View in EDS HTML Full Text PDF Full Text

Universal Behavior Screening and Early Warning System Indicators in Middle Schools

Saved in:
Bibliographic Details
Title: Universal Behavior Screening and Early Warning System Indicators in Middle Schools
Language: English
Authors: Emily Graybill (ORCID 0000-0002-6912-156X), Scott Lewis, Ella Anghel (ORCID 0000-0001-6332-7826), Sofia Awan, Brian Barger, Ashley Salmon
Source: Psychology in the Schools. 2025 62(9):2955-2968.
Availability: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
Peer Reviewed: Y
Page Count: 14
Publication Date: 2025
Sponsoring Agency: Substance Abuse and Mental Health Services Administration (SAMHSA) (DHHS/PHS)
Contract Number: SM083659
Document Type: Journal Articles
Reports - Research
Education Level: Junior High Schools
Middle Schools
Secondary Education
Descriptors: Screening Tests, Student Behavior, Early Intervention, Middle School Students, At Risk Students, Potential Dropouts, Underachievement, Psychological Patterns, Symptoms (Individual Disorders), Behavior Problems, Emotional Problems, Racial Differences, Gender Differences, Attendance Patterns, Predictor Variables
DOI: 10.1002/pits.23515
ISSN: 0033-3085
1520-6807
Abstract: Early warning indicator and intervention systems (EWS) have been promoted to identify students at risk of school underperformance or dropout. Current EWS systems typically include administrative data on attendance, behavior incidents requiring disciplinary action, and course performance. This study tested whether specific emotional and behavioral risk symptoms measured by a student self-report universal screener administered in the fall can predict the three EWS indicators after controlling for fall behavioral incidents and whether they account for some of the variance in EWS attributed to demographic characteristics. Using data from 3307 middle school students, we found that after accounting for fall disciplinary issues, conduct problems predicted poorer student outcomes, but hyperactivity/inattention was predictive of better course performance. Peer problems predicted lower performance in some courses, while emotional problems predicted better performance as well as fewer behavioral issues. We also found that these symptoms accounted for some of the variance in EWS attributed to race and gender. The results suggest that student self-report universal screening can complement existing EWS measures and potentially identify at-risk students who may not otherwise be identified through traditional EWS indicators.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1479953
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwEQFFAsAeJ44JvyonNyxocPAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDA1rAR20uFDu-0uGiwIBEICBm2S3lcRqjwt8i_5_TEcEE0uEAgXtm1GcRZANuMvFmSmFLMIL9MGQ8D4ovcded-FaMHB4QqsHIH8MzWmxO06ELxwigQrVrGJrrtdKQkJtOa5oE3NptrAPQg70nfLlRmg-CXC0MaTaybq46e1z3k6hvkPq5yAVXSVQpA6fBFoOB1ezxHCEZ0CWq4cYdS0o_oeUxkoo52NlrSFm7QH3
Text:
  Availability: 1
  Value: <anid>AN0187257390;pis01sep.25;2025Aug14.01:14;v2.2.500</anid> <title id="AN0187257390-1">Universal Behavior Screening and Early Warning System Indicators in Middle Schools </title> <p>Early warning indicator and intervention systems (EWS) have been promoted to identify students at risk of school underperformance or dropout. Current EWS systems typically include administrative data on attendance, behavior incidents requiring disciplinary action, and course performance. This study tested whether specific emotional and behavioral risk symptoms measured by a student self‐report universal screener administered in the fall can predict the three EWS indicators after controlling for fall behavioral incidents and whether they account for some of the variance in EWS attributed to demographic characteristics. Using data from 3307 middle school students, we found that after accounting for fall disciplinary issues, conduct problems predicted poorer student outcomes, but hyperactivity/inattention was predictive of better course performance. Peer problems predicted lower performance in some courses, while emotional problems predicted better performance as well as fewer behavioral issues. We also found that these symptoms accounted for some of the variance in EWS attributed to race and gender. The results suggest that student self‐report universal screening can complement existing EWS measures and potentially identify at‐risk students who may not otherwise be identified through traditional EWS indicators.</p> <p>Summary: School district early warning system indicators often do not include a measure of emotional and behavior risk or a student self‐report measure of early risk.Among middle schoolers, students with internalizing concerns such as peer and emotional problems may not be at risk on other early warning system indicators – such as attendance, behavior, and course performance ‐ and would likely be missed as needing early intervention without behavior screening data.Student self‐report universal behavior screening measuring internalizing and externalizing concerns can complement existing early warning systems to identify at‐risk students early, particularly those with internalizing concerns who may not be at risk on other early warning system indicators.</p> <p>Keywords: early warning systems; mental health; middle school; universal screening</p> <hd id="AN0187257390-2">Introduction</hd> <p>School dropout is a public health issue with community‐ and individual‐level risk and protective factors (Lansford et al. [<reflink idref="bib33" id="ref1">33</reflink>]; Shin et al. [<reflink idref="bib51" id="ref2">51</reflink>]). Over 2 million high school students in the United States (U.S.) drop out of school each year. This equates to over 1000 students per school day (National Center for Education Statistics [NCES] [<reflink idref="bib43" id="ref3">43</reflink>]). These data underscore the urgent need for schools to build large data systems within a public health framework that facilitate dropout prevention activities and early identification and intervention for students at risk of dropping out (Freudenberg and Ruglis [<reflink idref="bib16" id="ref4">16</reflink>]).</p> <p>Increasingly, schools use their vast sets of administrative data to identify students at risk for dropout and implement early intervention to mitigate that risk (Neild et al. [<reflink idref="bib44" id="ref5">44</reflink>]; Regional Educational Laboratory Pacific [<reflink idref="bib47" id="ref6">47</reflink>]). Early Warning Indicator and Intervention Systems (EWS) are data systems used to identify students at risk of dropout and low school engagement (Balfanz and Byrnes [<reflink idref="bib3" id="ref7">3</reflink>]). They grew from the idea that students show early signs of disengagement with school and that early intervention by educators may prevent dropout (Bruce et al. [<reflink idref="bib7" id="ref8">7</reflink>]; Neild et al. [<reflink idref="bib44" id="ref9">44</reflink>]). In a sign of their effectiveness, the U.S. Department of Education has recognized the use of EWSs to identify students at an increased risk of dropping out of school (U.S. Department of Education [<reflink idref="bib53" id="ref10">53</reflink>]).</p> <p>Most commonly, EWSs use schools' administrative data, specifically attendance, behavioral issues, and course performance, also known as the "ABCs", to identify at‐risk students (Bruce et al. [<reflink idref="bib7" id="ref11">7</reflink>]). These three ABC domains are widely used because they are highly predictive of high school completion (Neild et al. [<reflink idref="bib44" id="ref12">44</reflink>]; Pinkus [<reflink idref="bib46" id="ref13">46</reflink>]; Wentworth and Nagaoka [<reflink idref="bib56" id="ref14">56</reflink>]). However, relying solely on ABCs to identify at‐risk students has several disadvantages. First, the data are only available several months into the school year at the earliest, which delays the start of interventions (de Vasconcelos et al. [<reflink idref="bib54" id="ref15">54</reflink>]) and promotes a "wait‐to‐fail" model, flagging students who are already struggling (Cavanaugh [<reflink idref="bib10" id="ref16">10</reflink>]). While ABC indicators from the prior spring can be used to inform interventions in the following year, information from the previous year might not reflect a student's current difficulties, and it still allows schools to intervene only among students who have struggled for long periods of time.</p> <p>Second, traditional EWS indicators do not include student self‐report data. Most middle and high school students can recognize their own disengagement (Archambault et al. [<reflink idref="bib1" id="ref17">1</reflink>]) even if it is not observable by teachers or administrative data (Lovelace et al. [<reflink idref="bib37" id="ref18">37</reflink>]). Indeed, some research has explored the relationship between student self‐report screening data and student outcomes, suggesting that student self‐report data add unique information to the identification of student risk (Margherio et al. [<reflink idref="bib38" id="ref19">38</reflink>]).