. Academic resilience has recently become a more prominent topic in the literature (Morales, 2014), yet factors promoting academic resilience among FGCS are less discussed. Understanding how academic resilience impacts the overall academic success of students who are low SES and minority status (Morales, 2014) is important as these are prominent characteristics of the FGCS population.
Motivation has been shown to be a predictive factor of academic success at various levels (Edgar et al., 2019) and is also found to be a factor promoting academic resilience when in a disadvantaged environment (Alfaro et al., 2009; Stuber, 2011). Evidence in the literature shows that intrinsic motivation has a positive impact on the performance and success of FGCS. For example, Trevino and DeFreitas (2014) reviewed previous studies on Latino FGCS showing that academic resilience and persistence are strongly related to intrinsic motivation and academic achievement.
While having intrinsic motivation is associated with FGCS’ academic success and completion (Próspero & Vohra-Gupta, 2007), FGCS may be going to college for different reasons. K. V. T. Bui (2002) found that FGCS are more likely to report wanting to go to college to help their family financially, gain respect/status, and bring honor to their family. Irlbeck and her colleagues (2014) conducted a case study to better understand FGCS’ motivations and support systems. Semistructured interviews were conducted to determine what factors led to college enrollment. Family and peer support systems, self-motivation, and integration into school groups/activities were themes across participants (Irlbeck et al., 2014). One participant shared how her motivation stemmed from witnessing the challenges her mother faced after her father died at a young age. Other participants mentioned teachers, family members, and family situations as motivating factors for pursuing a degree. These findings suggest that extrinsic motivations is also a factor for aspiring FGCS to attend college (Próspero & Vohra-Gupta, 2007; Sidle & McReynolds, 2009).
Focus and Approach of the Present Study
The purpose of this study is to investigate the relationship between motivation to attend high school and college preparation and enrollment among 10th-grade students whose parents did not go to college. Building on previous work by Abar et al. (2012), which examined the influence of school motivation among 10th-grade students using data from Education Longitudinal Study of 2002 (ELS 2002), this study is filling an important gap in the literature by exploring the precollege motivation among students whose parents did not go to college, as these students become FGCS when they successfully reach postsecondary education. This study adapted the self-determination theory (SDT) and organismic integration theory (OIT) framework (Abar et al., 2012; Ryan & Deci, 2017) to investigate high school students’ reasons for attending school and how they are related to late–high school dropout. Based on the SDT and OIT theoretical framework, student motivation can be categorized as intrinsic motivation, extrinsic motivation, and amotivation. Particularly, extrinsic motivation can be further categorized into externally regulated behaviors, introjected regulation of behaviors, identification, and integrated regulation based on varying levels of autonomy in the behaviors (Ryan & Deci, 2017).
Methodologically, we used a person-centered approach, latent class analysis (LCA; Collins & Lanza, 2009) to classify individuals into distinct motivation orientation groups. The person-centered approach is an alternative way to unpack complex relations among variables by identifying groups of people based on their similarities on a set of indicators (Muthén & Muthén, 2000). Because the person-centered approach takes into account the heterogeneity of the population, it allows researchers to explore the relationship between variables more comprehensively. Students’ motivation orientation was identified through LCA, which examines the students’ item response patterns among the binary indicators of reasons for attending school. Students with a more similar motivation orientation were classified into the same subgroup.
Method
Participants
Data analyzed in this study were obtained from the ELS 2002 and from the first follow-up study in 2004 (Ingels et al., 2007). The sampling technique utilized by the ELS consisted of two stages. In the first stage, high schools were selected using the probability proportional to size technique. In the second stage, students from the selected schools were randomly selected, with each school yielding approximately 26 students. Observations from 3,997 10th-grade students, enrolled in 353 high schools, whose parents did not go to college were included at baseline. Of the 10th-grade students considered, 2,059 (51.5%) identified as female. For race/ethnicity, this study categorized student responses into five broad groups: Non-Latinx White (47.0%), Latinx (23.9%), Black or African American (14.2%), Asian (10.2%), and Other (4.7%). Additionally, 1,021 (25.8%) students stated that English was not their native language.
