Identifying the Determinants of Chronic Absenteeism: A Bioecological Systems Approach


by Michael A. Gottfried & Kevin A. Gee - 2017

Background/Context: Chronic school absenteeism is a pervasive problem across the US; in early education, it is most rampant in kindergarten and its consequences are particularly detrimental, often leading to poorer academic, behavioral and developmental outcomes later in life. Though prior empirical research has identified a broad range of determinants of chronic absenteeism, there lacks a single, unified theoretically driven investigation examining how such factors concurrently explain the incidence of chronic absenteeism among our nation’s youngest schoolchildren. Thus, it is difficult to determine the relative importance of one factor over another, hence making it challenging to develop appropriate supports and services to reduce school absences.

Purpose/Research Questions: Our study filled this critical void—we investigated multiple determinants of chronic absenteeism that were grounded, theoretically and empirically, in Bronfenbrenner’s bioecological model of development. Specifically, using data from the Early Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K:2011) and the method of hierarchical generalized linear modeling (HGLM), we analyzed how the co-occurrence of key (1) process, (2) person, and (3) context (micro-, meso-, exo- and macrosystem) factors was associated with kindergarteners’ probability of being chronically absent.

Findings/Results: Children who have poorer health, higher internalizing behaviors, and more frequent engagement in learning activities at home had higher odds of chronic absenteeism. Also, children from larger families and of lower socioeconomic status faced increased odds of chronic absenteeism. Conversely, children holding positive attitudes towards school had lowered odds of chronic absenteeism, a finding that remained robust across socioeconomic status groups. Finally, parent–school connections were associated with lowered odds of absenteeism.

Conclusions/Recommendations: Overall, our findings strongly suggested that addressing chronic absenteeism will require comprehensive and multifaceted approaches that recognize these multiple factors. With this theoretically grounded, more descriptive approach, it is more feasible to identify key factors and subsequently design policies and practices to prevent absence behavior.



Across the US, approximately 10 to 15% of students are chronically absent, which is generally defined as missing 10% or more (about 19 days) of the school year (Balfanz & Byrnes, 2012). The negative consequences of chronic absenteeism are pervasive and particularly detrimental for children’s development in early grades, ranging from lowered academic performance (Gottfried, 2010, 2011b, in press; Connolly & Olson, 2012; Gershenson, Jacknowitz, & Brannegan, 2014; Lehr, Sinclair, & Christenson, 2004; Rumberger, 1995) to increased risk for behavioral and developmental outcomes (Gottfried, in press; Ekstrom, Goertz, Pollack, & Rock, 1986; Finn, 1989; Johnson, 2005; Newmann, 1981). In fact, of all elementary school years, chronic absenteeism is highest in kindergarten (Balfanz & Byrnes, 2012).


Prior research has examined factors that correlate with chronic absenteeism, such as educational disengagement (Bealing, 1990; deJung & Duckworth, 1986; Harte, 1994; Lehr et al., 2004; Reid, 1983; Southworth, 1992), family structure (Catsambis & Beveridge, 2001; Fan & Chen, 2001; Jeynes, 2003; McNeal, 1999; Muller, 1993; Sampson & Laub, 1994), peer effects (Author, 2013), and student–teacher interactions (Allen, 2003; Bealing, 1990; Marvul, 2012). However, due to a critical lapse in prior research, there is no overarching consensus on which factors have the greatest association with school absences. One key reason for this gap in the field is that the factors of absenteeism have been analyzed in isolation from one another and, additionally, the research surrounding chronic absenteeism has been largely atheoretical. Therefore, the disjointed nature of the work in this area renders it difficult to design more holistic practices and policies surrounding absenteeism reduction that are theoretically grounded.


Our study fills this critical void—we investigated the multiple influences of chronic absenteeism, hence overcoming the issue of disjointedness of the field. More so, we did so by drawing upon core theoretical concepts underlying Bronfenbrenner’s bioecological model of development (Brofenbrenner & Morris, 2006), hence enhancing the theoretical basis of the field. We know of no other research that systematically examined how underlying ecological determinants of chronic absenteeism, ranging from individual demographic characteristics to important child–environment interactions, concurrently influence the incidence of chronic absenteeism.


Accordingly, we adopted a novel approach in our proposed study—to our knowledge, we were the first to examine chronic absenteeism in kindergarten through the theoretical lens of Bronfenbrenner’s bioecological model of development. Through this model, we investigated kindergarteners’ probability of being chronically absent within a developmental framework that encapsulated numerous factors: (1) processes capturing interactions between children and their environments; (2) person characteristics of the child; and (3) contexts which involve the proximal and distal environments of the child. By incorporating these aspects of the model into our analysis, it was possible to simultaneously address multiple factors of chronic absenteeism, from both theoretical and empirical vantage points, and thus begin to understand a much fuller, richer portrait of what drives this negative schooling behavior.


To conduct our analysis, we analyzed data from the Early Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K:2011), the most current, nationally representative sample of kindergarteners in the US. We used hierarchical generalized linear modeling (HGLM), which enabled us to examine select bioecological factors related to chronic absenteeism while simultaneously accounting for the clustering of children within schools. The new knowledge gained from our study is critical for influencing ways in which both policymakers and practitioners can target multiple factors that make children most susceptible to excessive absenteeism. Further, our work puts forth evidence grounded directly in developmental theory that can be leveraged to help guide the design of interventions early on in schooling, before the ramifications of chronic absenteeism become exacerbated.


BACKGROUND AND CONTEXT


Though chronic absenteeism is a phenomenon that occurs throughout schooling, it is most pervasive in kindergarten (Balfanz & Byrnes, 2012; Romero & Lee, 2007). As kindergarten is an extremely critical developmental period which establishes children’s foundational skills, attitudes and dispositions for lifelong success (Duncan et al., 2007; Olson, Sameroff, Kerr, Lopez, & Wellman, 2005; Posner & Rothbart, 2000), educators and early childhood specialists have expressed concern about children missing excessive amounts of this formative year of schooling (Author, 2015). In fact, this concern is borne out by evidence demonstrating the overwhelmingly negative consequences of chronic absenteeism in children’s early years of school. Prior empirical evidence demonstrates that children with higher school absences early on perform more poorly in school (Dryfoos, 1990; Finn, 1993; Gottfried, 2009, 2011a; Lehr, Hansen, Sinclair, & Christenson, 2003; Stouthamer-Loeber & Loeber, 1988); have a larger probability of high school dropout (Rumberger, 1995); face an increased chance of future unemployment (Alexander, Entwisle, & Horsey, 1997; Broadhurst, Paton, & May-Chahal, 2005; Kane, 2006); are eventually more likely to use tobacco, alcohol, and other drugs (Hallfors et al., 2002); and exhibit greater behavioral issues, including social disengagement and alienation (Gottfried, in press; Ekstrom et al., 1986; Finn, 1993; Johnson, 2005; Newmann, 1981). Moreover, these ramifications are exacerbated for children who are from families of lower socioeconomic status (SES) (Balfanz & Legters, 2004; Fine, 1994; Orfield & Kornhaber, 2001).


Unsurprisingly, given this string of negative outcomes attributed to absenteeism, many researchers have attempted to identify factors that are associated with this detrimental behavior. Doing so provides insight into how these negative academic and developmental consequences can be mitigated. A growing body of prior empirical evidence demonstrates that the correlates of chronic absenteeism are multifaceted, ranging from individual characteristics of children themselves to their immediate environments.