</p> <p>Finally, consistent findings point to the association between mental health and dropout risk (Hjorth et al. [<reflink idref="bib24" id="ref20">24</reflink>]). Cross‐sectional and longitudinal research shows that untreated mental health concerns are one risk factor for leaving school (Butterworth and Leach [<reflink idref="bib8" id="ref21">8</reflink>]; Hjorth et al. [<reflink idref="bib24" id="ref22">24</reflink>]; Lawrence and Adebowale [<reflink idref="bib34" id="ref23">34</reflink>]). Conversely, both experimental and associational studies show that students are less likely to drop out if they receive needed mental health support and have positive mental wellness (Heller et al. [<reflink idref="bib23" id="ref24">23</reflink>]; Lawrence and Adebowale [<reflink idref="bib34" id="ref25">34</reflink>]), and students who had recovered from earlier mental health concerns are at a lower risk of dropping out. These findings suggest that by intervening early with youth with such concerns, educators may be able to prevent school dropout (Rodriguez and Conchas [<reflink idref="bib49" id="ref26">49</reflink>]), highlighting the importance of early identification of and intervention for students' mental health concerns. Yet, the typical EWS in schools only includes observed behavioral problems like discipline referrals instead of emotional risk indicators such as universal behavioral and emotional risk screening (hereafter referred to as universal behavior screening). There is also minimal empirical literature on the inclusion of an emotional and behavior risk data set within districts' EWS to guide schools on using such data to inform prevention and early intervention activities (Davis et al. [<reflink idref="bib13" id="ref27">13</reflink>]).</p> <hd id="AN0187257390-3">Universal Behavior Screening and Early Warning Systems</hd> <p>Universal behavior screeners can address many of the issues that exist with traditional EWSs. Screeners can be easily administered early in the school year, allowing for earlier identification and intervention. They reveal students' perspectives on their own emotional and behavioral status (when using a student self‐report screener), complementing the external perspective of administrative data. Moreover, they allow for the detection of students with internalizing concerns who are usually under‐identified using traditional EWSs despite their heightened risk (e.g., Graybill et al. [<reflink idref="bib21" id="ref28">21</reflink>]). School‐based screeners often consist of externalizing and internalizing subscales, and some screeners further expand on these general subscales by including subscales or items targeting specific symptoms such as peer rejection, hyperactivity, and academic behavior (e.g., Drummond [<reflink idref="bib14" id="ref29">14</reflink>]; Izumi and Eklund [<reflink idref="bib25" id="ref30">25</reflink>]).</p> <p>In recent years, more research has examined how universal screeners relate to EWS indicators. These studies have shown that emotional and behavioral risk screeners can predict office discipline referrals (ODRs; Gregory et al. [<reflink idref="bib22" id="ref31">22</reflink>]; Graybill et al. [<reflink idref="bib21" id="ref32">21</reflink>]), attendance (Jones et al. [<reflink idref="bib27" id="ref33">27</reflink>]), and academic attainment (Eklund et al. [<reflink idref="bib15" id="ref34">15</reflink>]). However, existing studies have rarely examined all three ABC indicators together alongside student self‐report screening data. Since currently, EWS are used together in schools to identify at‐risk students, it is important to also test whether screeners can predict all three of them together.</p> <p>In addition, these studies usually focus on internalizing and externalizing scores, even though many screeners offer subscales targeting more specific issues. Investigating how specific subscales relate to the ABC data can help practitioners understand the complex relationship between risk indicators and focus on the most informative subscales data to include in their EWSs. For example, Young et al.'s (2021) work examined specific screener scores (aggression, peer rejection, flat emotion, etc.) and their association with attendance, academic performance, and in‐school and out‐of‐school suspensions. They found that even within the externalizing and internalizing groups of symptoms, there were specific subscales whose relationships with ABCs diverged in direction and strength. However, Young et al. ([<reflink idref="bib57" id="ref35">57</reflink>]) only used a teacher‐report screener. Exploring this effect with a student‐report screener is warranted, as it may offer unique benefits with less burden on teachers.</p> <hd id="AN0187257390-4">Use of Office Discipline Referrals as Screeners</hd> <p>Schools typically use their ODR data as their behavior metric in their EWS. Some studies show that ODRs accrued by a student in early fall predict spring ODRs and other end‐of‐year outcomes (McIntosh et al. [<reflink idref="bib41" id="ref36">41</reflink>]). However, the use of Fall ODR data still means that interventions can be applied in November at the earliest, missing out on months of potential preventative work with at‐risk students. Therefore, it is important to test the advantage of universal screening relative to Fall ODR count in predicting student outcomes.</p> <hd id="AN0187257390-5">Disparities in Dropout Rates and EWS</hd> <p>Dropout rates vary by race, ethnicity, and gender, with American Indian/Alaska Native, Hispanic, and Black youth experiencing the highest rates of dropout (National Center for Educational Statistics [NCES] [<reflink idref="bib42" id="ref37">42</reflink>]). Across most races, males are more likely to drop out than females (National Center for Educational Statistics [<reflink idref="bib42" id="ref38">42</reflink>]). The same differences in racial and gender data exist with the ABC indicators in EWS. Black students and males are more likely to receive discipline referrals (Liu et al. [<reflink idref="bib36" id="ref39">36</reflink>]) and have lower achievement scores (Assari et al. [<reflink idref="bib2" id="ref40">2</reflink>]). These disparities and the underlying racial bias justify the need to understand how universal screening and ABC data trends vary by race and gender (Chin et al. [<reflink idref="bib11" id="ref41">11</reflink>]). Universal screening data are collected systematically and used to inform a preventative and proactive data‐based decision‐making process for all students. Data from a universal screening process may not include the inconsistencies and bias inherent in the assignment of ODRs as a discipline strategy or in the review of ODRs to make decisions about individual students (Tara C. Raines et al. [<reflink idref="bib52" id="ref42">52</reflink>]; Barger et al. [<reflink idref="bib5" id="ref43">5</reflink>]).</p> <hd id="AN0187257390-6">Current Study</hd> <p>In the current study, we explore the contribution of a student self‐report universal behavior screener within an EWS to better understand the value added by including early mental health data in an EWS. More specifically, we examine the relationship between universal behavior screening subscales with the original EWS indicators, attendance, behavior, and course performance in a middle school setting. We also extend existing research by comparing the screeners' subscales to Fall ODRs and as predictors of end‐of‐year ABCs and by testing whether the universal screening accounts for some of the gender and race disparities in ABCs.</p> <p>To do so, we analyzed data from students in six middle schools who completed the self‐report version of the Strengths and Difficulties Questionnaire (SDQ; Goodman [<reflink idref="bib19" id="ref44">19</reflink>]), focusing on the SDQ subscales of emotional problems, peer problems, conduct problems, and hyperactivity‐inattention. While many existing studies focus on high school students' data, research on EWS suggests that risk indicators of middle schoolers can predict later dropout risk (Balfanz et al. [<reflink idref="bib4" id="ref45">4</reflink>]). As we were mostly interested in early risk detection, we focused on middle schoolers. This study may inform future efforts to enrich EWSs to support data‐based decision‐making within districts' public health tiered system of support framework to facilitate the early identification of and intervention for students at risk of dropout.</p> <p>Our research questions are:</p> <hd id="AN0187257390-7">RQ1</hd> <p>Do the universal screener subscales predict attendance, behavior, and course performance?</p> <hd id="AN0187257390-8">RQ2</hd> <p>Do the universal screener subscales predict attendance, behavior, and course performance over and above Fall ODRs?</p> <hd id="AN0187257390-9">RQ3</hd> <p>Do the universal screeners capture some variance in EWS indicators attributed to gender or race?