Measures
Reasons for attending school
In the ELS 2002 Student Questionnaire, students were asked to indicate their reasons for attending school on a list of nine questions, on a 4-point Likert-type scale (strongly agree to strongly disagree). Among the nine questions, seven questions were selected to be used in this study as they are closely related to the SDT and OIT framework (Ryan & Deci, 2017): two survey questions capturing intrinsic motivations, two capturing identified/introjected regulation, two capturing external regulation, and one capturing amotivation (Abar et al., 2012). For the purpose of this study, these items were dichotomized with 1 = agree and 0 = disagree.
College preparation
Five somewhat sequential indicators from the ELS 2002 Student Questionnaire were used to reflect students’ college preparation during high school (Choy, 2001). These five indicators are students’ college degree aspiration in 10th grade, academic preparedness for college-level work (math and reading proficiency), taking college entrance examinations (SAT and/or ACT), and filling out applications. All of these indicators were coded as binary variables, with 1 = yes and 0 = no.
Postsecondary enrollment
Postsecondary enrollment was a dichotomous variable indicating whether a student had enrolled in a postsecondary educational program within 2 years from their high school graduation date. In this sample, 43% (n = 1,435) of the respondents did not enroll in postsecondary education 2 years after graduation. Additionally, it was also of interest to determine whether there were differences between student motivation profiles in how many months had elapsed prior to enrolling (time to enrollment) in a postsecondary program. Of note, the 2-year window was chosen out of practicality as only 25 months of postsecondary enrollment data were available, starting from January of the students’ 12th-grade academic year and ending in January 2 years later.
Control variables
To better understand the relationship between school motivation and college preparation and enrollment, we control for students’ demographics, cultural capital, and parental involvement. As discussed previously, a student’s race/ethnicity, SES, cultural capital, and parental involvement have been found to affect a student’s college preparation and enrollment. In addition, covariates included native language (English or non-English) and student gender, as prior literature has observed a gender gap in college enrollment (Doherty et al., 2016) and lower college enrollment for English language learners (Kanno & Cromley, 2016).
Demographic Variables
Students’ sex was coded as a binary variable (0 = female, 1 = male). Students’ race/ethnicity was coded into four binary variables indicating Asian American, Black American, Latinx, and Other. White was the reference group. Students’ native language was coded as a binary variable (0 = nonnative English speaker, 1 = native English speaker). Students’ SES was a composite index score made up of five separate variables, which are parents’ educational levels, occupations, and overall family income (Ingels et al., 2007). SES was used as a continuous variable in the analysis.
Cultural capital and parental involvement
In the ELS 2002 data, there is not a set of items deliberately designed to measure cultural capital. In this study, we used one item as a proxy of the potential social and cultural resources students may have. This item is about outside school activities (i.e., how often do you spend time on taking music, art, language, and dance classes outside of school?). Previous research (Wilbur & Roscigno, 2016) has operationalized cultural capital as the frequency of students’ participation in music, art, foreign language classes, and dance lessons. This item is rated on a 4-point scale with 1 = rarely or never participating and 4 = participating every day or almost every day. In terms of parental involvement, we used four items asking how often students discussed things at school with their parents (i.e., selecting courses or programs at school, school activities or events of particular interest, things studied in class, and grades; Wilbur & Roscigno, 2016). We computed an index score for parental involvement for each student by taking the mean of the four items (α = .81) and used the index score as a continuous variable in the analysis.
Analysis
All analyses were conducted with Mplus Version 7.3. The robust maximum likelihood estimator was used as it is the default estimator in Mplus. Because the ELS 2002 data were collected using a multistage sampling approach, data were clustered in the sampling unit. To account for the nonindependence of observation due to sampling, a sandwich estimator was used to compute the standard errors. In addition, because ELS 2002 data were collected longitudinally, weights are needed to adjust for the survey nonresponse across different time points. In this study, we use the base year to the second follow-up panel weight, as we linked students’ data from the base-year survey to their postsecondary enrollment information during the second follow-up data collection.
