Discipline referrals and access to secondary level support in elementary and middle schools: patterns across African-American, Hispanic-American, and White students.
Subject:
Book publishing (Case studies)
Child psychopathology (Case studies)
Education grants (Case studies)
Multiculturalism (Case studies)
African Americans (Surveys)
African Americans (Case studies)
Teachers (Surveys)
Teachers (Case studies)
Decision-making (Case studies)
Disabled students (Surveys)
Disabled students (Case studies)
Children (Behavior)
Children (Case studies)
Authors:
Vincent, Claudia G.
Tobin, Tary J.
Hawken, Leanne S.
Frank, Jennifer L.
Pub Date:
08/01/2012
Publication:
Name: Education & Treatment of Children Publisher: West Virginia University Press, University of West Virginia Audience: Professional Format: Magazine/Journal Subject: Education; Family and marriage; Social sciences Copyright: COPYRIGHT 2012 West Virginia University Press, University of West Virginia ISSN: 0748-8491
Issue:
Date: August, 2012 Source Volume: 35 Source Issue: 3
Topic:
Canadian Subject Form: Child behaviour
Product:
Product Code: 2731000 Book Publishing; 9105113 Institutional Grants NAICS Code: 51113 Book Publishers; 92311 Administration of Education Programs SIC Code: 2731 Book publishing
Organization:
Government Agency: United States. Department of Education; United States. National Center for Education Statistics Company Name: Guilford Publications Inc. Organization: University of Oregon
Geographic:
Geographic Scope: United States; New York Geographic Code: 1U2NY New York

Accession Number:
301649977
Full Text:
Abstract

Given documented racial/ethnic disproportionality in disciplinary referrals and strong recommendations to base provision of secondary level supports on data, we explored whether students from various racial/ethnic groups have equitable access to secondary supports. We disaggregated discipline data from 155 elementary and 46 middle schools by student race/ethnicity and behavioral risk level to assess the extent to which different racial/ethnic groups were disproportionately represented among students at each risk level and students receiving secondary support. Outcomes indicated that Hispanic-American and White students were underrepresented among students with multiple disciplinary referrals, while African-American students were over-represented. African-American students were over-represented among students receiving secondary support in elementary schools but were less likely to receive this support in middle schools. Across all schools, number of referrals as well as race/ethnicity emerged as statistically significant predictors of access to secondary level support. Limitations to the current investigation and recommendations for future research are provided.

It has become a well-known fact that students from non-White backgrounds, especially African-American and Hispanic-American students, experience poorer discipline and academic outcomes in the United States public school system than their White peers. For African-American students, research has documented disproportionately high numbers of office discipline referrals (Bradshaw, Mitchell, O'Brennan, & Leaf, 2010; Kaufman et al., 2010; Skiba, et al., 2011; Skiba, Peterson, & Williams, 1997; Vincent, Tobin, Swain-Bradway, & May, 2011), comparatively harsher punishments for behavioral violations (Glackman et al., 1978; Gregory, 1995; Shaw & Braden, 1990; Skiba, Michael, Nardo & Peterson, 2002; Skiba & Peterson, 2000), and increased odds for being suspended Or expelled (KewelRamani, Gilbertson, Fox, & Provasnik, 2007; Krezmien, Leone, & Achilles, 2006; Wallace, Goodkind, Wallace, & Bachman, 2008). These behavioral outcomes are accompanied by comparatively lower reading and math achievement (Lee, 2000; Lee, Grigg, & Donahue, 2007) and overidentification for special education services (Coutinho & Oswald, 2000; Harry & Klingner, 2006; Waitoller, Artiles, & Cheney, 2010; Zhang, Katsiyannis, & Herbst, 2004). For Hispanic-American students, research has documented disproportionately high rates of suspension beginning in middle school (Skiba et al., 2011), comparatively high levels of anxiety and depression (Fletcher, 2008; McLaughlin, Hilt, & Nolen-Hoeksema, 2007; Varela, Sanchez-Sosa, Biggs, & Luis, 2008; Zayas, Lester, Cabassa, & Fortuna, 2005), and high drop-out rates (Stillwell, 2010). Similar to African-American students, Hispanic-American students lag far behind their White peers in reading and math at the elementary, middle, and high school level (Aud, Fox, & KewalRamani, 2010).