First, at the individual level, educational disengagement and alienation from school (Bealing, 1990; deJung & Duckworth, 1986; Harte, 1994; Lehr et al., 2004; Reid, 1983) and poor health (Allen, 2003; Pourat & Nicholson, 2009) are important predictors of chronic absenteeism. Second, Reid (1982) shows that family influences, such as family structure, father’s occupation, mother’s work status, and free lunch status, are all related to absence patterns. Specifically, household size as a measure of family structure is speculated to be linked to different rates of absenteeism, due to availability in parental supervision (Sampson & Laub, 1994). Relatedly, low parental involvement is also linked to higher student absences (Catsambis & Beveridge, 2001; Fan & Chen, 2001; Jeynes, 2003; McNeal, 1999; Muller, 1993). Ready (2010) and Romero and Lee (2007) both find evidence that family SES (as measured by occupation, mother’s work status, and free lunch status) is negatively related to absences. Moreover, children of younger mothers tend to have higher rates of absences. Third, in addition to the family, other sociocontextual factors such as peers’ academic, demographic, and behavioral characteristics are associated with absenteeism (Gottfried, 2011, 2013; Rothman, 2001). Additionally, school-level factors such as teacher–pupil relations, availability of health personnel, and program interventions are also highlighted as critical (Allen, 2003; Bealing, 1990; Marvul, 2012).


Despite mounting evidence in the research, there is no general consensus on which factors are most influential in predicting chronic absenteeism. This is partially because many of the aforementioned factors have been examined in isolation from one another. That is, there lacks a single, unified theoretically driven research agenda that jointly examines multiple factors of chronic absenteeism. There are two critical issues with a lack of an atheoretical, disjointed research agenda exploring factors of chronic absenteeism: first, no one has extracted the influence of each factor while controlling for the joint influence of other potential factors; and second, there lacks a foundational explanation as to why we would expect to see the patterns that we do. Therefore, it is difficult to design policy and practice surrounding the prevention of chronic absenteeism without further insight in these two issues.


We address both of these issues. As for the first issue, when viewed holistically, the empirical evidence on the determinants of chronic absenteeism reviewed above suggests not only that there are myriad factors linked to chronic absenteeism, but that these factors can be conceived of in a broader embedded system involving (1) the individual child; (2) his/her surrounding contexts, both proximal (i.e., family and school) and distal (i.e., SES); and (3) the interactions that children have within and across those contexts. Moreover, the extant research evidence suggests that these numerous factors should be studied in conjunction with one another rather than in isolation. Related (and linked to the second issue raised above), we argue that the empirical evidence, when collectively viewed, has implicitly suggested that chronic absenteeism can be conceptualized within a theoretical framework that shares salient features of Bronfenbrenner’s bioecological model of development (Brofenbrenner & Morris, 2006). This model, and prior variants of the model (sometimes referred to as the ecological model of human development [Bronfenbrenner, 1993, p. 37]) recognizes that children’s development over time is shaped by characteristics of children themselves, the environments in which they are embedded at multiple levels, and the processes that they engage in with these multiple environments (Bronfenbrenner, 1993; Bronfenbrenner & Morris, 2006). Accordingly, we ground our own study by drawing upon key elements of the bioecological model; however, rather than implicitly relying on the model, we explicitly and systematically allow the model to guide and inform our thinking about factors influencing chronic absenteeism. This is a substantial departure from prior absenteeism research.


For the purposes of our study, we focus on three components of the bioecological model: process, person, and context (Bronfenbrenner & Morris, 2006, p. 795). Drawing extensively from the work of Bronfenbrenner and Morris (2006), we define and explain how we conceived of each component of the model. Later, in our Method section, we describe the measures we used to operationalize these components.


PROCESS


Processes capture regularly occurring interactions between children, people, and their environments. More specifically, processes occur as the result of “progressively more complex reciprocal interaction between an active, evolving biopsychological human organism and the persons, objects, and symbols in its immediate external environment” (Bronfenbrenner & Morris, 2006, p. 797). Further, as Bronfenbrenner and Morris (2006) argue, for processes to actually be influential they must occur regularly over an extended period of time. They note that patterns of processes can be found in activities such as “reading, learning new skills, problem solving…performing complex tasks and acquiring new knowledge and know-how” (p. 797). We focused on reciprocal “enduring patterns of proximal process[es]” (p. 797) that regularly occurred in children’s lives, including (a) the degree to which parents engaged with their children at home; (b) the extent to which children engaged in learning activities at home during which they interacted with “objects and symbols” that required “attention, exploration, manipulation, elaboration and imagination”1 (p. 798); and (c) the degree to which children interacted positively with others in the school environment. These types of proximal processes have been shown to be the cornerstone of the developing child and to be potentially relevant to absenteeism behavior.


PERSON


These factors capture particular dispositions of children, their biopsychological attributes, as well as their ascribed characteristics. Specifically, we looked at what Bronfenbrenner and Morris (2006, pp. 810–813) describe and define as person forces (p. 810), resource characteristics (pp. 812–13) and demographic characteristics (p. 816).


Forces are “active behavioral dispositions” (p. 810) that can function to support or impede proximal processes that underlie development. Forces that support proximal process are known as “developmentally generative,” while those that impede are known as “developmentally disruptive” (p. 810). Generative forces that influence the processes we address include approaches to learning (i.e., an eagerness to learn new things, working independently and adapting to changes in routines, etc.), self-control, and the degree to which a child likes school. All of these forces enhance the likelihood that the child is able to engage in proximal processes (i.e., engagement with their learning environment) shaping their attitude towards school and perhaps ultimately influencing absenteeism behavior. On the other hand, disruptive forces include externalizing and internalizing behaviors—both can make it difficult to “engage in proximal processes requiring progressively more complex patterns of interaction over time” (Bronfenbrenner & Morris, 2006, p. 810). Further, these behavioral dispositions can enhance the likelihood that a child will be chronically absent.


Resource characteristics (Bronfenbrenner & Morris, 2006, p. 812) capture “biopsychological liabilities and assets” (p. 812) that are critical in shaping children’s development and their capacity to engage in their surroundings. Also, they can be considered “developmental assets in the form of ability knowledge, skills, and experience” (p. 812). As noted earlier, these types of resource characteristics have been shown to be strong predictors of chronic absenteeism. Examples of resource characteristics include a child’s disability status, academic ability, and overall health status.


Finally, our study focuses on a set of demographic factors (p. 814) specified by Bronfenbrenner and Morris (2006) which include age, gender, and race/ethnicity. They note that these factors “are so pervasive in affecting future development that their possible influence routinely needs to be considered in relation to the particular phenomenon under investigation” (p. 814).


CONTEXT


Contexts (or systems) (micro- meso-, exo- and macrosystems) involve children’s proximal and distal environments. The microsystem represents “patterns of activities, roles, and interpersonal relationships” within the “immediate environment” (i.e., school, home, etc.) of the child (p. 814). In our study, we focused on the role of two influential microsystems: children’s immediate families and their schools. We incorporated aspects of these microsystems into our study in two different ways: (1) via the proximal processes described above that occurred primarily in the microsystems of the home and school; and (2) by including specific attributes of the family microsystem. Finally, we examined children’s prior exposure to the microsystems of their daycare environments.


The mesosystem captures the “linkages and processes taking place between two or more settings containing the developing person” (Bronfenbrenner, 1993, p. 40). We examined the mesosystem comprised of the intersections between the microsystems of the family and school. Thus, in our study, the mesosystem captures the home–school connections, such as the degree to which parents engage with schools. Prior research has shown that establishing linkages between the microsystems of the home and the school, especially via school and family partnership programs, can significantly decrease chronic absenteeism (Sheldon & Epstein, 2004).