</p> <hd id="AN0187257390-10">Methods</hd> <p></p> <hd id="AN0187257390-11">Context</hd> <p>Data for this study were collected as part of a school‐based mental health project in the Southeastern U.S. This is a 5‐year project designed to help schools build infrastructure related to mental health, including strengthening the mechanisms of identifying at‐risk students and referring them to further support. The schools participating in this project administered the SDQ in September 2021, at the beginning of the 2021–2022 school year, the first year of the project. The screening data were then utilized to inform school‐level decision‐making within a multi‐tiered system of support framework. The schools reviewed school‐level trends in the screening data to inform their implementation of schoolwide programs. Individual‐level student risk was also reviewed to inform decisions about targeted and individualized interventions. Because these data were collected in Year 1 of a multi‐year infrastructure‐building project, it is not expected that the implementation of the project influenced trends in the student data reported in this article. Students' guardians provided consent for them to participate in the universal behavior screening process and this study was approved by Rutgers University Institutional Review Board.</p> <p>In this study, we included data from 3037 students within six middle schools (grades 6th through 8th) who had screening and administrative data available. Approximately 47% of respondents were female, 50% were male, and 3% were either Nonbinary or did not report their gender. Most students were Black or African American (76%), representing the demographics of the school district. The remaining 24% of students were grouped as "Other Race." Black students were used as the reference group in the models exploring differences by race due to Black students comprising the majority of the sample and also being of high interest to the target school district (Johfre and Freese [<reflink idref="bib26" id="ref46">26</reflink>]). All students in the sample were eligible for free‐ or reduced‐price lunch. The students' demographics are also presented in Table 1.</p> <p>1 Table Sample's demographics.</p> <p> <ephtml> <table><thead valign="bottom"><tr valign="bottom"><th /><th /><th>Frequency</th><th>Percentage</th></tr></thead><tbody valign="top"><tr><td>Gender</td><td>Female</td><td>1437</td><td>47</td></tr><tr><td /><td>Male</td><td>1511</td><td>50</td></tr><tr><td /><td>Nonbinary</td><td>30</td><td>1</td></tr><tr><td /><td>Prefer not to say</td><td>57</td><td>2</td></tr><tr><td /><td>Unknown</td><td>2</td><td>0.1</td></tr><tr><td>Race</td><td>American Indian or Alaska Native</td><td>45</td><td>2</td></tr><tr><td /><td>Asian</td><td>25</td><td>1</td></tr><tr><td /><td>Black or African American</td><td>2340</td><td>77</td></tr><tr><td /><td>Hispanic</td><td>117</td><td>4</td></tr><tr><td /><td>Multi‐racial</td><td>182</td><td>6</td></tr><tr><td /><td>Native Hawaiian or Other Pacific Islander</td><td>9</td><td>0.3</td></tr><tr><td /><td>Other/Unknown</td><td>21</td><td>1</td></tr><tr><td /><td>White</td><td>298</td><td>10</td></tr></tbody></table> </ephtml> </p> <hd id="AN0187257390-12">Measures</hd> <p></p> <hd id="AN0187257390-13">EWS Data</hd> <p>Data on the students' attendance, behavior, and course performance were included in the analyses. Attendance data consisted of both excused and unexcused absences. As research suggests that more than 20 absences a year significantly reduces graduation rates (Bruce et al. [<reflink idref="bib7" id="ref47">7</reflink>]), we chose to dichotomize this variable such that students with 20 or more absences were coded as one and all others were coded as 0. Before dichotomization, each student had 14.40 absences on average (standard deviation = 13.36). After dichotomization, 6.99% of students had more than 20 absences. Behavior outcome measures consisted of student‐level ODR counts, and they were also dichotomized such that more than two ODRs were coded as 1 and all others were coded as 0, following Bruce et al.'s ([<reflink idref="bib7" id="ref48">7</reflink>]) recommendations based on a summary of the literature and several case studies. The mean number of ODRs before October was 0.48 (standard deviation = 0.86), with 3.82% of students had at least two ODRs before October. Throughout the year, the mean number of ODRs was 2.53 (standard deviation = 2.09), and 17.50% had at least two ODRs. Course performance data included the state‐wide end‐of‐grade test for English language arts (ELA) and math, where performance was categorized by the state‐testing system into four categories, representing beginning (<reflink idref="bib1" id="ref49">1</reflink>), developing (<reflink idref="bib2" id="ref50">2</reflink>), proficient (<reflink idref="bib3" id="ref51">3</reflink>), and distinguished (<reflink idref="bib4" id="ref52">4</reflink>) learners. Table 2 presents the percentages of the achievement levels.</p> <p>2 Table Student's performance.</p> <p> <ephtml> <table><thead valign="bottom"><tr valign="bottom"><th /><th>% Beginning (1)</th><th>% Developing (2)</th><th>% Proficient (3)</th><th>% Distinguished (4)</th></tr></thead><tbody valign="top"><tr><td>ELA Achievement</td><td>37.91</td><td>38.33</td><td>21.00</td><td>2.76</td></tr><tr><td>Math Achievement</td><td>50.97</td><td>34.84</td><td>11.23</td><td>2.96</td></tr></tbody></table> </ephtml> </p> <hd id="AN0187257390-14">Strengths and Difficulties Questionnaire (SDQ)</hd> <p>The SDQ (Goodman [<reflink idref="bib18" id="ref53">18</reflink>]) is commonly used to screen for the behavioral and emotional risk of children aged 2–17. It was selected as the screener in this study due to its strong psychometric properties (Goodman [<reflink idref="bib19" id="ref54">19</reflink>]) and its affordability for the school district. The SDQ consists of 25 questions answered on a 3‐point Likert scale (<emph>Not True, Somewhat True, Certainly True</emph>). These 25 attributes then inform a five‐subscale structure: emotional problems (e.g. 'I worry a lot'), conduct problems ('I usually do as I am told,' reverse coded), hyperactivity‐inattention ('I am restless, I cannot stay still for long'), peer problems ('I would rather be alone than with people of my age'), and prosocial behavior ('I am kind to younger children'). All scales were coded such that higher scores represent higher levels (i.e., more mental health concerns). Table 3 presents descriptive statistics of the four scale scores.</p> <p>3 Table Descriptive statistics for the SDQ subscales.</p> <p> <ephtml> <table><thead valign="bottom"><tr valign="bottom"><th /><th>Mean</th><th>Standard deviation</th><th>Minimum</th><th>Maximum</th><th>Range</th><th>Skewness</th><th>Kurtosis</th></tr></thead><tbody valign="top"><tr><td>Emotional problems</td><td>3.59</td><td>2.49</td><td>0</td><td>10</td><td>10</td><td>0.40</td><td>−0.63</td></tr><tr><td>Conduct problems</td><td>2.28</td><td>1.87</td><td>0</td><td>10</td><td>10</td><td>0.77</td><td>0.11</td></tr><tr><td>Hyperactivity‐inattention</td><td>4.07</td><td>2.30</td><td>0</td><td>10</td><td>10</td><td>0.23</td><td>−0.46</td></tr><tr><td>Peer problems</td><td>3.02</td><td>1.85</td><td>0</td><td>10</td><td>10</td><td>0.53</td><td>0.06</td></tr></tbody></table> </ephtml> </p> <hd id="AN0187257390-15">Data Analysis</hd> <p>Structural equation modeling (SEM) was used to construct our models. The outcomes of these models were absences (dichotomized), ODRs (dichotomized), ELA performance categories, and math performance categories. In the first model (Figure 1), we included twenty manifest variables (items), each loaded on one of four latent variables representing the SDQ subscales (hyperactivity/inattention, conduct problems, emotional problems, and peer problems), five items for each subscale. This model was used to answer RQ1.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/PIS/01sep25/pits23515-fig-0001.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="pits23515-fig-0001.jpg" title="1 Predicting EWS outcomes from the four SDQ subscales (Model 1). (R) indicates a reverse‐coded item." /> </p> <p></p> <p>The second model (Figure 2) was similar, except we added the dichotomized Fall ODRs as a predictor (i.e., all ODRs before October 31st), and the behavior outcomes included all other ODRs. When using ODRs as a predictor and outcome, we used the same categories reported above. This model was used to answer RQ2.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/PIS/01sep25/pits23515-fig-0002.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="pits23515-fig-0002.jpg" title="2 Predicting EWS outcomes from the four SDQ subscales and Fall ODRs (Model 2). (R) indicates a reverse‐coded item." /> </p> <p></p> <p>Finally, to understand trends by demographics, we compared Model 3 (Figure 3), which included only race and gender as predictors of EWS outcomes, to Model 4 (Figure 4), which also includes the SDQ items and subscales. Specifically, we examined the standardized coefficients of race and gender between Model 3 and Model 4 to see if the SDQ subscales account for some of their effect on the EWS outcomes.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/PIS/01sep25/pits23515-fig-0003.