While documentations of racially disproportionate educational outcomes abound, our understanding of why these outcomes occur and what might change them remains limited. General guidelines to redress racial inequity in educational outcomes include culturally responsive evidence-based behavior support delivered within a response to intervention (RtI) framework as a vehicle to increase students' time engaged with academic instruction (Cartledge, Singh, & Gibson, 2008; Klingner et al., 2005; Skiba et al., 2008). The key recommendations for culturally responsive delivery of evidence-based behavior support within an RtI framework are (a) availability of support structures of varying intensity; (b) continuous data collection for screening, diagnostic, and progress monitoring purposes; (c) interpretation of those data to determine a student's responsiveness to existing support mechanisms; (d) open discussions of race in data interpretations; (e) early and culturally specific intervention in the form of additional support for students who are at risk of behavioral failure; and (f) family involvement (Gresham, Lane, & Lambros, 2000; Hawken, Vincent, & Schumann, 2008; Lane, Kalberg, & Menzies, 2009; Schumann & Burrow-Sanchez, 2010; Skiba et al., 2008).

Response to Intervention and School-wide Positive Behavior Support

One approach to student behavior support delivery that uses the RtI framework to prevent students' behavioral failure and maximize their access to academic instruction is school-wide positive behavior support (SWPBS). SWPBS involves (a) a whole school effort to proactively teach and positively acknowledge appropriate behavior (Bradshaw, Koth, Thornton, & Leaf, 2009), (b) making data-based decisions regarding students' support needs (Kaufman et al., 2010; Sugai & Lewis, 1999; Technical Assistance Center on Positive Behavioral Interventions and Supports, 2010; Tobin, 2006), and (c) having in place procedures for identifying students who need additional support and being able to promptly provide the level of support needed (Fairbanks, Sugai, Guardino, & Lathrop, 2007; Sailor, Dunlap, Sugai, & Homer, 2009; Shinn, Walker, & Stoner, 2002). Full implementation of SWPBS merges universal (primary level) support for all students at all times in all settings with secondary level support for groups of students requiring more assistance and tertiary level support for individual students with the most intense behavioral challenges.

Secondary level support is intended for students who, despite their exposure to universal support, continue to receive disciplinary referrals. The key features of secondary level support are (a) continuous availability to all students, (b) increased adult contact, and (c) increased monitoring of behavioral performance. For example, Second Step (Committee for Children, 2010), Check and Connect (Christenson et al., 2008), and Check, Connect, and Expect (Cheney et al., 2010) are secondary interventions meeting these criteria.

Many schools implementing SWPBS use Check-in/Check-out (CICO; Crone, Hawken, & Homer, 2010) as a secondary level intervention to support students who are unsuccessful with universal support alone. Also known as the Behavior Education Program, CICO is a popular secondary level intervention because it can be efficiently provided to students who need more than universal support (Hawken, 2006; Hawken, Adolphson, MacLeod, & Schumann, 2009; Hawken & Homer, 2003; Hawken & Johnson, 2007; Hawken, MacLeod, & Rawlings, 2007; Hawken, Pettersson, Mootz, & Anderson, 2005; Hawken et al., 2008; March & Horner, 2002; McIntosh & Av-Gay, 2007; McIntosh, Campbell, Carter, & Dickey, 2009). A student who is placed on CICO checks in with a coordinator (teacher or other school staff member) at the beginning of the day, receives a daily progress report (DPR) to present to each teacher who uses it to provide feedback (points) at the end of each class period, and checks out at the end of the day with the CICO coordinator. Students are encouraged to take their DPR home to solicit feedback from their parent(s) and return it signed by a parent to the CICO coordinator the next morning (Crone et al., 2010). The critical features of the CICO secondary level intervention are (a) increased adult contact through checking in and out with the CICO coordinator as well as receiving feedback from the teacher via the DPR, (b) close monitoring of behavioral performance through points earned, and--if available--(c) regular parental feedback through parents' review of the DPR. The overall goal of CICO is to decrease a student's time involved in disciplinary violations and increase his or her time engaged with the academic curriculum.