Similar to the mesosystem, the exosystem “comprises the linkages and processes taking place between two or more settings” (Bronfenbrenner and Morris, 2006, p. 818). Yet in contrast to the mesosystem, at least one of the multiple settings does not include the immediate environment surrounding the child. However, the setting that does not include the child indirectly influences processes within the child’s immediate setting. An example of this is a parent’s work environment away from home (Bronfenbrenner, 1993). The child is not directly part of the parent’s work setting, yet, the demands placed on parents in their work environment (i.e., time engaged in work away from the home) may indirectly influence the child via the amount of time parents have to care for their child at home (Weiss, 2005). With respect to absenteeism, if parents’ workplace setting severely constrains the amount of time they can spend at home, their ability to be involved in their child’s education is impeded; given that parental involvement is a key predictor of chronic absenteeism (Sheldon & Epstein, 2004), relatively lower levels of involvement and engagement may lead to higher likelihood of absenteeism. There are additional contexts not directly involving the child that have been posited to indirectly influence absenteeism, including broader labor market conditions as well as school district policies on absenteeism (Kearney, 2008).


Finally, the macrosystem is comprised of the “overarching pattern of micro-, meso- and exosystems characteristic of a given culture or subculture” which includes “belief systems, bodies of knowledge, material resources, customs, life-styles, opportunity structures…and life course options embedded in each of these broader systems” (Bronfenbrenner, 1993, p. 40). In our study, we focus on the macrosystem involving the socioeconomic status of families by comparing how process, person, and context factors differentially influence absenteeism behaviors across three distinct SES groups (Tudge, Odero, Hogan, & Etz, 2003).


While prior studies on absenteeism have been motivated by examining multiple factors of absenteeism (see e.g., Chang & Romero, 2008; Kearney, 2008; Rothman, 2001), none have considered how process, person, and context factors co-occur to jointly influence absenteeism in a theoretically grounded model. This is a critical oversight, as prior research that focuses on a single or small set of factors has provided only a limited understanding of what factors might be correlated with differences in chronic absenteeism rates. We, however, provide a more extensive and holistic examination of factors of chronic absenteeism. By grounding our work within a bioecological framework of development, we can determine not only which factors are significantly linked to absenteeism, but also the relative importance of each. Doing so allows schools and policy leaders to identify key factors, discern between their relative influence, and subsequently channel resources in a way that addresses how to best reduce the continuation of the damage caused by missing school. Given this motivation, in our study, we asked: What were the bioecological factors associated with chronic absenteeism among kindergarteners in the US? With this question, we are the first, to our knowledge, to apply bioecological theory to understand salient predictors of chronic absenteeism using a nationally representative sample of kindergarten children.


METHOD


PARTICIPANTS


We relied on a comprehensive dataset of kindergarten students in the US. Created by the National Center for Education Statistics (NCES) at the U.S. Department of Education, the Early Childhood Longitudinal Study, Kindergarten Class of 2010–2011 (ECLS-K:2011) represents the most recent data collected on kindergarten students and their families, classrooms, and schools. The students in ECLS-K:2011 represent a diverse range of school types, socioeconomic levels, and racial and ethnic backgrounds (for a description of the data collection and sampling procedures of ECLS-K:2011, see Mulligan, Hastedt, & McCarroll, 2012).


The analytic sample contains N=9,350 students. Note that sample sizes have been rounded to the nearest tens based on the restrictions set up by NCES’s policy. The analytic sample is limited to children who had nonmissing information on all measures and on school identification number. We compared mean differences for all measures in our analyses based on students who were included and excluded from the sample. For our set of measures, mean differences did not arise between groups, giving us a degree of confidence that observations were missing at random.


MEASURES


Table 1 describes the predictor variables we included in our analysis. Table 2 presents descriptive statistics for dependent and independent measures in this study. The outcome of interest, i.e., chronic absenteeism, is derived from the teacher’s assessment of the student. In the spring survey of ECLS-K:2011, a teacher was asked to report the number of absences of the student in the sample over the course of the school year. There are a discrete number of answer choices on the teacher assessment, including 0, 1 to 4, 5 to 7, 8 to 10, 11 to 19, and 20 or more.


Table 1. Predictor Variables from the Early Childhood Longitudinal Study-Kindergarten Class of 2010–2011 (ECLS-K:2011) Used to Represent Components of Bronfenbrenner’s Bioecological Model of Development as They Relate to Chronic Absenteeism

Bioecological Component

Predictor

Process

Reciprocal interactions occurring between children and other individuals, symbols and objects in their immediate contexts.

Child’s:

Learning activities at home (interactions with objects and symbols)

Parents’ involvement in home (parent–child interactions)

Interpersonal behavior at school (child–peer interactions)

Person

The dispositions of children, their biopsychological attributes, as well as their ascribed characteristics

Forces (“active behavioral dispositions” that promote positive development)

Approaches to learning (i.e., keeps belongings organized; shows eagerness to learn new things; works independently)

Self-control

Liking school


Resource characteristics (“biopsychological liabilities and assets”; “developmental assets in the form of ability knowledge, skills, and experience”)

Externalizing problems

Internalizing problems

English language learner (ELL) status

Disability status

Math ability

Body mass index (BMI)

Health status (Excellent, Very Good; Good; Fair)


Demographic factors

Gender

Race/ethnicity (White, Black, Latino, Asian, Other)

Age at kindergarten entry

Microsystem

The immediate environments of children

Child’s:

Number of siblings

Mother’s marital status (at time of birth)

Socioeconomic status

Parents’ marital status

Out-of-home pre-K care

Exosystem

Connection between two or more settings, one of which does not contain the child

Parents’ weekly hours of employment

Mesosystem

Connections between the microsystems of home and school

Child’s parents’ attendance at:

Back-to-school night

PTA/PTO meeting

Parent–teacher conference

Volunteer event

Macrosystem

“Overarching pattern of micro-, meso- and exosystems characteristic of a given culture or subculture…”

Household socioeconomic status (SES)


Table 2. Descriptive Statistics (n=9,350)


[39_21802.htm_g/00002.jpg]


Some researchers define a threshold for chronic absenteeism, suggesting that it begins after missing a cumulative of 2 weeks of school, while others indicate that chronic absenteeism occurs after missing more than 18 days of school (e.g., Gottfried, 2014; Balfanz & Byrnes, 2012). To be the most inclusive of all possible definitions, the chronic absenteeism measure in our study equals 1 if a student has missed more than 2 weeks of school (i.e., 11 or more days) and 0 otherwise. As shown in Table 2, approximately 12% of the sample was chronically absent, which is consistent with prior research (Gottfried, 2014; Balfanz & Byrnes, 2012).


ELEMENTS OF THE BIOECOLOGICAL MODEL OF DEVELOPMENT


Using Bronfenbrenner and Morris (2006) as a guide, we systematically identified a set of relevant bioecological factors representing process, person, and context that were captured in the ECLS-K:2011 dataset. To further guide our selection of factors most salient to absenteeism, we relied on the extant empirical and theoretical literature. We incorporate many of these factors as independent variables in our analysis. In Table 1, we summarize the constructs we used.


Although it is not possible to incorporate every measure representing the different elements of the bioecological model in our present study,2 based on our constructs, we argue that our study still holds tremendous value and insight for understanding chronic absenteeism from a bioecological standpoint. As Bronfenbrenner and Morris (2006) note: “Even when the theoretical and operational requirements of the bioecological model are not met in full, the results can still contribute to understanding the forces that shape human development.” (p. 813).


Process


As we previously discussed in our Background and Context section, we focused on processes reflecting the several types of “reciprocal interactions” between children and “the persons, objects, and symbols” in their “immediate external environment” (Bronfenbrenner & Morris, 2006, p. 797). To capture these patterns of proximal processes, we focused on three measures: (a) children’s learning activities at home; (b) the involvement that parents have with children at home (parent–child interaction); and (c) the way children interact with others at school (child–child interactions).