LCA
In this study, we first performed an unconditional LCA on the seven binary indicators of students’ reasons to attend school. We examined the model fit indices to determine an optimal number of classes. Student-level characteristics, which included gender, race/ethnicity, an indicator of whether English as their native language, and variables about their SES, cultural capital, and parental educational involvement, were reported by latent classes.
A series of LCA models with a different number of class solutions were analyzed and compared to determine the best fitting model. Several model fit indices were used to inform model selection, including Akaike’s information criterion (AIC; Akaike, 1987), the Bayesian information criterion (BIC; Schwarz, 1978), and the sample size–adjusted BIC (aBIC; Sclove, 1987), adjusted Lo–Mendell–Rubin test (aLRT; Lo et al., 2001), and entropy. The information criterion indices (i.e., AIC, BIC, and aBIC) are useful for model comparisons as they take into account the goodness of fit and parsimony of a model. As suggested in the literature, smaller AIC, BIC, or aBIC values indicate a better model fit (Denson & Ing, 2014). The aLRT is similar to the likelihood ratio difference test that compares model fit between K class and K − 1 class solutions. A significant p value indicates that the model with K class had a better model fit than a more parsimonious model with K − 1 class (Lo et al., 2001). Finally, the entropy value was used to examine the quality of the LCA classification and the extent to which individuals have been correctly assigned to a latent class. Entropy values are bounded between 0 and 1. Values approaching 1 indicate good classification accuracy (Geiser, 2013). In addition to the statistical indicators, we take into account the size, interpretability, and meaningfulness of the latent class as suggested in the literature (e.g., Denson & Ing, 2014; Lanza et al., 2013).
LCA with college preparation indicators as a distal outcome
To examine whether students’ school motivation orientation is a predictor of students’ college preparation, we added indicators of college preparation as a distal outcome to the conditional LCA model. These binary distal outcome variables include students’ degree aspirations in 10th grade, academic skills (math and reading) in 10th grade, taking the SAT/ACT in 12th grade, and ever applying for a college. A three-step approach (Asparouhov & Muthén, 2014) was used to evaluate the relationship between the latent class variable and college preparation and enrollment outcomes while controlling for the covariates.
LCA with college enrolment as a distal outcome
To examine whether students’ school motivation orientation was a predictor of students’ college postsecondary enrollment, two approaches were taken to explore college enrollment as a dichotomous variable and as time to enrollment by month. The binary enrollment variable was handled as a binary distal outcome variable using the procedure previously mentioned. For the time-to-enrollment variable, a latent discrete-time survival analysis using the proportional hazard model (Muthén & Masyn, 2005; Raykov et al., 2017) was conducted to estimate the conditional probability of students enrolling in postsecondary education in this given period. In Mplus, the latent discrete-time survival model was parameterized as a one-parameter logistic item response theory model (Raykov et al., 2017) with the hazard probability and covariate effect being time invariant. Specifically, students’ postsecondary status was a set of binary variables (1 = enrolled, 0 = not enrolled) observed each month during the 2 years.
The discrete-time survival analysis was selected to account for censoring. In this study, individuals who did not enroll in postsecondary education within the 2-year span and survey nonrespondents are right-censored (Muthén & Masyn, 2005). Three coding patterns were used to reflect individuals who reached postsecondary education during the period, individuals who were lost to follow-up during the time period, and individuals who did not reach postsecondary education during the time period (Muthén & Masyn, 2005). The hazard function in this latent discrete-time model represents the probability of enrolling in postsecondary education, for the first time, during the 2-year time period. Then, latent class membership was included as a predictor of postsecondary enrollment in the latent hazard function. The covariates (i.e., gender, race, ELS, math and reading achievement, and highest expected level of academic achievement) were included in the model to control for individual differences.
