February 2021: Improved Educational Achievement as a Path to Desistance

Grigorenko, E. L., Hart, L., Hein, S., Kovalenko, J., & Naumova, O. Y. (2019). Improved educational achievement as a path to desistance. New Directions for Child and Adolescent Development, 165, 1–25.

Summary by Julia D. Kovalenko, M.A., and Elena L. Grigorenko, Ph.D.


Juvenile delinquency is a serious concern in the United States, which has the largest juvenile justice (JJ) system in the world. Among JJ-involved youth, one of the most commonly observed diagnoses are learning disabilities (LD). The aim of the article published by Grigorenko and colleagues (2019) was twofold: (1) Review the literature on the connection between learning difficulties/disabilities and delinquency, and (2) provide examples of such connections with data available through the Houston LD Hub. The Houston LD Hub is a member of the LD Research Centers (including Texas Center for Learning Disabilities) and Hub network that focuses on JJ youth.

The research literature contains four main lines of research on the association between learning difficulties/disabilities and delinquency. The first two research lines include studies on the general population: (1) examining academic difficulties as a risk factor for delinquent behavior and (2) focusing on the diagnosis or presence of LD as a risk for entry into the JJ system. The other two lines focus specifically on JJ-involved youth: (3) the presentation of academic difficulties among JJ-involved youth and (4) the prevalence of LDs among these youth.

Academic Difficulties and LD as Risk Factors for Delinquency in the General Population

Reviews of the literature, with some exceptions, suggest that the risk of being involved in delinquent behavior is substantially greater for youth with LD compared to youth without LD. Poor school performance and school truancy have been linked to delinquent behaviors. In adolescents, academic difficulties are disproportionately represented among those who are suspended, expelled, and/or drop out of high school, which in turn have been associated with delinquency in general and serious offending in particular. The literature indicates that both LD and emotional disturbances (ED) were observed to be significant risk factors for delinquency. A study found that youth diagnosed with ED and LD in 1990 and 2005 had significantly higher arrest rates in the adult justice system, compared with other students with disabilities (Newman, Wagner, Cameto, Knokey, & Shaver, 2010). Some research is less conclusive, with a few studies reporting no association between LD and delinquency (Malmgren, Abbott, & Hawkins, 1999; White & Loeber, 2008).

A number of theories attempt to explain the connection between delinquent behavior and LD. The school failure hypothesis suggests that the student may experience consistent academic failure due to LD, which leads to a negative self-image and marginalization from school. The continuous experience of negativity leads to delinquent behaviors. The susceptibility theory indicates that personality and predispositions in youth with LD make them susceptible to situations and opportunities that increase the risk for delinquency. These two theories are considered too simplistic because they do not account for the complex path to delinquency. The third hypothesis takes a developmental approach to attempt to overcome these limitations by considering (1) early difficulties in language development as the basis for challenges with school readiness and reading acquisition; and (2) the manifestation of delinquency as a result of a continuously unfolding accumulation of disruptive behavior, peer rejection, and school suspension/expulsion.

Academic Difficulties and LD Among JJ-Involved Youth

As a whole, JJ-involved youth are characterized by lower academic performance and levels of competencies in all academic domains, compared to the general population of their peers. It has been observed that nearly 70% of incarcerated youth suffer from at least one diagnosable disability, although many of them are underdiagnosed. The prevalence of youth with disabilities as defined by IDEIA, Individuals with Disabilities Education Improvement Act (20 U.S. code chapter 33 §1401 (30)), in the JJ system has been estimated to range from 28% to 46%, which is significantly higher than the general population estimates, ranging from 8% to 12%. The overrepresentation of LD among JJ-involved youth is consistent with earlier research indicating that youth with LD have a lower age of first offense and are 220% more likely to be adjudicated than adolescents in the general population.

Deficient academic skills may also lead to difficulties in navigating the system at all stages of the JJ process, as well as in transitioning successfully back into the community without recidivating. JJ-involved youth with a special education diagnosis upon discharge from a residential facility are 2.8 times more likely to be placed in a residential facility again and 2.2 times less likely to be engaged in either employment or enrolled in school, compared to a sample of their peers without disabilities.

In summary, there is research connecting learning difficulties/disabilities to delinquency both in the general population and among JJ-involved youth. However, there is a substantial amount of variation in and shortage of (1) estimates of the strength and direction of this association; (2) understanding of the causal mechanisms of this association; and (3) policies for accommodating youth with LD in the JJ system and upon their community re-entry. The program of research underway by the Houston LD Hub seeks to address these shortcomings of the literature.