The measures of learning and parental involvement at home were replicated from Votruba-Drzal, Li-Grining, and Maldonado-Carreño (2008). Both measures assessed the child’s home environment, activities, and cognitive stimulation. The number of learning activities in which children engaged at home comprised of 15 dichotomously scored items that assessed whether in the past month the child visited a bookstore, took music lessons, or attended tutoring lessons. Parental involvement was measured on a 4-point Likert scale comprising 10 items that assessed the frequency with which parents engaged the child in playing games, singing songs, reading books, and doing arts and crafts.


Finally, we included a child’s score on the teacher-rated interpersonal skills scale, which measures the frequency with which a child has been getting along with people, forming and maintaining friendships, helping other children, showing sensitivity to the feelings of others, and expressing feelings, ideas, and opinions in positive ways. This scale is continuous and represents the average of a series of questions pertaining to the frequency of a particular item ranging from 1 (never) to 4 (very often). We do acknowledge that these proximal processes are “one-sided” (Bronfenbrenner & Morris, 2006, p. 800) since we do not fully capture the reciprocal nature of the interactions occurring (i.e., the degree to which children responded to their parents’ learning activities, or how other children responded to the child’s positive behavior).


Person


As indicated in Table 1, the first set of child-level factors pertain to person forces, or behaviors and dispositions that influence the above proximal processes. Consistent with Bronfenbrenner and Morris (2006), we classify these forces as either (a) developmentally generative or (b) developmentally disruptive. These forces are based on a set of socioemotional scales derived from the teacher’s assessment of student behavior in both the fall and spring kindergarten survey waves. NCES based all socioemotional scales on the Social Skills Rating System (SSRS) developed by Gresham and Elliot (1990); however, they modified these original scales and created Teacher Social Rating Scales (SRS) in ECLS-K:2011. Each scale is continuous and represents the average of a series of questions pertaining to the frequency of a particular item ranging from 1 (never) to 4 (very often). NCES reports that all scales have high internal consistency, with the reliability coefficients ranging from 0.79 to 0.89.


For developmentally generative person forces, we focus on the child’s (a) approaches to learning and (b) self-control, which are captured via teacher’s ratings in the Teacher Social Rating Scales (SRS). The first scale rates a child’s frequency of organization, eagerness to learn new things, independent work ability, adaptability to change, persistence in completing tasks, and ability to pay attention. The second scale measures the frequency of the student’s ability to control his or her temper, respect others’ property, accept peers’ ideas, and handle peer pressure. A higher score on these scales indicates higher positive socioemotional skills (i.e., positive regression coefficients represent positive outcomes). Finally, we included a parental survey rating of the frequency with which a child expresses liking school. This measure ranges from 1 to 3 with 3 representing the greatest frequency of expressing liking school.


For developmentally disruptive person forces, we focus on the problem behaviors scales of the SRS which include externalizing and internalizing problem behaviors in school. The externalizing problem behaviors scale measures the frequency with which a child argues, fights, gets angry, acts impulsively, and disturbs ongoing activities. The internalizing problem behaviors scale rates the presence of anxiety, loneliness, low self-esteem, and sadness. A higher score on these two problem behaviors scales reflects a lower outcome (i.e., negative coefficients represent positive outcomes).


The person resource characteristics we included capture the “biopsychological liabilities and assets” of the child. These are indicators for whether a child is an English language learner (ELL) based on the primary language spoken at home and whether a child has a diagnosed disability based on school individualized education program (IEP) records. Additional characteristics included math ability at kindergarten entry based on scores scaled using item-response theory (IRT). Finally, to account for children’s physical health, we included both body mass index (BMI) and parents’ rating of their child’s health, ranging from fair to excellent. According to prior literature, poor health may be a potential liability—children who are less healthy might have greater absences (Allen, 2003; Bloom, Dey, & Freeman, 2006).


Finally, we have included a typical set of demographic factors explicitly mentioned by Bronfenbrenner and Morris (2006), including a child’s age at kindergarten entry in months, ethnicity/race, and gender.


Context: Microsystem


The measures of proximal process we addressed previously function primarily within the microsystems of the family and school. We also included additional characteristics of family structure, including the number of siblings, whether a student’s mother was married at his/her birth, and whether a student’s parents were currently married. Moreover, we included family SES, as measured by a composite scale constructed by NCES based on family income, parental education attainment, and parental occupation.


Finally, we included characteristics of the microsystem of a child’s childcare environment­­ prior to kindergarten: relative, nonrelative, center-based care, Head Start, and multiple types. Additionally, hours of nonparental care was included. A similar set of indicators (except for Head Start) were included to account for the type of care a student had received outside of kindergarten time (e.g., afterschool care).


Context: Exosystem


We included parents’ reported working hours as a way to address the exosystem. As noted earlier, the parental workplace and by extension, the time parents spend at work is not part of a child’s immediate environment (it is, however, part of a parent’s microsystem), but it may have an indirect effect on a child.


Context: Mesosystem


The measures we used to focus on the linkages between the microsystems of family and school included four indicators that capture the extent to which parents reported their engagement in school-related activities: i.e., whether they had (a) attended back-to-school night; (b) attended a parent–teacher association (PTA) or parent–teacher organization (PTO) meeting; (c) attended parent–teacher conference; or (d) volunteered at school.


Context: Macrosystem


Empirical work of Tudge et al. (2003) guided our approach to incorporating the macrosystem in our analysis. Tudge et al. (2003) focused on two groups of families based on their socioeconomic status, noting that “Bronfenbrenner argued that to understand development, the research design must involve ‘a contrast between at least two macrosystems…’ Thus, one can satisfy the minimum requirement by conducting cross-cultural research as it is typically understood, or by examining groups that are distinguished by race, ethnicity, or social class within a single society” (p. 47). Accordingly, in our research design we stratified our analyses by low, middle, and high SES groups. This allowed us to understand how selected process, person, and contextual factors underlying chronic absenteeism behaviors function across SES.


DATA ANALYTIC METHOD


Hierarchical Generalized Linear Modeling (HGLM)


To estimate the relationship between our chronic absenteeism outcome and selected measures representing process, person, and context variables, we fitted the following 2-level hierarchical generalized linear model (HGLM)3 (Raudenbush & Bryk, 2002) model to account for the nesting of children (i) ­in schools (j):


Level-1 (Child)
      [39_21802.htm_g/00004.jpg]

                                                                                                                                                           (1)

Level-2 (School)

[39_21802.htm_g/00006.jpg]

[39_21802.htm_g/00008.jpg], for q=1,…,Q

(2)


Substituting level-2 into level-1 and rearranging yields:

[39_21802.htm_g/00010.jpg]

(3)

where [39_21802.htm_g/00012.jpg]is the probability of chronic absenteeism, [39_21802.htm_g/00014.jpg]represents a set of level-1 variables capturing our selected process, person, and contextual predictors and [39_21802.htm_g/00016.jpg] represents a random effects at level-2. We constrained the effects of the level-1 predictors ([39_21802.htm_g/00018.jpg]) to be fixed at level 2, assuming that they are invariant across schools and let the intercept ([39_21802.htm_g/00020.jpg]) be random across schools. To facilitate the interpretation of our models, we present our coefficient estimates as odds ratios by exponentiating our parameter estimates for the fixed effects in our models (i.e., [39_21802.htm_g/00022.jpg]).