Results
Preliminary Analysis
We first examined students’ reasons for attending school descriptively. On average, students were likely to endorse the following statements “Education is important to get a job later” (95% agree) and “I’m learning skills that I will need for a job” (85% agree) as the reasons for attending schools. We also graphically inspected the pattern of missing data in the study variables using the naniar package in R version 0.4.0. No systematic patterns of missing data were observed in these variables.
LCA
We identified the three-class solution as the best fitting model and used it for the subsequent analyses. The three-class solution had the highest entropy among the competing models. Although the AIC and BIC values continued to decrease beyond three classes, the differences became negligible and the contextual interpretability more difficult (Table 1).
Latent Class Analysis Fit Statistics.
Note. −2LL = −2 × log likelihood; AIC = Akaike information criterion; BIC = Bayesian information criterion; aBIC = adjusted Bayesian information criterion; aLRT p = adjusted Lo–Mendell–Rubin likelihood ratio test p value.
The fit statistics for the final model were bolded.
The first class consisted of individuals exhibiting high levels of school motivation and a low level of amotivation. This class is labeled as a highly motivated (HM) group (53%). These students reported that they were motivated to attend school for a combination of intrinsic and identified/introjected reasons as well as for externally regulated reasons such as social activities. Students in this group found classes interesting and challenging, were satisfied by the learning in school, and felt education was important to their future job. In addition, students in the HM group considered school to be a place to meet friends and participate in team/club activities.
Similar to the first class, the second class consisted of individuals exhibiting a high level of school motivation with the exception of a low level of external regulation and amotivation. This class is labeled as intrinsic-identified/introjected (II) group (15%). Students in this class were characterized by attending schools for a combination of intrinsic (i.e., classes are interesting and challenging) and identified/introjected reasons (i.e., education is important for getting a good job later on). However, students in this latent class did not regard school as a place to meet friends.
The third class consisted of individuals with low levels of intrinsic motivations who valued education for its utility. This class is labeled as low intrinsic motivation (LIM) group (32%). Students in this class reported attending school in order to obtain a job in the future and felt school was a place to meet friends. Compared to the other groups, students in this group also showed a relatively higher probability of amotivation or feeling there was nothing better to do than school (Figure 1).
Probabilities of endorsing reasons for attending high school. Classes are interesting and challenging (intrinsic motivations).
Demographic Characteristics of the Classes
To further understand the characteristics of the three school motivation groups, we examined the differences in the demographic variables among the groups (Table 2). Students in the HM group had the highest proportion of Asian American students (13%) and students whose native language was not English (30%). Students in the II group had the highest proportion of female students (59%) and the highest proportion of Black/African American students (31%). Students in the LIM group had the highest proportion of male students (53%), the highest proportion of White students (61%), and the highest proportion of native English speakers.
Descriptive Statistics of the Study Variables by Latent Class.
Note. N = 3,769 (listwise deletion is applied to the covariates in the hypothesis testing). [Means] and (standard deviations) of the scale score are reported. SES is a standardized scale. Cultural capital is measured by how often students take art, music, foreign language, and dance lessons, on a 4-point rating scale (1 = rarely or never and 4 = everyday or almost everyday). Parental educational involvement is measured by how often students discussed school courses, activities, things learned in school, and grades, with parents on a 3-point scale (1 = never, 2 = sometimes, 3 = often). SES = socioeconomic scale; ns = not significant.
a Reference group for the categorical variables.
b Other includes American Indian/Alaska Native (n = 42, 1.06%) and more than one race (n = 14, 3.62%).
c Continuous variable.
*p < .05, **p < .01, ***p < .001.
Linking Class Membership to College Preparation
The results from the analyses suggest that respondents from the three motivation orientation classes had significantly different levels of college preparation, controlling for their demographic characteristics, cultural capital, and parental involvement. Figure 2 summarizes the conditional probability of enrolling in college for each latent class, given responses to a specific item. Overall, college preparation indicators were similar between the HM and II groups, except for math and reading proficiency. A much lower percentage of students in the II group were proficient in math (21%) and reading (26%) in 10th grade. Compared to those in the LIM group, students in the HM and II group had significantly higher odds of having college degree aspirations in 10th grade, taking or planning to take the SAT/ACT in 12th grade, and applying for college.