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="pits23515-fig-0003.jpg" title="3 Predicting EWS outcomes from race and gender (Model 3)." /> </p> <p></p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/PIS/01sep25/pits23515-fig-0004.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="pits23515-fig-0004.jpg" title="4 Predicting EWS outcomes from race, gender, and the four SDQ subscales (Model 4). (R) indicates a reverse‐coded item." /> </p> <p></p> <p>All SEM analyses were performed using the lavaan package (version 0.6–17) in R version 4.3.2 (Rosseel [<reflink idref="bib50" id="ref55">50</reflink>]). The models were estimated using diagonally weighted least squares, with mean and variance adjustment (WLSMV) to accommodate categorical variables (Graybill et al. [<reflink idref="bib20" id="ref56">20</reflink>]; Barger et al. [<reflink idref="bib5" id="ref57">5</reflink>]; Li [<reflink idref="bib35" id="ref58">35</reflink>]). Following Kline ([<reflink idref="bib29" id="ref59">29</reflink>]), we estimated Models 1, 2, and 4 iteratively revising them by allowing items to covary until an acceptable fit was obtained. Model 3 only included demographics and no items, so this step was not necessary. We evaluated the models' fit to the data using comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR). A model was considered as having an acceptable fit if its CFI > 0.90, RMSEA < 0.05, and SRMR < 0.08 (Byrne [<reflink idref="bib9" id="ref60">9</reflink>]). We also report the models' χ<sups>2</sups>, though as this parameter is sensitive to sample size, we did not use it for fit evaluation (Kline [<reflink idref="bib29" id="ref61">29</reflink>]). Note that Model 3 was just identified, so meaningful fit statistics were not reported for that model.</p> <hd id="AN0187257390-20">Results</hd> <p>A zero‐order correlation table is presented in Appendix A. All models were well‐fitting after minor modifications. In all models, we allowed two pairs of items on the peer problems subscale to covary (solitary‐bullied, good friend‐popular), and for three items in the conduct problems (obedient) and the hyperactivity‐inattention (reflective, persistent) subscales to covary. In Model 4, we also had to allow all predictors to covary (i.e., race, gender, and all subscales).</p> <p>Model 1 was used to answer RQ1 by checking whether the screener subscales predict our ABC indicators. The final fit statistics for this model were χ<sups>2</sups>(<reflink idref="bib223" id="ref62">223</reflink>) = 1839.82, <emph>p</emph> < 0.001, CFI = 0.946, RMSEA = 0.049, SRMR = 0.06. Figure 5 presents the standardized coefficients for the significant paths in Models 1; Note that all of the figures presented in this paper do not show the covarying items for a cleaner presentation. The other coefficients, along with specific <emph>p</emph>‐values and other statistics, are available in Appendix B.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/PIS/01sep25/pits23515-fig-0005.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="pits23515-fig-0005.jpg" title="5 Model 1 predicting attendance, behaviors, and ELA and math performance from the four SDQ subscales. (R) indicates a reverse‐coded item." /> </p> <p></p> <p>We found that emotional problems predicted lower rates of ODRs and higher ELA achievement (β = −0.422, 0.170, respectively). Peer problems only predicted lower ELA scores (β = −0.145). Hyperactivity‐Inattentiveness predicted higher achievement both in ELA (β = 0.313) and in math (β = 0.328). The conduct problems subscale was the strongest predictor of all four outcomes: it was positively associated with absences (β = 0.172) and ODRs (β = 0.534) and negatively associated with ELA (β = −0.457) and math scores (β = −0.438).</p> <p>Model 2 was structured to test whether the relationships presented in Model 1 exist after controlling for Fall ODRs, thus answering RQ2. The fit statistics of Model 2 were χ<sups>2</sups>(<reflink idref="bib241" id="ref63">241</reflink>) = 1910.49, <emph>p</emph> < 0.001, CFI = 0.945, RMSEA = 0.048, SRMR = 0.06. It is presented in Figure 6 (Appendix C contains a full description of the model). We found that in Model 2, Fall ODRs were also a significant predictor of all outcomes: absences (β = 0.337), Spring ODRs (β = 0.421), ELA scores (β = −0.243), and math scores (β = −0.243). The screener subscales generally remained significant predictors in the same direction as in Model 1, though their effect size was slightly weaker. There were, however, several exceptions. The peer problems subscale now significantly predicted absences (β = 0.124) and Spring ODRs (β = 0.118), and was a stronger predictor of ELA scores (β = −0.172). In addition, the conduct problem subscale no longer significantly predicted absences but was still a strong predictor of ODRs and achievement scores. When comparing the subscales' effect size to that of the fall ODRs, we found that fall ODRs had the strongest effects on absences and spring ODRs, but that the conduct problems and hyperactivity subscales were better predictors of both achievement variables. The other screener subscales had weaker effects.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/PIS/01sep25/pits23515-fig-0006.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="pits23515-fig-0006.jpg" title="6 Model 2 predicting attendance, behaviors, and ELA and math performance from the four SDQ subscales and Fall ODRs. (R) indicates a reverse‐coded item." /> </p> <p></p> <p>Given the surprising results with respect to the hyperactivity‐inattentiveness subscale, we suspected a potential suppression effect, where the association between hyperactivity‐inattentiveness and the outcomes are accounted for by the conduct problems scores. First, we examined the bivariate correlations between the hyperactivity‐inattentiveness scores and the outcomes; we found significant and positive, albeit weak, correlations with absences (<emph>r</emph> = 0.06) and the number of ODRs (<emph>r</emph> = 0.05), and nonsignificant correlations with ELA and math scores. Running models 1 and 2 without conduct problems also showed that hyperactivity‐inattentiveness significantly predicted more absences (β = 0.089 in Model 1; n.s. in Model 2) and ODRs (β = 0.241, 0.106 for Models 1 and 2, respectively). However, hyperactivity‐inattentiveness was still positively associated with ELA (β = 0.068, 0.145) and Math achievement (β = 0.091, 0.167).</p> <p>Finally, we compared Model 3 containing only race and gender as predictors to Model 4 adding the screener subscales to see whether they account for some of the observed relationships between demographics and the EWS outcomes, thus answering RQ3. Models 3 and 4 are presented in Figures 7 and 8, respectively (full results are available in Appendix D). The fit statistics for Model 4 were χ<sups>2</sups>(<reflink idref="bib256" id="ref64">256</reflink>) = 2038.20, <emph>p</emph> < 0.001, CFI = 0.943, RMSEA = 0.048, SRMR = 0.06.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/PIS/01sep25/pits23515-fig-0007.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="pits23515-fig-0007.jpg" title="7 Model 3 predicting attendance, behaviors, and ELA and math performance from the race and gender." /> </p> <p></p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/PIS/01sep25/pits23515-fig-0008.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="pits23515-fig-0008.jpg" title="8 Model 4 predicting attendance, behaviors, and ELA and math performance from race, gender, and the four SDQ subscales. (R) indicates a reverse‐coded item." /> </p> <p></p> <p>In Model 3, race was a significant predictor of Spring ODRs (β = 0.203), ELA scores (β = −0.223), and math scores (β = −0.223), but not absences. That is, Black students had more ODRs and lower achievement scores than other students. Gender was a significant predictor of Spring ODRs (β = −0.144) and ELA scores (β = 0.096), but not absences or math scores. That is, females had fewer ODRs and higher ELA scores than males.</p> <p>In Model 4, after accounting for the screener subscales, Black students still had significantly more Spring ODRs (β = 0.171) and lower ELA and math scores (β = −0.193, −0.210, respectively). These effect sizes are smaller relative to Model 3. Gender was a significant predictor of math scores (β = −0.100), but not Spring ODRs or ELA scores. That is, controlling for the screeners accounted for the gender differences in ODRs and ELA scores and created new differences such that females now had lower math achievement scores.</p> <hd id="AN0187257390-25">Discussion</hd> <p>Existing EWS data structures, designed to identify students at risk of dropout, often rely on attendance, behavior, and course performance data. While useful, recent studies and policies suggest that these indicators could be complemented with a student self‐report behavior screener, particularly as part of an early universal screening process (e.g., Riglin et al. [<reflink idref="bib48" id="ref65">48</reflink>]). This will help schools add preventative measures to identify and support at‐risk students. However, existing studies have not directly tested how specific behavioral and emotional risk indicators (i.e., behavior screener subscales) are associated with the ABCs, a crucial step for understanding the utility of integrating student self‐report screening into an EWS.