Prior to fitting the full model as specified in equation (3), we first fit an unconstrained (no predictor) model as a benchmark to determine the proportion of total variability in our outcome that was attributable to between school differences (i.e., the intracluster correlation coefficient). Given this approach, the unconditional intracluster correlation coefficient (ICC) for our unconstrained model was [39_21802.htm_g/00024.jpg] = 0.17. Note that in a logistic HLM model, the level-1 error variance is calculated as [39_21802.htm_g/00026.jpg], an approach that treats the dichotomous outcome as a latent variable (Snijders & Bosker, 2012). Based on this ICC, approximately 17% of the variation in the chronic absenteeism outcome was due to school-to-school differences. The remaining 83% was due to between-children differences, which indicates that children’s characteristics are somewhat correlated within school. This degree of interdependence justifies our use of HGLM as our estimation strategy. We then fit models augmenting this unconditional model with sets of predictors previously described in our measures section. We used Stata 13 to fit all HGLM models to our data using maximum likelihood estimation (MLE) (StataCorp, 2013).


RESULTS


MAIN FINDINGS


Table 3 presents the odds ratios and standard errors for the model predicting chronic absenteeism. The model accounts for the multilevel structure of the data of students clustered within schools. Note that the coefficients are presented as odds ratios. A coefficient larger than a value of 1 indicates higher odds of chronic absenteeism, whereas a coefficient lower than 1 suggests lower odds. An odds ratio that approximates 1 would not be associated with any change in odds.


Table 3. Bioecological Predictors of Chronic Absenteeism (Main Model)


[39_21802.htm_g/00028.jpg]


Process


Given the focal role of processes in shaping development as described above, we first focus on the most salient proximal process in our model that was significantly related to chronic absenteeism. We found that children with higher levels of home learning activities had higher odds of chronic absenteeism. While this result may seem counterintuitive at first, we theorized that parents providing a larger number of home activities might be more likely to accommodate absences for two reasons. First, some highly involved parents might have the capacity to stay at home with an absent child, such as a mother who does not work (Author, 2015). Second, these parents might feel as though missed in-school time can be supplemented with home learning activities; hence, the opportunity cost of missing school in these families declines. As this study is associative in nature, the correlation between these two measures is not necessarily surprising.


Person


There were several important findings with respect to person forces. As expected, students who displayed greater frequencies of approaches to learning had lower odds of being chronically absent. In other words, positive attitudes towards school were associated with lower instances of missing school. Greater instances of other teacher-rated behaviors, however, were associated with differences in school absences. For instance, students with higher frequencies of externalizing behaviors were less likely to be chronically absent. It is plausible that parents rely on schools to provide care for students with higher instances of behavioral issues. Moreover, students with higher internalizing behaviors were more likely to be chronically absent. This finding is consistent with literature on disengaged or anxious students who have a higher likelihood of being absent from school (Ekstrom et al., 1986). These findings for externalizing and internalizing behaviors are not contradictory. Children who act out are likely to attend school (perhaps due to parental agency), and those who are more anxious or lonely tend not to. Finally, students who expressed liking school were less likely to be chronically absent. This final finding—based on a parent-rated measure—corroborates the teacher-rated approaches to learning scale also described in this section. Hence, there is a great amount of consistency across multiple measures in this domain.


Of all person resources measures, only health was significantly related to chronic absenteeism. Compared to students in excellent health (the reference group), students with only good or fair health ratings had higher odds of chronic absenteeism. Students with the lowest health rating were most likely to be chronically absent compared to any other health level, a finding consistent with extant literature (Allen, 2003; Bloom et al., 2006). Hence, health remained a persistent and highly significant predictor of chronic absenteeism, even after accounting for our wide span of factors.


Among our set of person demographic characteristics, only race was a statistically significant predictor of chronic absenteeism. Asian students and those from other racial/ethnic categories (besides Blacks and Hispanics) were more likely to be chronically absent relative to their White counterparts. However, we interpret these findings for race with caution, given the smaller sample sizes of these demographic groups in the ECLS-K dataset.


Context: Microsystem


Both family structure and SES predicted differences in chronic absenteeism. Specifically, students with more siblings tended to have lower odds of chronic absenteeism. As Author (2014) suggested, in larger families, there might be more accountability when it comes to getting more than one child to school. Additionally, students from higher SES families had lower odds of chronic absenteeism, as corroborated by the literature (Nauer, Mader, Robinson, & Jacobs, 2014). Childcare experiences were also linked to differences in odds of chronic absenteeism. Across both prekindergarten and kindergarten sets of care variables, children in center-based care were less likely to be chronically absent in kindergarten. Author (2015) suggests that this is because center-based care provides families with the necessary impetus to help solidify family logistics surrounding getting children to kindergarten as well as to help students develop the mindset of attending school on a regular, daily basis. Note that the indicator for hours of care, while significant, was associated with an odds of almost one-to-one.


Context: Exosystem


While the measure of hours of maternal employment was statistically significant, there was little practical significance given that the odds ratios were approximately one-to-one. The effect of paternal employment was not statistically significant.


Context: Mesosystem


While home parental involvement was not statistically significantly associated with chronic absenteeism, several attributes in the mesosystem were linked to missing school. The direction of the effect of these mesosystem factors was as expected: students whose parents attended back-to-school nights or PTA/PTO meetings were less likely to be chronically absent. We speculate that parents who were more invested in or involved with the school would have a greater likelihood of ensuring their children attended school.


THE INTERACTIVE ROLE OF SES


Within policy and practice dialogue, lower SES is supported as having a major influence on students’ chronic absenteeism (Nauer et al., 2014). Children from low SES backgrounds often face greater health issues (Allen, 2003), fewer family resources (Chang & Romero, 2008), and more negative attitudes about going to school (Chang & Romero, 2008). Hence, these challenges that children from lower SES families face are often associated with a higher prevalence of chronic absenteeism (Balfanz & Letgers, 2004; Nauer et al., 2014; Orfield & Kornhaber, 2001). Therefore, understanding how SES moderates our model will better inform researchers, policymakers, and practitioners about how to make adjustments and guide policy to address specific needs for those students with limited opportunities.


To develop a more nuanced understanding of how being in a particular SES group might moderate our original findings, we stratified our sample of students into tertiles based on the NCES SES measure and then refitted our model from Table 3 on each SES subsample (low SES, middle SES, and high SES). Since each model was separately estimated by socioeconomic subgroup, each represents a fully interacted model. Hence, the findings here represent interaction effects between SES as a categorical measure (low, middle, or high) and all other factors in the model.


Table 4. Models by Socioeconomic Status


[39_21802.htm_g/00030.jpg]


Table 4 presents the odds ratios for our findings for each SES group. For clarity, we omit reporting standard errors in the table. As in Table 3, the only process measure that is statistically significant is home learning activities. Interestingly, the microsystem measure of SES appears to moderate process effects; it was only in highest SES families that students with greater home learning activities were more likely to be chronically absent (Table 4). This provides even greater support for the speculation based on the findings in Table 3. The effect we detected may be because higher SES parents feel that they can supplement missed school time with their ability to be involved in the child’s learning outside of school—in other words, the opportunity costs of missed school time are lower for highly involved higher SES families. Or, it might be the case that most of home learning activities of the highest SES households, such as sports, music recitals, or travel, often involve missing school.


We also found that parent and family measures in micro- and mesosystems were moderated by SES. For instance, we found two key features of the family microcontext that interact to shape absenteeism outcomes: SES and siblings. For the middle SES group, the number of siblings, as a measure of family structure, influenced absenteeism rates. Hence, once our models were stratified by SES, a more refined portrait emerged as to how family structure might play a significant role shaping chronic absenteeism. Similarly, when SES, again a microcontextual factor, was interacted with parent–school factors of the mesosystem, the largest benefits (i.e., smaller odds of chronic absenteeism) emerged for families from the higher SES subgroup. Therefore, students from highest SES families had the lowest odds of chronic absenteeism when their families were involved in school activities. Students from the low SES subgroup did benefit when their parents attended a back-to-school night, however, the odds were significant, but not as large.