College preparation and enrollment by latent class membership. Note. *Indicates data were collected at the second follow-up survey in 2006.
Linking Class Membership to Postsecondary Enrolment
Controlling for students’ demographic characteristics, cultural capital, and parental involvement, our results suggest that students’ odds of enrolling in postsecondary education during the 2-year period after January of their 12th-grade year varied significantly across latent classes. Students in the HM group had the highest odds of enrolling in postsecondary education (odds = 1.99), followed by the II (odds = 1.82) and LIM (odds = 0.86) groups.
Among the different classes, students in the HM group were observed to have a higher likelihood of reaching postsecondary education than students in the LIM group (Wald χ2 = 14.14, p < .001), whereas no significant differences were observed in the other class comparisons. This suggested that students in the HM group were observed to reach postsecondary education more quickly than students in the LIM group after controlling for students’ demographic characteristics, cultural capital, and parental involvement.
Discussion
In this study, we used a person-centered approach to identify reasons for attending school for high school students whose parents did not attend college. We found three distinct groups of motivation orientations among these 10th-grade students, suggesting that the reasons for attending school are heterogeneous among them. Consistent with findings reported by Abar et al. (2012), which showed that the majority of 10th graders (60%) enjoy classes, are satisfied with their learning, and value education, over half (53%) of the sample of 10th graders in our study were categorized in the HM group.
The smallest subgroup in our study was the II group (15%). Students in this subgroup reported that they were motivated to go to school because they enjoyed learning and valued education yet do not consider school to be a place to meet friends or to join a team or club. As we observed, a higher proportion of females and students of color were included in the II subgroup. It is possible that these students did not experience connectedness in school. Previous research suggests that students’ gender and race play a role in their experiences and psychological adjustment in high school. For example, Akos and Galassi (2004) found that boys felt more connected than girls in high school and are more likely to report that friends in school helped to adjust to high school life. Gruber and Fineran (2016) also observed that more peer-related problems, including bullying and sexual harassment, were experienced by girls than boys during high school. Similarly, students of color are at higher risk of experiencing bullying and peer victimization during adolescence (Nansel et al., 2001). These previous findings provide some insight into the experiences of students in the II group.
Despite these differences in motivation to attend school, students in the HM and II groups reported similar college degree aspirations in 10th grade, SAT/ACT test-taking, and college planning in 12th grade. Further, a similar percentage of students in these two groups had ever applied for college and enrolled in a postsecondary institution within 2 years after 12th grade. This finding is consistent with previous research showing that students’ intrinsic motivation is positively associated with their academic achievement (Khalaila, 2015; Kumar & Sangeetha, 2019).
Math and reading skills in 10th grade differed between the two groups. Particularly, students who reported less social support from peers (II group) reported much lower scores in reading and math, which aligns with previous studies about peer influences on academic achievement (Berthelon et al., 2019). Dijkstra and Gest (2015) also showed an association between students’ academic achievement, prosocial behavior, and bullying with peer acceptance and how students feel about their school. Peer dynamics and effects related to gender and race/ethnicity discussed previously may be manifested by the low academic skills we observed in these intrinsically motivated students (II group).
The second largest subgroup, representing about one third of the students, was the LIM group. Students in this subgroup go to school to prepare themselves for getting a job and to meet friends. They are more likely to be male and White compared to the other two groups. In terms of college preparation, students with LIM are less likely to aspire to obtain a college degree, to take or plan to take the SAT/ACT in 12th grade, and to apply for college compared to students in the other two groups. We observed that less than half of the students with LIM enroll in postsecondary education, within 2 years after their 12th grade. These findings suggest that, in 10th grade, this group of students does not have a strong desire to go to college. We also observed that of the three groups, students in the LIM subgroup reported the lowest levels of cultural capital and parental education involvement at home. This is consistent with previous literature suggesting that going to college is a process deeply intertwined with students’ habitus, family cultural capital, and school contexts (Roksa & Robinson, 2017).