</p> <p>Our study focused on student self‐report screening data collected within middle schools. We explored the diverging associations between the subscales of a self‐report screener, the SDQ, and the ABC metrics used in traditional EWSs. We also tested whether these subscales predict these end‐of‐year outcomes over and above another early possible EWS metric, Fall ODRs. Finally, we tested whether the subscales account for some of the variance in ABC attributed to race and gender, thus potentially accounting for some of the race and gender bias that could be associated with the ABCs.</p> <p>Our findings indeed show that different subscales have different relationships with ABC metrics. The conduct problems subscale on the SDQ had the strongest association with the ABC data and predicted poorer student outcomes across all metrics. These findings align with prior studies suggesting that students with behavioral concerns are more likely to display behaviors that result in disciplinary action (Kearney [<reflink idref="bib28" id="ref66">28</reflink>]; Lane et al. [<reflink idref="bib32" id="ref67">32</reflink>]).</p> <p>However, hyperactivity‐inattention was predictive of higher ELA and math scores and did not predict absences or ODRs. Testing for a possible suppression effect revealed that when not accounting for conduct problems, hyperactivity‐inattention predicted more absences and behavior problems but was still predictive of higher achievement. This surprising result is contrary to findings on the negative impact of externalizing symptoms on achievement and other ABC outcomes (e.g., Masten et al. [<reflink idref="bib40" id="ref68">40</reflink>]; Zimmermann et al. [<reflink idref="bib58" id="ref69">58</reflink>]), and specifically the role of hyperactivity‐inattentiveness in this relationship (Galéra et al. [<reflink idref="bib17" id="ref70">17</reflink>]). However, more recent works suggest that evidence on the link between hyperactivity‐inattention symptoms and achievement is inconclusive (see a review in Kulkarni et al. [<reflink idref="bib31" id="ref71">31</reflink>]) and is especially weak when achievement is measured using standardized testing as opposed to teachers' assessment (e.g., Okano et al. [<reflink idref="bib45" id="ref72">45</reflink>]), as in our case. While this does not explain why our findings show a positive (rather than null) relationship between hyperactivity‐inattention and achievement, it shows that this relationship is complex and warrants further exploration.</p> <p>The findings concerning the emotional problems and peer problems subscales were surprising, as well. While others find that internalizing symptoms are associated with fewer negative outcomes across the ABC indicators (e.g., Graybill et al. [<reflink idref="bib21" id="ref73">21</reflink>]), our findings show a mixed relationship: emotional problems like stress and anxiety were associated with fewer ODRs and higher ELA scores, but peer problems were only associated with lower ELA scores. This further demonstrates how ABC outcomes may be unlikely to be able to identify students with internalizing issues, further supporting the use of student self‐report screeners as a unique contributor to the identification of at‐risk students.</p> <p>We also found that most of these relationships remained significant and relatively strong even after controlling for Fall ODRs. While early ODRs could detect at‐risk students early in a less resource‐intensive way, they do not capture the full range of emotional and behavioral risks identified by tools like the SDQ. In particular, although the conduct problems scale should measure similar behaviors to those manifested in ODRs, the scale seems to capture slightly different risk factors that predict poorer student outcomes over and above Fall ODRs. That is, our findings support the use of a universal screener for identifying and ideally intervening with students at risk of various behavior and academic problems. Furthermore, behavior screening data are available earlier in the school year, allowing for earlier implementation of prevention and early intervention programs for students identified as at risk. Therefore, the results of this study point to universal screeners as the better alternative.</p> <p>Finally, we found that the screener subscales account for some of the variance in ABC outcomes related to race and gender. ABC outcomes associated with ODRs seem consistently biased in favor of White and female students (e.g., Bradshaw et al. [<reflink idref="bib6" id="ref74">6</reflink>]; Martinez et al. [<reflink idref="bib39" id="ref75">39</reflink>]). This seems to be the case in teacher‐reported behavioral screeners, as well (e.g., Kulkarni and Sullivan [<reflink idref="bib30" id="ref76">30</reflink>]). Our findings show that student‐reported emotional and behavioral health issues can account for the association between ABC outcomes and demographics, again supporting their use to complement traditional EWS to reduce inequities in schools' prevention of dropout. Universal self‐report behavior screeners provide districts with a preventative, proactive and systematically collected data set to inform unbiased decision‐making about students (Tara C. Raines et al. [<reflink idref="bib52" id="ref77">52</reflink>]).</p> <p>In addition to the aforementioned strengths of this study, it extends the literature by focusing on a sample of majority Black, low‐SES schools, contributing to the findings' generalizability to high‐need schools, which may not exhibit the same patterns as other schools. For instance, Young et al. ([<reflink idref="bib57" id="ref78">57</reflink>]) who focused on a suburban, mostly White sample did not find any associations between internalizing scores and ABCs.</p> <p>This study demonstrated that universal behavior screeners like the SDQ provide valuable early identification of students at risk for social, emotional, and behavioral problems, potentially leading to more effective interventions to prevent dropout and later mental health concerns. While it is important to consider the challenges in implementing a universal screening process, including the need for resources and other logistical challenges (Villarreal and Peterson [<reflink idref="bib55" id="ref79">55</reflink>]), we found that incorporating these screeners into EWS can help schools proactively address student needs before significant issues develop, aligning with schools' goals to reduce dropout rates and improve student outcomes (Cruden et al. [<reflink idref="bib12" id="ref80">12</reflink>]).</p> <p>Yet, the study was not without limitations. First, we did not test whether the screener contributes to the prediction of dropout beyond traditional ABC outcomes. As this is the central reason for implementing EWSs, such analyses are required before drawing conclusions regarding the utility of universal screening for enhancing EWSs and contributing to early intervention that prevents dropout. Second, while our study's unique design and sample are strengths that can enrich the existing literature (e.g., using a student‐report instrument, a sample of mostly Black, low‐income, middle school students), the results might not generalize to other contexts. The decision on whether to implement universal screening in particular schools should, therefore, rely on evidence relevant to the schools' characteristics. The sample's uniqueness might also account for our unusual findings concerning hyperactivity‐inattentiveness, further supporting the need for future research in high‐need schools. Finally, school dropout is a systemic issue influenced by disparities in access to resources within certain student populations. Future research could examine data reflective of those social determinants that predict school completion.</p> <p>In spite of these limitations, the study's findings have multiple implications for school psychology. The data reveal the complexity of the relationship between different types of behavioral and emotional symptoms and observable ABC data, enhancing our understanding of early mental health issues and their relationships to student outcomes. The results also demonstrate how symptoms and outcomes correlate in different populations and when using different measures, providing guidance for exploring this issue in different contexts. Finally, it adds to the evidence of the potential of student self‐report universal screening to complement traditional EWSs within districts' public health tiered system of support models. Schools should, therefore, consider implementing student self‐report behavior screening to enhance their dropout prevention programs, especially if the screening can be universal and early.