As addressed previously, empirical evidence links several person-level measures to absenteeism. Interestingly, there were several person resources factors that were shaped by SES. The role of ELL and disability status emerged as significant predictors for students in the lowest SES group; both low SES ELL and disabled students were less likely to be chronically absent. One plausible explanation is that the schools provide access to services for both ELL students and students with disabilities—services that the families of limited financial means rely on and could not get elsewhere. Therefore, low SES families are more likely to ensure that their children attend school to access these services. Alternatively, at school, these children can receive specialized services that families may not be able to receive or afford at home. There were differences in findings by race/ethnicity; however, the sample sizes in each category reduce the confidence in them and are hence not discussed here.


The findings in this section help us understand how and when these person-level factors were shaped by SES. Beginning with the latter, irrespective of SES, expression of liking school was linked to lower chronic absenteeism. Similarly, health was not differentiated across SES groups. Together, the lack of interaction effects between SES and these person-level factors serve to highlight the importance of school-going attitudes as well as the importance of strong health, regardless of SES.


DISCUSSION


Grounding our work in Bronfenbrenner’s bioecological model of development, we analyzed the relationship between process, person, and context factors and chronic absenteeism for a recent national cohort of kindergarteners drawn from the Early Childhood Longitudinal Survey (ECLS-K:2011). We used the method of hierarchical generalized linear modeling (HGLM) to account for students nested within schools. The bioecological model was influential to our study because it allowed us to systematically identify salient correlates of chronic absenteeism. Our results, as theoretically supported by the bioecological model of development, demonstrated that children’s absenteeism behaviors were not tied to one single attribute or context. Rather, the risk factors for chronic absenteeism simultaneously involved attributes of children, their environments, and the interactions they engage in within those environments. The bioecological model offered a highly relevant framework with which to investigate absenteeism behaviors since it provided us with a holistic theoretical perspective—one lacking in prior absenteeism research—with which to understand children’s absenteeism.


Using the bioecological model to guide our selection of correlates, we found that the probability of chronic absenteeism was associated with a fairly consistent set of factors. The fact that there was consistency among these factors was in itself an important finding. As mentioned in the introduction, most research in the area of absenteeism focused on a single or limited set of factors associated with absences. Therefore, in prior research, it was difficult to evaluate the relative importance of each factor; consequently, the field was left with a slew of statistically significant relationships without having any guide to help determine which factor might be more important than others. This makes it difficult to design and develop policies and programs to support absenteeism reduction. This study took a different perspective, however. Here, we focused more carefully on the “whole child,” enabling us to identify and test the significance of numerous factors simultaneously. In doing so, these following, noteworthy findings emerged.


First, children who displayed greater frequencies of approaches to learning (or who expressed liking school) were less likely to be chronically absent, whereas those who displayed greater frequencies of internalizing behaviors were more likely to be chronically absent. Prior research shows that students who are disengaged from school or feel alienated or isolated were more chronically absent (Ekstrom et al., 1986; Finn, 1993; Johnson, 2005; Newmann, 1981), and our findings corrobate this. From a policy and practice perspective, these findings suggest that efforts to reduce chronic absenteeism vis-a-vis student resource factors might be targeted at boosting engagement towards learning and positive feelings towards school. In fact, there has been prior success with interventions to improve school engagement. Previous efforts, such as “Check & Connect” (Lehr et al., 2004), have shown that monitoring and addressing student engagement is linked to lower instances of absenteeism. Further efforts in this line would be supported by the evidence in this present study.


Kindergarteners’ health status also emerged as significant. Regardless of SES, children in poorer states of health had higher odds of being chronically absent relative to their healthier peers, as found in prior research (Allen, 2003; Jackson, Vann, Kotch, Pahel, & Lee, 2011; Krenitsky-Korn, 2011; Pourat & Nicholson, 2009). While it is widely accepted that access to important health care services and federal health programs for young children, such as Medicaid, can overcome children’s impaired school readiness (Currie, 2005) as well as absenteeism (Zhang, 2012), health-related coverage and services still remain out of reach for many children. Given our finding of a strong health-attendance relationship, encouraging participation for children and families in programs (both on and off school sites) that support health and well-being of children is critical to both health and successful school-going behavior. Additionally, it is important to identify which particular health issues are most likely to be associated with chronic absenteeism. For example, children with asthma are 3.2 times more likely to be chronically absent than their peers without asthma (Krenitsky-Korn, 2011). Therefore, there remain opportunities to address not only general issues pertaining to access to health care but also ways to develop school supports for specific health needs.


We found that children in families with more siblings tended to have lower odds of being chronically absent. Children with additional siblings (especially older ones) might be under more supervision, which has been shown to be critical in reducing absenteeism (Sampson & Laub, 1994). Or, it might be the case that children with many siblings can walk to school together or take public transportation together, particularly in more densely populated areas (Gottfried, 2010, 2014); this, however, may not be possible for children with fewer or no siblings. Determining the precise mechanisms underlying this finding remains open for future research. Nonetheless, in going forward, varying supports might be considered. For example, school and community supports might ensure that children in small families attend school as frequently as children in larger families who have more “in-house” support. Prior research has suggested that a stronger school–community–family partnership can lower child absenteeism (Sheldon, 2007). Our findings, therefore, support interventions in which schools and communities could work in tandem with smaller families to promote good attendance by offering programs to walk children to school, escort them on public transportation, and carpool.


Finally, several meso-contextual factors proved to be significant predictors of chronic absenteeism. For example, children with parents who were more involved at school tended to have lower odds of chronic absenteeism. Higher rates of absenteeism are linked to families in which parents are absent from, unaware of, or uninvolved in their children’s schooling (Catsambis & Beveridge, 2001; Fan & Chen, 2001; Jeynes, 2003; McNeal, 1999; Muller, 1993). Our findings underscore this. Given that parental detachment from school has been found to increase child absenteeism (Lehr et al., 2004; Sheldon, 2007), schools might continue identifying new ways to spur parental engagement. Perhaps the set of straightforward indicators utilized in this study to measure the role of meso-context might aid schools in identifying ways to boost the importance of being present in school activities. But for those families who cannot be part of school functions for various reasons, such as work schedules, awareness is critical. Research suggests that increased communication to the parents about a child’s schooling can help to reduce absence rates (Epstein & Sheldon, 2002). Hence, our study would support efforts to increase both involvement in and awareness about school activities.


Finally, a key set of findings in our study was related to the moderating role of SES. As established previously, SES holds significant predictive power in increasing and/or reducing chronic absenteeism (e.g., Gottfried, 2009; Nauer et al., 2014; Ready, 2010). As hypothesized, children from lower SES families tended to have higher odds of being chronically absent. This relationship has been previously established and discussed extensively in the literature on chronic absenteeism. However, the value of this finding in this particuar study comes from the fact that even after accounting for a wide span of factors in the bioecological model, SES still continued to emerge as a significant interactive factor, as seen in Table 4. Understanding how effects differ by subgroups provides a platform for both researchers and policymakers to develop an agenda that attemps to parse the factors that not only matter the most, but determine exactly for whom they matter.


Given our findings, our study supports continued efforts at reducing chronic absence, particularly in kindergarten, and provides theoretically grounded evidence as to which factors may be more important in considering ways to reduce absenteeism. As mentioned, of all elementary school years, kindergarten has the highest chronic absenteeism rates and hence, many children are missing the most amount of school during one of the most formative and critical years for their development (Juel, 1988; Pianta & Walsh, 1996; Smith, 1997). Determining how to promote positive influences and mitigate negative factors during this critical year will thus serve to not only better shape research, policy, and practice around this detrimental schooling behavior but aid in creating stronger interventions and policies to curb absenteeism as well.