Our results also provide information about differences in time to enrollment among the three types of motivation orientations. We observed a significant difference in the baseline hazard function between the HM and LIM groups, after controlling for students’ demographic characteristics, cultural capital, and parental involvement. A total of 85% of the students were identified in these two groups. This finding suggested students’ school motivation in 10th grade may be an early predictor of the time students took to reach postsecondary education. This finding is consistent with previous research that both intrinsic and extrinsic motivation are factors for inspiring FGCS to attend college (Sidle & McReynolds, 2009).
Practical Implications
The results from this study depict a demographic picture of high school students with diverse motivation toward school as well as how these diverse motivations are associated with academic performance (i.e., math and reading skills), college degree aspirations, college preparation, and attendance in postsecondary education. Such information allows high school teachers and counselors to identify particular groups of students who might be vulnerable to lower academic performance and reduced rates of attendance in postsecondary education. For example, high school teachers and counselors may recognize students who belong to II group could encounter more difficulties in peer relationships, which could affect their academic performance, such as math and reading skills. The higher percentage of women and students of color in this group also provides information for high school teachers and counselors to identify the targeted population.
In addition, the struggles of students who belong to the LIM group seem to lie in the LIM in school, which are associated with a lack of aspiration and lower enrollments in postsecondary education. Yet, students in this group demonstrated high extrinsic motivation for going to school, such as “education is important to get a job later” and “school is a place to meet friends.” From the SDT and OIT framework, extrinsically motivated behaviors can be internalized when students are exposed to a context of relatedness and feeling component in the learning activities (Ryan & Deci, 2017). Therefore, when working with students in the LIM group, school counselors and teachers could help students identify meaningful reasons for going to college. Along with a good relationship and support for their autonomy, students would be more likely to internalize and integrate their motivation for attending college.
Furthermore, appropriate interventions could be developed by high school teachers and counselors to enhance peer relationships in school. For example, peer support groups for girls and students of color could be provided to increase a sense of belongingness among these students. High school teachers and counselors could provide educational programs that (a) promote values of postsecondary education through demonstrating diverse career trajectories and required levels of education and (b) introduce college environment, resources, and culture.
Limitations and Future Research
This study uses data from the ELS 2002 and the first follow-up study in 2004. While the data provide a large-scale, nationally representative sample and an excellent opportunity to explore the link between school motivation and postsecondary outcomes, there might be some changes since the data were collected. The data are limited in providing an up-to-date picture of recent 10th graders whose parents did not go to college. Future studies may consider replicating the current research with more recent large-scale data sets.
Another limitation of this study is that it focuses only on the association between students’ school motivation in high school and postsecondary enrollment; it is unknown whether their precollege motivation would also predict college persistence. More studies are needed to explore the association of precollege motivation and academic outcomes among FGCS, particularly whether students’ precollege motivation would be an early indicator of college persistence and retention, which would provide implications for career counseling professionals at the college level.
In addition, this study provides an overall picture of how high school motivation can serve as an early predictor of college enrollment among students whose parents did not go to college. Reflecting known trends, we noticed that students’ demographic backgrounds (i.e., gender, ethnic, or SES) play a distinct role in these students’ college preparation and enrollment (Cataldi et al., 2018). Future studies should explore differences in students’ college preparation and enrollment outcomes by demographic groups to improve understanding of how gender, race, and native language impact motivation and college-related outcomes.
Finally, this study did not explore the types of institutions and the academic majors of students in this sample once they enrolled in college. Little is known about high school motivation and choices major (e.g., science, technology, engineering, and mathematics majors). Future studies should also consider that motivation in high school may be associated with the choice of college major.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