</p> <hd id="AN0187257390-26">Acknowledgments</hd> <p>This paper was developed in part under grant number SM083659 from the Substance Abuse and Mental Health Services Administration (SAMHSA), U.S. Department of Health and Human Services (HHS). The views, policies, and opinions expressed are those of the authors and do not necessarily reflect those of SAMHSA or HHS.</p> <hd id="AN0187257390-27">Ethics Statement</hd> <p>All procedures have been approved by Rutgers University's IRB (protocol number 000043).</p> <hd id="AN0187257390-28">Conflicts of Interest</hd> <p>The authors declare no conflicts of interest.</p> <hd id="AN0187257390-29">Data Availability Statement</hd> <p>Data are confidential and cannot be shared without permission from the school district that provided them.</p> <p>GRAPH: Appendix A.</p> <p>GRAPH: Appendix B.</p> <p>GRAPH: Appendix C.</p> <p>GRAPH: Appendix D.</p> <ref id="AN0187257390-30"> <title> References </title> <blist> <bibl id="bib1" idref="ref17" type="bt">1</bibl> <bibtext> Archambault, I., M. Janosz, J. Morizot, and L. Pagani. 2009. " Adolescent Behavioral, Affective, and Cognitive Engagement in School: Relationship to Dropout." Journal of School Health 79, no. 9 : 408 – 415.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref40" type="bt">2</bibl> <bibtext> Assari, S., A. Mardani, M. Maleki, S. Boyce, and M. Bazargan. 2021. " Black‐White Achievement Gap: Role of Race, School Urbanity, and Parental Education." Pediatric Health, Medicine, and Therapeutics 12 : 1 – 11. https://doi.org/10.2147/PHMT.S238877.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref7" type="bt">3</bibl> <bibtext> Balfanz, R., and V. Byrnes. 2019. " Early Warning Indicators and Intervention Systems: State of the Field." In Handbook of Student Engagement Interventions, edited by J. Fredericks, A. Reschly, and S. Christenson, 45 – 55. Elsevier Science & Technology.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref45" type="bt">4</bibl> <bibtext> Balfanz, R., L. Herzog, and D. J. Mac Iver. 2007. " Preventing Student Disengagement and Keeping Students on the Graduation Path in Urban Middle‐Grades Schools: Early Identification and Effective Interventions." Educational Psychologist 42, no. 4 : 223 – 235. https://doi.org/10.1080/00461520701621079.</bibtext> </blist> <blist> <bibl id="bib5" idref="ref43" type="bt">5</bibl> <bibtext> Barger, B., E. Graybill, A. Roach, and K. Lane. 2020. " Differential Item and Test Functioning of the SRSS‐IE12 Across Race, Ethnicity, Gender, and Elementary Level." Assessment for Effective Intervention 47, no. 2 : 79 – 88. https://doi.org/10.1177/1534508420976830.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref74" type="bt">6</bibl> <bibtext> Bradshaw, C. P., M. M. Mitchell, L. M. O'Brennan, and P. J. Leaf. 2010. " Multilevel Exploration of Factors Contributing to the Overrepresentation of Black Students in Office Disciplinary Referrals." Journal of Educational Psychology 102, no. 2 : 508 – 520. https://doi.org/10.1037/a0018450.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref8" type="bt">7</bibl> <bibtext> Bruce, M., J. M. Bridgeland, J. H. Fox, and R. Balfanz. 2011. The Use of Early Warning and Intervention Systems to Build a Grad Nation. Center for Organization of Schools. https://eric.ed.gov/?id=ED526421.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref21" type="bt">8</bibl> <bibtext> Butterworth, P., and L. S. Leach. 2018. " Early Onset of Distress Disorders and High‐School Dropout: Prospective Evidence From a National Cohort of Australian Adolescents." American Journal of Epidemiology 187, no. 6 : 1192 – 1198. https://doi.org/10.1093/aje/kwx353.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref60" type="bt">9</bibl> <bibtext> Byrne, B. M. 1994. Structural Equation Modeling With EQS and EQS/Windows: Basic Concepts, Applications, and Programming. Sage.</bibtext> </blist> <blist> <bibtext> Cavanaugh, B. 2016. " A Preliminary Investigation Examining the Use of Minor Discipline Referral Data to Identify Students at Risk for Behavioral Difficulties: Observations Within Systems of Schoolwide Positive Behavior Interventions and Supports." Journal of Applied School Psychology 32, no. 4 : 354 – 366. https://doi.org/10.1080/15377903.2016.1207739.</bibtext> </blist> <blist> <bibtext> Chin, M. J., D. M. Quinn, T. K. Dhaliwal, and V. S. Lovison. 2020. " Bias in the Air: A Nationwide Exploration of Teachers' Implicit Racial Attitudes, Aggregate Bias, and Student Outcomes." Educational Researcher 49, no. 8 : 566 – 578. https://doi.org/10.3102/0013189X20937240.</bibtext> </blist> <blist> <bibtext> Cruden, G., K. Kelleher, S. Kellam, and C. H. Brown. 2016. " Increasing the Delivery of Preventive Health Services in Public Education." Supplement, American Journal of Preventive Medicine 51, no. 4 Suppl 2 : S158 – S167. https://doi.org/10.1016/j.amepre.2016.07.002.</bibtext> </blist> <blist> <bibtext> Davis, M. H., M. A. M. Iver, R. W. Balfanz, M. L. Stein, and J. H. Fox. 2019. " Implementation of an Early Warning Indicator and Intervention System." Preventing School Failure 63, no. 1 : 77 – 88. https://doi.org/10.1080/1045988X.2018.1506977.</bibtext> </blist> <blist> <bibtext> Drummond, T. 1994. The Student Risk Screening Scale (SRSS). Josephine County Mental Health Program.</bibtext> </blist> <blist> <bibtext> Eklund, K., T. L. Renshaw, E. Dowdy, et al. 2009. " Early Identification of Behavioral and Emotional Problems in Youth: Universal Screening Versus Teacher‐Referral Identification." California School Psychologist 14, no. 1 : 89 – 95. https://doi.org/10.1007/bf03340954.</bibtext> </blist> <blist> <bibtext> Freudenberg, N., and J. Ruglis. 2007. " Reframing School Dropout as a Public Health Issue." Preventing Chronic Disease 4, no. 4 : 107.</bibtext> </blist> <blist> <bibtext> Galéra, C., M. Melchior, J. F. Chastang, M. P. Bouvard, and E. Fombonne. 2009. " Childhood and Adolescent Hyperactivity‐Inattention Symptoms and Academic Achievement 8 Years Later: The GAZEL Youth Study." Psychological Medicine 39, no. 11 : 1895 – 1906. https://doi.org/10.1017/S0033291709005510.</bibtext> </blist> <blist> <bibtext> Goodman, R. 1997. " The Strengths and Difficulties Questionnaire: A Research Note." Journal of Child Psychology and Psychiatry, and Allied Disciplines 38, no. 5 : 581 – 586. https://doi.org/10.1111/j.1469-7610.1997.tb01545.x.</bibtext> </blist> <blist> <bibtext> Goodman, R. 2001. " Psychometric Properties of the Strengths and Difficulties Questionnaire." Journal of the American Academy of Child and Adolescent Psychiatry 40, no. 11 : 1337 – 1345. https://doi.org/10.1097/00004583-200111000-00015.</bibtext> </blist> <blist> <bibtext> Graybill, E., A. Roach, and B. Barger. 2020. " Factor Structure of the Self‐Report Strength and Difficulties Questionnaire in a Diverse U.S. Sample." Journal of Psychopathology and Behavioral Assessment 43, no. 2 : 388 – 398. https://doi.org/10.1007/s10862-020-09833-4.</bibtext> </blist> <blist> <bibtext> Graybill, E., A. Salmon, B. Barger, and A. T. Roach. 2022. " Examining the Predictive Utility of the Self‐Report Strengths and Difficulties Questionnaire With Middle School Students." International Journal of Mental Health : 1 – 13. https://doi.org/10.1080/00207411.2022.2038983.</bibtext> </blist> <blist> <bibtext> Gregory, C., E. C. Graybill, B. Barger, A. T. Roach, and K. Lane. 2020. " Predictive Validity of the Student Risk Screening Scale‐Internalizing and Externalizing (SRSS‐IE) Scores." Journal of Emotional and Behavioral Disorders 29, no. 2 : 105 – 112. https://doi.org/10.1177/1063426620967283.</bibtext> </blist> <blist> <bibtext> Heller, S., H. A. Pollack, R. Ander, and J. Ludwig. 2013. Preventing Youth Violence and Dropout: A Randomized Field Experiment (No. w19014). National Bureau of Economic Research.</bibtext> </blist> <blist> <bibtext> Hjorth, C. F., L. Bligrav, L. S. Frandsen, et al. 2016. " Mental Health and School Dropout Across Educational Levels and Genders: A 4.8‐year Follow‐Up Study." BMC Public Health 16 : 976. https://doi.org/10.1186/s12889-016-3622-8.</bibtext> </blist> <blist> <bibtext> Izumi, J. T., and K. Eklund. 2023. " Universal Screening for Social–Emotional and Behavioral Risk: Differential Item Functioning on the SAEBRS." School Psychology 38 : 238 – 246. https://doi.org/10.1037/spq/0000542.</bibtext> </blist> <blist> <bibtext> Johfre, S. S., and J. Freese. 2021. " Reconsidering the Reference Category." Sociological Methodology 51, no. 2 : 253 – 269. https://doi.org/10.1177/0081175020982632.</bibtext> </blist> <blist> <bibtext> Jones, C., E. Graybill, B. Barger, and A. T. Roach. 2020. " Examining the Predictive Validity of Behavior Screeners Across Measures and Respondents." Psychology in the Schools 57 : 923 – 936. https://doi.org/10.1002/pits.22371.</bibtext> </blist> <blist> <bibtext> Kearney, C. A. 2002. " Identifying the Function of School Refusal Behavior: A Revision of the School Refusal Assessment Scale." Journal of Psychopathology and Behavioral Assessment 24, no. 4 : 235 – 245. https://doi.org/10.1023/A:1020774932043.</bibtext> </blist> <blist> <bibtext> Kline, R. B. 2015. Principles and Practice of Structural Equation Modeling (Fourth Edition. Guilford Publications).