FUTURE RESEARCH


Future endeavors can build upon our study in several meaningful ways. First, one potential limitation of this study is that it focused exclusively on a cross section of kindergarteners. While kindergarteners face the highest incidence of chronic absenteeism, a further adaptation of the bioecological model could be to analyze factors shaping how absenteeism behaviors change over time. In incorporating changes over time, future research might consider factors contributing to persistent chronic absenteeism gaps and compare whether the factors in any single year are different from those that affect chronic absenteeism as children grow older.


A second potential limitation is that this study focused on the correlates of chronic absenteeism. We do not causally attribute any of the factors to absenteeism; yet, the strength of our design yields important insights into the patterns of absenteeism and the likely influences that deserve further study. Hence, our study is useful in guiding future research designs—with some drawing upon experimental or quasiexperimental methods—that will help distinguish correlates of chronic absenteeism from causal impacts. For instance, it might be possible to randomly assign health interventions to families and children to determine the role that they play in reducing school absenteeism. Or, one could implement a school engagement program like “Check & Connect” (Lehr et al., 2004) to determine the effects of monitoring and addressing attitudes towards school. Quasiexperimentally, it could be possible to test for the effects of having another sibling in the family, with the appropriate panel data allowing one to monitor children over time and hence changes to family structure over time. While many of the factors in this study cannot necessarily be randomly assigned, such as number of siblings, there are certainly ways to address the effects of these factors.


Finally, in our study, it was not possible to determine the underlying mechanisms through which these significant factors influenced chronic absenteeism. Research designs based on qualitative methods have a strong potential to provide additional insight into mechanisms driving the consistent findings in our study.


In conclusion, chronic absenteeism in early education poses enormous challenges for young students as it constrains their ability to achieve positive educational, health, and developmental successes, both in the short and long term. Therefore, future work in this area will allow for us to further detect origins of this problem in school and reduce its negative consequences.


Acknowledgement


The authors of this study received funding from the Spencer Foundation. This article reflects the work of the authors and not of the granting agency.


Notes


1. According to Bronfenbrenner and Morris (2006), these types of learning activities “can be carried on in the absence of other persons” and are considered “solo activities” (p. 814).

2. Bronfenbrenner’s bioecological model also incorporates the element of time. However, we do not explicitly include time given that as of the writing and completion of this study, our primary outcome is measured at a single time point.

3. HGLMs are similar to hierarchical linear models, except that HGLMs are used in cases where standard assumptions of linearity and normality do not hold. In our case, our chronic absenteeism outcome is dichotomous, which makes HGLM an appropriate method with which to analyze our data. As noted in Raudenbush & Bryk (2002), HGLMs are also referred to as generalized linear mixed models or generalized linear models with random effects (p. 292).


References


Alexander, K. L., Entwisle, D. R., & Horsey, C. S. (1997). From first grade forward: Early foundations of high school dropout. Sociology of Education, 70(2), 87–107.


Allen, G. (2003). The impact of elementary school nurses on student attendance. Journal of School Nursing, 19(4), 225–231.


Balfanz, R., & Byrnes, V. (2012). The importance of being there: A report on absenteeism in the nation’s public schools. Baltimore: Johns Hopkins University Center for Social Organization of Schools.


Balfanz, R., & Legters, N. (2004). Locating the dropout crisis. Which high schools produce the nation’s dropouts? Where are they located? Who attends them? Baltimore, MD: Johns Hopkins University, Center for Research on the Education of Students Placed At Risk.


Bealing, V. (1990). Pupil perceptions of absenteeism in the secondary school. Maladjustment and Therapeutic Education, 8(1), 19–34.


Bloom, B., Dey, A. N., & Freeman, G. (2006). Summary health statistics for U.S. children: National Health Interview Survey, 2005. Washington, DC: U.S. Department of Health and Human Services: National Center for Health Statistics.


Broadhurst, K., Paton, H., & May-Chahal, C. (2005). Children missing from school systems: Exploring divergent patterns of disengagement in the narrative accounts of parents, carers, children and young people. British Journal of Sociology of Education, 26(1), 105–119.


Bronfenbrenner, U. (1993). Ecological models of human development. In M. Gauvain & M. Cole (Eds.), Readings of the development of children (2nd ed., pp. 37–43). New York, NY: Freeman.


Bronfenbrenner, U., & Morris, P. (2006). The bioecological model of human development. In R. Lerner & W. Damon (Eds.), Handbook of child psychology: Vol. 1. Theoretical models of human development (6th ed., pp. 793–828). Hoboken, NJ: John Wiley & Sons, Inc. http://doi.org/10.1002/9780470147658.chpsy0114


Catsambis, S., & Beveridge, A. A. (2001). Does neighborhood matter? Family, neighborhood, and school influences on eighth-grade mathematics achievement. Sociologcial Focus, 34(4), 435–457.


Chang, H. N., & Romero, M. (2008). Present, engaged, and accounted for: The critical importance of addressing chronic absence in the early grades. New York, NY: Natiaonl Center for Children in Poverty.


Connolly, F., & Olson, L. S. (2012). Early elementary performance and attendance in Baltimore City schools’ pre-kindergarten and kindergarten. Baltimore, MD: Baltimore Education Research Consortium.


Currie, J. (2005). Health disparities and gaps in school readiness. The Future of Children, 15(1), 117–138.


deJung, J. E., & Duckworth, K. (1986). Measuring student absences in the high schools. Eugene, OR: Center for Educational Policy and Management.


Dryfoos, J. G. (1990). Adolescents at risk: Prevalence and prevention. New York, NY: Oxford University Press.


Duncan, G. J., Dowsett, C. J., Claessens, A., Magnuson, K., Huston, A. C., Klebanov, P., … Japel, C. (2007). School readiness and later achievement. Developmental Psychology, 43(6), 1428–1446.


Ekstrom, R. B., Goertz, M. E., Pollack, J. M., & Rock, D. A. (1986). Who drops out of high school and why? Findings from a national study. Teachers College Record, 87(3), 356–373.


Epstein, J. L., & Sheldon, S. B. (2002). Present and accounted for: Improving student attendance through family and community involvement. Journal of Educational Research, 95, 308–318.


Fan, X., & Chen, M. (2001). Parental involvement and students’ academic achievement: A meta-analysis. Educational Psychology Review, 13(1), 1–22.


Fine, M. (1994). Chartering urban school reform. In M. Fine (Ed.), Chartering urban school reform: Reflections on public high schools in the midst of change (pp. 5–30). New York, NY: Teachers College Press.


Finn, J. D. (1989). Withdrawing from school. Review of Educational Research, 59(2), 117–142. http://doi.org/10.3102/00346543059002117


Finn, J. D. (1993). School engagement and students at risk. Washington, DC: U.S. Department of Education, National Center for Education Statistics.


Gershenson, S., Jacknowitz, A., & Brannegan, A. (2016). Are student absences worth the worry in U.S. primary schools? Education Finance and Policy. Online Early Access.


Gottfried, M. A. (2010). Evaluating the relationship between student attendance and achievement in urban elementary and middle schools: An instrumental variables approach. American Educational Research Journal, 47(2), 434–465.


Gottfried, M. A. (2011b). The detrimental effects of missing school: Evidence from urban siblings. American Journal of Education, 117(2), 147–182.


Gresham, F. M., & Elliot, S. M. (1990). Social skills rating system. Circle Pines, MN: American Guidance Service.


Hallfors, D., Vevea, J. L., Iritani, B., Cho, H., Khatapoush, S., & Saxe, L. (2002). Truancy, grade point average, and sexual activity: a meta-analysis of risk indicators for youth substance use. Journal of School Health, 72(5), 205–11.