The effectiveness of CICO has been empirically documented. CICO has been effective in reducing problem behavior at the elementary (Cheney et al., 2009; Fairbanks et al., 2007; Filter, Benedict, Horner, Todd, & Watson, 2007; Hawken et al., 2007; McCurdy, Kunsch, & Reibstein, 2007; Stage, Cheney, Flower, Templeton, & Waugh, in review; Todd, Kaufman, Meyer, & Homer, 2008), middle (Hawken, 2006; Hawken & Homer, 2003; March & Homer, 2002), and high school level (Swain-Bradway & Homer, 2010). CICO has also been effective with students in urban settings (McCurdy et al., 2007). Following implementation of CICO, students have shown increases in academically engaged time (Hawken & Homer, 2003; Swain-Bradway & Homer, 2010), as well as a decreased need for additional behavior support and referrals to special education (Hawken et al., 2007). The majority of the studies cited above found CICO to be effective with more than 65% of the students who received the intervention.

Schools use various strategies to identify students who might benefit from CICO. One readily available indicator that a student needs more than universal support alone is frequency of office discipline referrals (ODR). In general, students with 0-1 ODR are deemed at low risk for behavioral failure and successful with universal support alone, students with 2-5 ODR are considered at medium risk and eligible for secondary level support, and students with 6+ ODR are considered at high risk and in need of tertiary level support (Homer, Sugai, Todd, & Lewis-Palmer, 2005). Routinely collected by many schools, ODR provide useful information about the type of behavioral violation a student engages in and when and where it tends to occur. Although ODR are limited by potential differences in teacher tolerance and tend to capture mostly externalizing behavior (Hawken et al., 2008; Nelson, Gonzales, Epstein, & Benner, 2003), they do serve as a valid and reliable behavioral metric for identifying students at elevated risk for behavioral failure. It is important to consider that ODR represent events involving (a) a student who engages in a specific behavior, (b) a teacher who observes the behavior, and (c) an administrator who delivers a consequence (Sugai, Sprague, Homer, & Walker, 2000). As such, ODR do not only reflect student behavior, but also teacher and staff behavior.

Research contributing to the evidence-base for SWPBS has primarily focused on changes in rates of ODR. Aggregated across entire school populations, ODR rates from schools implementing the universal tier of SWPBS consistently show a downward trend (Bohanon et al., 2006; Bradshaw, Koth, et al., 2009; Bradshaw, Mitchell, & Leaf, 2010; Bradshaw, Reinke, Brown, Bevans, & Leaf, 2008; Homer et al., 2009; Muscott, Mann, & LeBrun, 2008). However, disaggregated by student race and ethnicity, ODR rates from schools implementing SWPBS show that disciplinary inequity remains essentially unchanged (Kaufman et al., 2010; Vincent, Cartledge, May & Tobin, 2009; Vincent et al., 2011) and referral patterns by race/ethnicity mirror those found across the United States (Aud et al., 2010).

Access to Secondary Support for Students from Various Racial/Ethnic Backgrounds

If implementation of the universal level of SWPBS is not associated with greater discipline equity across ethnicities, one solution to try to "equalize" disciplinary outcomes would be to provide non-White students who are at risk access to secondary level support to stop or slow the trajectory in relation to acquiring ODR. Therefore, our primary interest was in exploring if students who receive multiple ODR (a) get access to secondary level support at a rate proportionate to their representation at the risk level and (b) get access promptly. Secondarily, we were interested in exploring if students from different racial/ethnic groups who get access to secondary level support benefit equitably. Thus, the focus of this study was to investigate whether race/ethnicity predicts access to secondary level support, and if the outcome of secondary level support differed across students from different racial/ethnic backgrounds.