</bibtext> </blist> <blist> <bibtext> Kulkarni, T., and A. L. Sullivan. 2022. " Academic Achievement and Relations to Externalizing Behavior: Much Ado About Nothing? " Journal of School Psychology 94 : 1 – 14. https://doi.org/10.1016/j.jsp.2022.07.004.</bibtext> </blist> <blist> <bibtext> Kulkarni, T., A. L. Sullivan, and J. Kim. 2021. " Externalizing Behavior Problems and Low Academic Achievement: Does a Causal Relation Exist? " Educational Psychology Review 33, no. 3 : 915 – 936. https://doi.org/10.1007/s10648-020-09582-6.</bibtext> </blist> <blist> <bibtext> Lane, K. L., W. P. Oakes, E. W. Carter, W. E. Lambert, and A. B. Jenkins. 2013. " Initial Evidence for the Reliability and Validity of the Student Risk Screening Scale for Internalizing and Externalizing Behaviors at the Middle School Level." Assessment for Effective Intervention 39, no. 1 : 24 – 38. https://doi.org/10.1177/1534508413489336.</bibtext> </blist> <blist> <bibtext> Lansford, J. E., K. A. Dodge, G. S. Pettit, and J. E. Bates. 2016. " A Public Health Perspective on School Dropout and Adult Outcomes: A Prospective Study of Risk and Protective Factors From Age 5 to 27." Journal of Adolescent Health 58, no. 6 : 652 – 658. https://doi.org/10.1016/j.jadohealth.2016.01.014.</bibtext> </blist> <blist> <bibtext> Lawrence, K. C., and T. A. Adebowale. 2023. " Adolescence Dropout Risk Predictors: Family Structure, Mental Health, and Self‐Esteem." Journal of Community Psychology 51 : 120 – 136. https://doi.org/10.1002/jcop.22884.</bibtext> </blist> <blist> <bibtext> Li, C.‐H. 2015. " Confirmatory Factor Analysis With Ordinal Data: Comparing Robust Maximum Likelihood and Diagonally Weighted Least Squares." Behavior Research Methods 48, no. 3 : 936 – 949. https://doi.org/10.3758/s13428-015-0619-7.</bibtext> </blist> <blist> <bibtext> Liu, J., M. S. Hayes, and S. Gershenson. 2024. " JUE Insight: From Referrals to Suspensions: New Evidence on Racial Disparities in Exclusionary Discipline." Journal of Urban Economics 141 : 103453. https://doi.org/10.1016/j.jus.2022.103453.</bibtext> </blist> <blist> <bibtext> Lovelace, M. D., A. L. Reschly, and J. J. Appleton. 2017. " Beyond School Records: The Value of Cognitive and Affective Engagement In Predicting Dropout and On‐Time Graduation." Professional School Counseling 21, no. 1 : 1096 – 2409. https://doi.org/10.5330/1096-2409-21.1.70.</bibtext> </blist> <blist> <bibtext> Margherio, S. M., S. W. Evans, and J. S. Owens. 2019. " Universal Screening In Middle and High Schools: Who Falls through the Cracks." School Psychology 34, no. 6 : 591 – 602. https://doi.org/10.1037/spq000037.</bibtext> </blist> <blist> <bibtext> Martinez, A., S. D. McMahon, and S. Treger. 2016. " Individual‐and School‐Level Predictors of Student Office Disciplinary Referrals." Journal of Emotional and Behavioral Disorders 24, no. 1 : 30 – 41. https://doi.org/10.1177/1063426615588289.</bibtext> </blist> <blist> <bibtext> Masten, A. S., G. I. Roisman, J. D. Long, et al. 2005. " Developmental Cascades: Linking Academic Achievement and Externalizing and Internalizing Symptoms Over 20 Years." Developmental Psychology 41, no. 5 : 733 – 746. https://doi.org/10.1037/0012-1649.41.5.733.</bibtext> </blist> <blist> <bibtext> McIntosh, K., J. L. Frank, and S. A. Spaulding. 2010. " Establishing Research‐Based Trajectories of Office Discipline Referrals for Individual Students." School Psychology Review 39, no. 3 : 380 – 394. https://doi.org/10.1080/02796015.2010.12087759.</bibtext> </blist> <blist> <bibtext> National Center for Educational Statistics. 2017. Trends in High School Dropout and Completion Rates in the United States. Institute of Education Sciences. https://nces.ed.gov/programs/dropout/ind_02.asp#:~:text=2017%20status%20dropout%20rates%20by%20race/ethnicity,the%20rate%20for%20Asian%20youth.</bibtext> </blist> <blist> <bibtext> National Center for Education Statistics. 2023. Status Dropout Rates. Condition of Education. U.S. Department of Education, Institute of Education Sciences. https://nces.ed.gov/programs/coe/indicator/coj.</bibtext> </blist> <blist> <bibtext> Neild, R. C., R. Balfanz, and L. Herzog. 2007. " An Early Warning System." Educational Leadership 65, no. 2 : 28 – 33.</bibtext> </blist> <blist> <bibtext> Okano, L., L. Jeon, A. Crandall, T. Powell, and A. Riley. 2020. " The Cascading Effects of Externalizing Behaviors and Academic Achievement Across Developmental Transitions: Implications for Prevention and Intervention." Prevention Science: The Official Journal of the Society for Prevention Research 21, no. 2 : 211 – 221. https://doi.org/10.1007/s11121-019-01055-9.</bibtext> </blist> <blist> <bibtext> Pinkus, L. 2008. Using Early‐Warning Data to Improve Graduation Rates: Closing Cracks in the Education System. Alliance for Excellent Education.</bibtext> </blist> <blist> <bibtext> Regional Educational Laboratory Pacific. 2016. Four Signs Your District is Ready for an Early Warning System. https://files.eric.ed.gov/fulltext/ED577836.pdf.</bibtext> </blist> <blist> <bibtext> Riglin, L., K. V. Petrides, N. Frederickson, and F. Rice. 2014. " The Relationship Between Emotional Problems and Subsequent School Attainment: A Meta‐Analysis." Journal of Adolescence 37 : 335 – 346. https://doi.org/10.1016/j.adolescence.2014.02.010.</bibtext> </blist> <blist> <bibtext> Rodriguez, L. F., and G. Q. Conchas. 2009. " Preventing Truancy and Dropout Among Urban Middle School Youth: Understanding Community‐Based Action From the Student's Perspective." Education and Urban Society 41, no. 2 : 216 – 247. https://doi.org/10.1177/0013124508325681.</bibtext> </blist> <blist> <bibtext> Rosseel, Y. 2012. " Lavaan: An R Package for Structural Equation Modeling." Journal of Statistical Software 48, no. 2 : 1. https://doi.org/10.18637/jss.v048.i02.</bibtext> </blist> <blist> <bibtext> Shin, R. Q., M. A. Kendall, and E. M. Vera. 2012. " Dropout Prevention: A (re)conceptualization Through the Lens of Social Justice." In The Oxford Handbook of Prevention in Counseling Psychology. Oxford University Press. https://doi.org/10.1093/oxfordhb/9780195396423.013.0014.</bibtext> </blist> <blist> <bibtext> Tara C. Raines, R., D. Bridget V. Dever, K. Randy W. Kamphaus, and R. Andrew T. Roach. 2012. " Universal Screening for Behavioral and Emotional Risk: A Promising Method for Reducing Disproportionate Placement In Special Education." Journal of Negro Education 81, no. 3 : 283 – 296.</bibtext> </blist> <blist> <bibtext> U.S. Department of Education. 2016, September. Issue Brief: Early Warning Systems. Office of Planning, Evaluation and Policy Development, Policy and Program Studies Service. https://www2.ed.gov/rschstat/eval/high-school/early-warning-systems-brief.pdf.</bibtext> </blist> <blist> <bibtext> de Vasconcelos, A. N., L. A. Freires, G. D. L. Loureto, et al. 2023. " Advancing School Dropout Early Warning Systems: The IAFREE Relational Model for Identifying At‐Risk Students." Frontiers in Psychology 14 : 1189283. https://doi.org/10.3389/fpsyg.2023.1189283.</bibtext> </blist> <blist> <bibtext> Villarreal, V., and L. S. Peterson. 2024. " Mental Health Screening: Recommendations From an Integrated Literature Review." Contemporary School Psychology 29 : 250 – 260. https://doi.org/10.1007/s40688-024-00501-y.</bibtext> </blist> <blist> <bibtext> Wentworth, L., and J. Nagaoka. 2020. " Early Warning Indicators in Education: Innovations, Uses, and Optimal Conditions for Effectiveness." Teachers College Record: The Voice of Scholarship in Education 122, no. 14 : 1 – 22. https://doi.org/10.1177/016146812012201401.</bibtext> </blist> <blist> <bibtext> Young, E. L., S. E. Moulton, and A. Julian. 2021. " Integrating Social‐Emotional‐Behavioral Screening With Early Warning Indicators in a High School Setting." Preventing School Failure: Alternative Education for Children and Youth 65, no. 3 : 255 – 265. https://doi.org/10.1080/1045988X.2021.1898319.</bibtext> </blist> <blist> <bibtext> Zimmermann, F., K. Schütte, P. Taskinen, and O. Köller. 2013. " Reciprocal Effects Between Adolescent Externalizing Problems and Measures of Achievement." Journal of Educational Psychology 105, no. 3 : 747 – 761. https://doi.org/10.1037/a0032793.