Harte, A. J. (1994). Improving school attendance: Responsibility and challenge. Toronto, Canada: Canadian Education Association.


Jackson, S. L., Vann, W. F., Kotch, J. B., Pahel, B. T., & Lee, J. Y. (2011). Impact of poor oral health on children’s school attendance and performance. American Journal of Public Health101, 1900–1906.


Jeynes, W. H. (2003). A meta-analysis: The effects of parental involvement on minority children’s academic achievement. Education and Urban Society, 35(2), 202–218. http://doi.org/10.1177/0013124502239392


Johnson, G. M. (2005). Student alienation, academic achievement, and WebCT use. Education Technology and Society, 8(2), 179–189.


Juel, C. (1988). Learning to read and write: A longitudinal study of 54 children from first through fourth grades. Journal of Educational Psychology, 80(4), 437–447.


Kane, J. (2006). School exclusions and masculine, working-class identities. Gender and Education, 18(6), 673–685.


Kearney, C. (2008). An interdisciplinary model of school absenteeism in youth to inform professional practice and public policy. Educational Psychology Review. Retrieved from http://link.springer.com/article/10.1007/s10648-008-9078-3


Krenitsky-Korn, S. (2011). High school students with asthma: attitudes about school health, absenteeism, and its impact on academic achievement. Journal of Pediatric Nursing, 37, 61–68.


Lehr, C. A., Hansen, A., Sinclair, M. F., &  Christenson, S. L. (2003). Moving beyond dropout towards school completion: An integrative review of data-based interventions. School Psychology Review, 32(3), 342–364.


Lehr, C. A., Sinclair, M. F., & Christenson, S. L. (2004). Addressing student engagement and truancy prevention during the elementary school years: A replication study of the check & connect model. Journal of Education for Students Placed at Risk (JESPAR), 9(3), 279–301. http://doi.org/10.1207/s15327671espr0903_4


Marvul, J. N. (2012). If you build it, they will come: A successful truancy intervention program in a small high school. Urban Education, 47(1), 144–169. http://doi.org/10.1177/0042085911427738


McNeal, R. B., Jr. (1999). Parental involvement as social capital: Differential effectiveness on science achievement, truancy, and dropping out. Social Forces, 78(1), 117–144. http://doi.org/10.2307/3005792


Muller, C. (1993). Parental involvement and academic achievement: An analysis of family resources available to the child. In B. Schneider & J. S. Coleman (Eds.), Parents, their children, and schools (pp. 77–114). Boulder, CO: Westview Press.


Mulligan, G. M., Hastedt, S., & McCarroll, J. C. (2012, July). First-time kindergartners in 2010–11: First findings from the kindergarten rounds of the Early Childhood Longitudinal Study, kindergarten class of 2010–11 (ECLS-K:2011) (NCES 2012-049). Washington, DC: U.S. Department of Education, National Center for Education Statistics.


Nauer, K., Mader, N., Robinson, G., & Jacobs, T. (2014). A better picture of poverty: What chronic absenteeism and risk load reveal about NYC’s lowest income elementary schools. New York, NY: Center for New York City Affairs.


Newmann, F. M. (1981). Reducing student alienation in high schools: Implications of theory. Harvard Educational Review, 51(4), 546–564.


Olson, S. L., Sameroff, A. J., Kerr, D. C. R., Lopez, N. L., & Wellman, H. M. (2005). Developmental foundations of externalizing problems in young children: The role of effortful control. Development and Psychopathology, 17(1), 25–45.


Orfield, G., & Kornhaber, M. L. (Eds.). (2001). Raising standards or raising barriers? Inequality and high-stakes testing in public education. New York, NY: Century Foundation Press.


Pianta, R. C., & Walsh, D. J. (1996). High-risk children in schools: Constructing sustaining relationships. New York, NY: Routledge.


Posner, M. I., & Rothbart, M. K. (2000). Developing mechanisms of self-regulation. Development and Psychopathology, 12(3), 427–441.


Pourat, N., & Nicholson, G. (2009, November). Unaffordable dental care is linked to frequent school absences. Los Angeles, CA: UCLA Center for Health Policy Research.


Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods. Thousand Oaks, CA: Sage Publications.


Ready, D. D. (2010). Socioeconomic disadvantage, school attendance, and early cognitive development: The differential effects of school exposure. Sociology of Education, 83(4), 271–286. http://doi.org/10.1177/0038040710383520


Reid, K. (1982). The self-concept and persistent school absenteeism. British Journal of Education Psychology, 52(2), 179–187.


Reid, K. (1983). Institutional factors and persistent school absenteeism. Educational Management Administration & Leadership, 11(1), 17–27.


Romero, M., & Lee, Y.-S. (2007). A national portrait of chronic absenteeism in the early grades. New York.


Rothman, S. (2001). School absence and student background factors : A multilevel analysis. International Education Journal, 2(1), 59–68.


Rumberger, R. W. (1995). Dropping out of middle school: A multilevel analysis of students and schools. American Educational Research Journal, 32(3), 583–625.


Sampson, R. J., & Laub, J. H. (1994). Urban poverty and the family context of delinquency: A new look at structure and process in a classic study. Child Development, 65(2), 523. http://doi.org/10.2307/1131400


Sheldon, S. B. (2007). Improving student attendance with school, family, and community partnerships. Journal of Educational Research, 100(5), 267–275. http://doi.org/10.3200/JOER.100.5.267-275


Sheldon, S. B., & Epstein, J. L. (2004). Getting students to school: Using family and community involvement to reduce chronic absenteeism. School Community Journal, 14(2), 39–56. Retrieved from http://eric.ed.gov/?id=EJ794822


Smith, S. S. (1997). A longitudinal study: The literacy development of 57 children. In C. K. Kinzer, K. A. Hinchman, & D. J. Leu (Eds.), Inquiries in literacy theory and practice: Forty-sixth yearbook of the National Reading Conference (pp. ). Chicago, IL: National Reading Conference.


Snijders, T., & Bosker, R. (2012). Multilevel analysis: An introduction to basic and advanced multilevel modeling (2nd ed.). Los Angeles: SAGE.


Southworth, P. (1992). Psychological and social characteristics associated with persistent absence among secondary aged school children with special reference to different categories of persistent absence. Personality and Individual Differences, 13(3), 367–376. http://doi.org/10.1016/0191-8869(92)90116-7


StataCorp. (2013). Stata statistical software: Release 13. College Station, TX: StataCorp LP.


Stouthamer-Loeber, M., & Loeber, R. (1988). The use of prediction data in understanding delinquency. Behavioral Sciences & the Law, 6(3), 333–354. http://doi.org/10.1002/bsl.2370060305


Tudge, J., Odero, D., Hogan, D., & Etz, K. (2003). Relations between the everyday activities of preschoolers and their teachers’ perceptions of their competence in the first years of school. Early Childhood Research Quarterly, 18(1), 42–64.


Votruba-Drzal, E., Li-Grining, C. P., & Maldonado-Carreño, C. (2008). A developmental perspective on full- versus part-day kindergarten and children’s academic trajectories through fifth grade. Child

Development, 79(4), 957–978.


Weiss, H. (2005). Preparing educators to involve families: From theory to practice. Thousand Oaks, CA: Sage Publications.


Zhang, S. (2012). Do our children become healthier and wiser? A study of the effect of medicaid coverage on school absenteeism. International Journal of Health Services, 42(4), 627–646.




Cite This Article as: Teachers College Record Volume 119 Number 7, 2017, p. 1-34
https://www.tcrecord.org ID Number: 21802, Date Accessed: 11/26/2021 6:19:37 PM

Purchase Reprint Rights for this article or review