The following research questions drove our analyses:

1. Are African-American, Hispanic-American, and White students disproportionately represented among students at low (0-1 ODR), medium (2-5 ODR) and high (6+ ODR) risk for behavioral failure in elementary and middle schools?

2. Are African-American, Hispanic-American, and White students at each risk level disproportionately represented among students receiving CICO?

3. Does the probability of a student gaining access to CICO vary with number of ODR received, with racial-ethnic background, or with both?

4. Do reductions in ODR for students at medium and high-risk levels, who receive CICO, vary with racial-ethnic background, with CICO start date, or with both?

Method

We relied exclusively on extant data to answer our research questions. As such, our method was situated within an exploratory data analysis (EDA) paradigm (Leinhardt & Leinhardt, 1980). This approach allowed us to examine data for patterns and relationships that could be used to guide future research focused on the provision of secondary behavior support. Although EDA allows continuous adjustment of research questions following observed patterns in the data, we limited our explorations to the research questions identified above.

Our extant data were collected with the School-wide Information System (SWIS; May et al., 2005). SWIS allows schools to record, aggregate, and review ODR data. SWIS users can gather information about the frequency, location, time, and type of behavioral offenses, demographic information about the general student body and individual students engaging in inappropriate behavior, and administrative consequences following behavioral offenses. Recently, SWIS has added a CICO module that allows schools to track the behavioral performance of students on CICO. The information available on SWIS-CICO includes number of students on CICO per school, demographic information about individual students receiving CICO, start date of CICO implementation, and ODR received by individual students on CICO. Before gaining access to SWIS-CICO, schools need to meet certain readiness criteria, including (a) implementing universal support to criterion, (b) having a CICO coordinator and training materials, (c) having a process for identifying students for CICO, (d) having a designated data entry person and schedule, and (e) having a team-based process for using CICO data for decision-making (Todd, Homer, & Rosetto Dickey, 2010). SWIS and SWIS-CICO data from schools that agree to share their data for research purposes become part of the SWIS research database housed at the University of Oregon.

Sample Size and Characteristics

We examined extant discipline data collected with SWIS during the 2009-2010 academic year. We included schools in our data set that (a) used SWIS-CICO, (b) recorded their overall school enrollment by ethnicity, (c) were public elementary or middle schools located in the United States, and (d) had agreed to share their data for research purposes. A total of 155 elementary and 46 middle schools met these criteria. For each school, we had school-level data (demographics, average ODR per 100 students per day, number of students on CICO) and student-level data (ethnicity, CICO start date, number of ODR received). Table 1 provides an overview of the geographic location of the sample, average enrollment, average number of ODR per 100 students per day, total students enrolled at each school level, and the total number of students on CICO by school level.

Because the majority of students enrolled in the elementary and middle schools meeting inclusion criteria were of Hispanic-American, African-American, or White background, we focused our analyses on these ethnicities. Table 2 provides an overview of the number of students at each risk tier enrolled by race/ethnicity and school level. It is important to note that for a sizeable proportion of students at each risk tier and school level race/ethnicity was recorded as "not listed" or "unknown." Our sample for research questions 1, 2, and 3 included all students of Hispanic-American, African-American, and White backgrounds. Thus, our elementary sample consisted of a total of n = 7323 students (n = 1505 Hispanic-American, n = 2332 African-American, and n = 3486 White students). Our sample size for the middle school analysis was n = 6024 (n = 1163 Hispanic-American, n = 1990 African-American, and n = 2871 White students). Because research question 4 focused on the benefits of CICO for students at medium and high risk for behavioral failure, only students with 2 or more ODR in semester 1 of the academic year (prior to December 31) who were of Hispanic-American, African-American, or White backgrounds and received CICO at any point during the academic year were included in the analyses. Our elementary school sample for research question 4 consisted of n = 372 students (n = 72 Hispanic-American, n = 162 African-American, and n = 138 White students) and our middle school sample consisted of n = 159 students (n = 35 Hispanic-American, n = 48 African-American, and n = 76 White students).