</bibtext> </blist> </ref> <aug> <p>By Emily Graybill; Scott Lewis; Ella Anghel; Sofia Awan; Brian Barger and Ashley Salmon</p> <p>Reported by Author; Author; Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib33" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib51" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib43" firstref="ref3"></nolink> <nolink nlid="nl4" bibid="bib16" firstref="ref4"></nolink> <nolink nlid="nl5" bibid="bib44" firstref="ref5"></nolink> <nolink nlid="nl6" bibid="bib47" firstref="ref6"></nolink> <nolink nlid="nl7" bibid="bib53" firstref="ref10"></nolink> <nolink nlid="nl8" bibid="bib46" firstref="ref13"></nolink> <nolink nlid="nl9" bibid="bib56" firstref="ref14"></nolink> <nolink nlid="nl10" bibid="bib54" firstref="ref15"></nolink> <nolink nlid="nl11" bibid="bib10" firstref="ref16"></nolink> <nolink nlid="nl12" bibid="bib37" firstref="ref18"></nolink> <nolink nlid="nl13" bibid="bib38" firstref="ref19"></nolink> <nolink nlid="nl14" bibid="bib24" firstref="ref20"></nolink> <nolink nlid="nl15" bibid="bib34" firstref="ref23"></nolink> <nolink nlid="nl16" bibid="bib23" firstref="ref24"></nolink> <nolink nlid="nl17" bibid="bib49" firstref="ref26"></nolink> <nolink nlid="nl18" bibid="bib13" firstref="ref27"></nolink> <nolink nlid="nl19" bibid="bib21" firstref="ref28"></nolink> <nolink nlid="nl20" bibid="bib14" firstref="ref29"></nolink> <nolink nlid="nl21" bibid="bib25" firstref="ref30"></nolink> <nolink nlid="nl22" bibid="bib22" firstref="ref31"></nolink> <nolink nlid="nl23" bibid="bib27" firstref="ref33"></nolink> <nolink nlid="nl24" bibid="bib15" firstref="ref34"></nolink> <nolink nlid="nl25" bibid="bib57" firstref="ref35"></nolink> <nolink nlid="nl26" bibid="bib41" firstref="ref36"></nolink> <nolink nlid="nl27" bibid="bib42" firstref="ref37"></nolink> <nolink nlid="nl28" bibid="bib36" firstref="ref39"></nolink> <nolink nlid="nl29" bibid="bib11" firstref="ref41"></nolink> <nolink nlid="nl30" bibid="bib52" firstref="ref42"></nolink> <nolink nlid="nl31" bibid="bib19" firstref="ref44"></nolink> <nolink nlid="nl32" bibid="bib26" firstref="ref46"></nolink> <nolink nlid="nl33" bibid="bib18" firstref="ref53"></nolink> <nolink nlid="nl34" bibid="bib50" firstref="ref55"></nolink> <nolink nlid="nl35" bibid="bib20" firstref="ref56"></nolink> <nolink nlid="nl36" bibid="bib35" firstref="ref58"></nolink> <nolink nlid="nl37" bibid="bib29" firstref="ref59"></nolink> <nolink nlid="nl38" bibid="bib223" firstref="ref62"></nolink> <nolink nlid="nl39" bibid="bib241" firstref="ref63"></nolink> <nolink nlid="nl40" bibid="bib256" firstref="ref64"></nolink> <nolink nlid="nl41" bibid="bib48" firstref="ref65"></nolink> <nolink nlid="nl42" bibid="bib28" firstref="ref66"></nolink> <nolink nlid="nl43" bibid="bib32" firstref="ref67"></nolink> <nolink nlid="nl44" bibid="bib40" firstref="ref68"></nolink> <nolink nlid="nl45" bibid="bib58" firstref="ref69"></nolink> <nolink nlid="nl46" bibid="bib17" firstref="ref70"></nolink> <nolink nlid="nl47" bibid="bib31" firstref="ref71"></nolink> <nolink nlid="nl48" bibid="bib45" firstref="ref72"></nolink> <nolink nlid="nl49" bibid="bib39" firstref="ref75"></nolink> <nolink nlid="nl50" bibid="bib30" firstref="ref76"></nolink> <nolink nlid="nl51" bibid="bib55" firstref="ref79"></nolink> <nolink nlid="nl52" bibid="bib12" firstref="ref80"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1479953
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Universal Behavior Screening and Early Warning System Indicators in Middle Schools
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Emily+Graybill%22">Emily Graybill</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-6912-156X">0000-0002-6912-156X</externalLink>)<br /><searchLink fieldCode="AR" term="%22Scott+Lewis%22">Scott Lewis</searchLink><br /><searchLink fieldCode="AR" term="%22Ella+Anghel%22">Ella Anghel</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0001-6332-7826">0000-0001-6332-7826</externalLink>)<br /><searchLink fieldCode="AR" term="%22Sofia+Awan%22">Sofia Awan</searchLink><br /><searchLink fieldCode="AR" term="%22Brian+Barger%22">Brian Barger</searchLink><br /><searchLink fieldCode="AR" term="%22Ashley+Salmon%22">Ashley Salmon</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Psychology+in+the+Schools%22"><i>Psychology in the Schools</i></searchLink>. 2025 62(9):2955-2968.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 14
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: SourceSuprt
  Label: Sponsoring Agency
  Group: SrcSuprt
  Data: Substance Abuse and Mental Health Services Administration (SAMHSA) (DHHS/PHS)
– Name: NumberContract
  Label: Contract Number
  Group: NumCntrct
  Data: SM083659
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Junior+High+Schools%22">Junior High Schools</searchLink><br /><searchLink fieldCode="EL" term="%22Middle+Schools%22">Middle Schools</searchLink><br /><searchLink fieldCode="EL" term="%22Secondary+Education%22">Secondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Screening+Tests%22">Screening Tests</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Behavior%22">Student Behavior</searchLink><br /><searchLink fieldCode="DE" term="%22Early+Intervention%22">Early Intervention</searchLink><br /><searchLink fieldCode="DE" term="%22Middle+School+Students%22">Middle School Students</searchLink><br /><searchLink fieldCode="DE" term="%22At+Risk+Students%22">At Risk Students</searchLink><br /><searchLink fieldCode="DE" term="%22Potential+Dropouts%22">Potential Dropouts</searchLink><br /><searchLink fieldCode="DE" term="%22Underachievement%22">Underachievement</searchLink><br /><searchLink fieldCode="DE" term="%22Psychological+Patterns%22">Psychological Patterns</searchLink><br /><searchLink fieldCode="DE" term="%22Symptoms+%28Individual+Disorders%29%22">Symptoms (Individual Disorders)</searchLink><br /><searchLink fieldCode="DE" term="%22Behavior+Problems%22">Behavior Problems</searchLink><br /><searchLink fieldCode="DE" term="%22Emotional+Problems%22">Emotional Problems</searchLink><br /><searchLink fieldCode="DE" term="%22Racial+Differences%22">Racial Differences</searchLink><br /><searchLink fieldCode="DE" term="%22Gender+Differences%22">Gender Differences</searchLink><br /><searchLink fieldCode="DE" term="%22Attendance+Patterns%22">Attendance Patterns</searchLink><br /><searchLink fieldCode="DE" term="%22Predictor+Variables%22">Predictor Variables</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1002/pits.23515
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0033-3085<br />1520-6807
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Early warning indicator and intervention systems (EWS) have been promoted to identify students at risk of school underperformance or dropout. Current EWS systems typically include administrative data on attendance, behavior incidents requiring disciplinary action, and course performance. This study tested whether specific emotional and behavioral risk symptoms measured by a student self-report universal screener administered in the fall can predict the three EWS indicators after controlling for fall behavioral incidents and whether they account for some of the variance in EWS attributed to demographic characteristics. Using data from 3307 middle school students, we found that after accounting for fall disciplinary issues, conduct problems predicted poorer student outcomes, but hyperactivity/inattention was predictive of better course performance. Peer problems predicted lower performance in some courses, while emotional problems predicted better performance as well as fewer behavioral issues. We also found that these symptoms accounted for some of the variance in EWS attributed to race and gender. The results suggest that student self-report universal screening can complement existing EWS measures and potentially identify at-risk students who may not otherwise be identified through traditional EWS indicators.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1479953
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1479953
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1002/pits.23515
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 14
        StartPage: 2955
    Subjects:
      – SubjectFull: Screening Tests
        Type: general
      – SubjectFull: Student Behavior
        Type: general
      – SubjectFull: Early Intervention
        Type: general
      – SubjectFull: Middle School Students
        Type: general
      – SubjectFull: At Risk Students
        Type: general
      – SubjectFull: Potential Dropouts
        Type: general
      – SubjectFull: Underachievement
        Type: general
      – SubjectFull: Psychological Patterns
        Type: general
      – SubjectFull: Symptoms (Individual Disorders)
        Type: general
      – SubjectFull: Behavior Problems
        Type: general
      – SubjectFull: Emotional Problems
        Type: general
      – SubjectFull: Racial Differences
        Type: general
      – SubjectFull: Gender Differences
        Type: general
      – SubjectFull: Attendance Patterns
        Type: general
      – SubjectFull: Predictor Variables
        Type: general
    Titles:
      – TitleFull: Universal Behavior Screening and Early Warning System Indicators in Middle Schools
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Emily Graybill
      – PersonEntity:
          Name:
            NameFull: Scott Lewis
      – PersonEntity:
          Name:
            NameFull: Ella Anghel
      – PersonEntity:
          Name:
            NameFull: Sofia Awan
      – PersonEntity:
          Name:
            NameFull: Brian Barger
      – PersonEntity:
          Name:
            NameFull: Ashley Salmon
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 09
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 0033-3085
            – Type: issn-electronic
              Value: 1520-6807
          Numbering:
            – Type: volume
              Value: 62
            – Type: issue
              Value: 9
          Titles:
            – TitleFull: Psychology in the Schools
              Type: main
ResultId 1