Learning Difficulties, Learning Disabilities, and Delinquency: Data from the Houston LD Hub

The Houston LD Hub was funded in 2017 by the U.S. National Institutes of Health and has working relationships with faculty at Baylor College of Medicine and University of Houston, practitioners at the Connecticut Court Support Services Division (CSSD), and the Harris County Juvenile Probation Department (HCJPD) in Texas. Studies conducted in Connecticut formed the foundation for the ongoing work now unfolding primarily in Texas.

The CSSD requested an analysis of juvenile detention centers and community residential centers in the state of Connecticut to: (1) describe the educational characteristics of detained children and youth; (2) describe the educational programs currently used in detention and assess their consistency with the framework of the State of Connecticut Department of Education; (3) characterize the community of teachers working with students in detention; (4) identify systemic obstacles and/or challenges to educating this population; (5) determine the pathways of educational records of detained children and youth; and (6) identify system barriers or challenges to delivering education to this population and teaching in detention settings. The analyses conducted upon this request initiated the research group’s interest in LD in JJ-involved youth.

Primary Findings

Nearly 70% of all teachers reported that juvenile detainees’ special learning needs are “underidentified.” Screening records supported teacher reports, with only 70 students of 713 screened (i.e., 9.9%) being identified as having special educational needs. Of these 70, only 8 (i.e., 11.4%) had an Individualized Education Program (IEP) or notes referring to an IEP in their files.

To follow up on those findings, the researchers systematically and randomly sampled 20% of youth admitted to juvenile detention centers in Connecticut in a year (Grigorenko, Edwards, & Chapman, 2015; Grigorenko, Sullivan, & Chapman, 2015). Of the sample of 371 youth, 96.8% had a mental health diagnosis, 7.2% had an LD diagnosis, and 0.8% had an LD diagnosis as their primary diagnosis.

The documented rates of LD were lower than expected based on the literature. Thus, the team developed an educational placement screener to assess levels of competence in reading and mathematics for juveniles in short-term juvenile detention facilities (Hart et al., 2012). Youth’s average level of performance on the screener was around grade 5, which is substantially below expectations based on their grade placements in their home schools. In some cases, the discrepancy in performance was as large as eight grades.

Later, the screener was computerized and administered to a sample of 1,337 youth detained in Connecticut (Grigorenko et al., 2015). A subsample (n = 410) of these youth received the Wide Range Achievement Test in addition to the educational screener. The results of both assessments suggest the prevalence of LD ranged from 13% to 40%, averaging 24.9%. Youth’s grade equivalence indicators were varied: 3.51 to 4.27 for word-level reading, 6.01 to 6.22 for text-level reading, and 5.24 to 5.64 for mathematics.

Subsequent to this work, the research group merged information from the CSSD and the Connecticut State Department of Education into an administrative dataset (Hein et al., 2017). These data contained records from seven academic years for 58,678 JJ-involved youth. The number of IDEIA diagnoses of LD in this sample ranged from 4.1% to 6.4%, depending on the year. However, upon examination of standardized achievement data on reading and math achievement, a much higher number of youths performed below expected levels.

Inconsistencies between established diagnoses and the actual level of performance on standardized tests led the team to conduct a study with one of the largest JJ systems in the country, the HCJPD. The new project includes a clinical trial and big (administrative) data. The clinical trial, Strategies for Enhancing Reading in Older Underperforming Students (SERIOUS), was in its second year at the time of the article’s publication, so no data were available to report. The program includes twenty-four 90-minute lessons that teach foundational reading skills and “learning how to learn” skills. Upon the completion of the intervention, youth are provided a cell phone with a series of word-based computer games to build automaticity with the skills learned. Progress is evaluated throughout the intervention. Furthermore, neurophysiological and molecular biomarkers are collected to gain insight into the etiology of the learning difficulties/disabilities and delinquency association.

An additional source of information is the administrative data for all 79,119 JJ-involved youth in the county from 2006 through 2018. A smaller sample included data specifically on youth with mental health diagnoses (n = 3,566). Based on the first type of data, the frequency of LD across the two samples was .01% and .03%, respectively. In the mental health sample, the frequency of LD diagnoses was 9%.

Learning Difficulties, Learning Disabilities, and Delinquency: Future Directions

Juvenile delinquency represents a serious societal problem. If not treated while placed, youth recidivate at the rate of 60–80%, but when treated, the average reduction of recidivism is about 9%. However, there are numerous obstacles to the selection and delivery of evidence-based treatments such as educational therapies in the JJ system. First, there are difficulties with accurate assessment of the needs of these youth, such as lack of routine assessments, no guidelines on which assessment instruments to use, and unclear and inconsistent definitions of LD. Second, JJ-involved youth have been reported to be resistant to intervention and have poor engagement, either due to limited motivation or lack of awareness/denial of their difficulties. Third, these youth have a lack of access to educational programs in general and special education programs in particular. Fourth, although academic achievement has been described as a key variable in successful reintegration, protective, wrap-around rehabilitative services often do not emphasize the importance of educational achievement. Fifth, there is a lack of trained JJ-system practitioners who can work with youth with special needs in general and LD in particular.

To summarize, there are many obstacles, including a shortage of research on the prevalence, typology, and etiology of LD among JJ-involved youth that led the research group to conduct the interventional research project within the Houston LD Hub.

Recommendations for Educators and Administrators in JJ Settings

Screen all youth entering the JJ system using reliable assessment tools to identify youth with LD when they enter the system.
Develop flexible instructional plans to best support and optimize the learning of JJ-involved students with LD. Because youth may not be at the grade level they are supposed to be, providing them with materials at their current level of functioning will help them catch up on what they have missed.
Provide information on the types of services that youth can request from schools (e.g. 504 plan or IEP) so that they are aware of how and where to seek services upon reintegration into society.
Communicate with the youth’s home school when possible to exchange information on the youth’s education, learning concerns, and progress. In particular, this is important when the youth is first placed into detention facilities and upon release.

Implications for Educators and Administrators in non-JJ settings

The research suggests that there is an association between LD and ED with JJ-involvement. Early identification of LD and ED is key to helping students avoid JJ-involvement.

It is important to identify and address both the academic and emotional-behavioral needs of students with disabilities. Providing underserved or underresourced families with information on how to seek services for the youth’s problems may be a preventative measure for future JJ-involvement.


Grigorenko, E. L., Edwards, L., & Chapman, J. (2015). Cannabis use among juvenile detainees: Typology, frequency and association. Criminal Behaviour and Mental Health, 25, 54–65. Retrieved from https://doi.org/10.1002/cbm.1913

Grigorenko, E. L., Macomber, M., Hart, L., Naples, A., Chapman, J. C., Foley-Geib, C.,…Wagner, R. (2015). Academic achievement among juvenile detainees. Journal of Learning Disabilities, 48, 359–368. https://doi.org/10.1177/0022219413500991

Grigorenko, E. L., Sullivan, T., & Chapman, J. (2015). An investigation of gender dif- ferences in a representative sample of juveniles detained in Connecticut. International Journal of Law and Psychiatry, 38, 84–91. Retrieved from https://doi.org/10.1016/j.ijlp. 2015.01.011

Hart, L., Naples, A., Chapman, J., Chart, H., D’Amaddio, A., Foley-Geib, C., . . . Grigorenko, E. L. (2012). The development and evaluation of the educational placement screener for youth in detention. European Journal of Psychological Assessment, 28, 321– 328.

Hein, S., Barbot, B., Square, A., Chapman, J., Geib, C. F., & Grigorenko, E. L. (2017). Violent offending among juveniles: A 7-year longitudinal study of recidivism, desistance, and associations with mental health. Law and Human Behavior, 41, 273–283. https://doi.org/10.1037/lhb0000241

Malmgren, K., Abbott, R. D., & Hawkins, J. D. (1999). LD and delinquency: Rethinking the “link”. Journal of Learning Disabilities, 32, 194–200.

Newman, L.,Wagner,M., Cameto, R., Knokey, A. M., & Shaver, D. (2010). Comparisons across time of the outcomes of youth with disabilities up to 4 years after high school. A report of findings from the National Longitudinal Transition Study-2 (NLTS2). www.nlts2.org/reports/2010_09/nlts2_report_2010_09_complete.pdf

White, N. A., & Loeber, R. (2008). Bullying and special education as predictors of serious delinquency. Journal of Research in Crime and Delinquency, 45, 380–397.