Analytical Procedures

We used descriptive analyses only to answer research questions 1 and 2. For research question 1 (racial/ethnic groups' representation at each risk tier), we calculated the proportion students of each race/ethnicity represented of the total enrollment, and the proportion students of each race/ethnicity represented of the total students at each risk tier. For example, if in a school whose student population was 18% Hispanic-American, 18% of students with 0-1 ODR, 10% of students with 2-5 ODR, and 30% of students with 6+ ODR were also Hispanic-American, Hispanic-American students would be proportionately represented among students at low risk, under-represented among students at medium risk, and over-represented among students at high risk for behavioral failure. We represented the magnitude of disproportionate discipline outcomes by subtracting the proportion of students of a given race/ethnicity and at a given risk tier from the proportion of students of the same race/ethnicity enrolled. Thus, a value of 0 represented proportionate representation, negative values represented under-representation, and positive values represented over-representation. To compare outcomes across school levels, we compiled these descriptive statistics separately for elementary schools and middle schools.

To answer research question 2 (racial/ethnic groups' representation among students receiving CICO), we performed similar calculations at the next level of disaggregation: We calculated the proportion of students of each race/ethnicity at each risk level who received CICO and subtracted it from the proportion of students of each race/ethnicity at each risk level. For example if in the school above, where 18% of students with 0-1 ODR, 10% of students with 2-5 ODR, and 30% of students with 6+ ODR were Hispanic-American, 25% of the students with 0-1 ODR who received CICO, 18% of students with 2-5 ODR who received CICO, and 5% of students with 6+ ODR who received CICO were also Hispanic-American, Hispanic-American students would be over-represented among students with access to CICO at the low-risk tier, proportionately represented among students with access to CICO at the medium risk tier, and under-represented among students with access to CICO at the high-risk tier. We again performed separate calculations for elementary and middle schools, to see if outcomes differed by school level.

To investigate if race/ethnicity predicted access to secondary level support (research question 3), we conducted logistic regression to examine if number of ODR received, racial-ethnic background, or both were significant predictors of CICO access. In logistic regression, prediction errors do not violate homoscedasticity; therefore normality is not assumed and predictors can be distributed in various ways. We conducted separate analyses for the elementary and middle school samples. Because we expected that students who accumulated multiple ODR early in the school year would be most likely to receive CICO, we used ODRs received during semester 1 of the school year (beginning of school year through December 31) as one predictor. Racial-ethnic background served as a second predictor, with "White" as the reference category. Because we had no a-priori hypothesis about which predictor would be stronger, both predictors were added into the equation at the same time. We interpreted the deviation statistic to evaluate the significance of our model against the null model; and we interpreted the significance of the exponentiated coefficients to evaluate odds ratios.

Research question 4 focused on reductions in ODR across time for students at medium and high risk for behavioral failure who received CICO. Repeated measures ANOVA with ODR at time 1 (prior to December 31) and ODR at time 2 (after January 1) could not be conducted because our data did not meet the basic assumption of normality. The distribution of our dependent variable at both points in time was strongly positively skewed, and square-root, log10, and inverse transformations failed to normalize the distributions. Thus, we computed ODR difference scores by subtracting Semester 1. ODR from Semester 2 ODR so that negative values represented decreases in ODR and positive values represented increases in ODR and conducted factorial ANOVA with ODR difference scores as the dependent variable. Between subject-factors were student race/ethnicity with three levels (Hispanic-American, African-American, White) and CICO start date with two levels: "early starters" (students who received CICO within approximately 4 months of becoming eligible) and "late starters" (students who did not received CICO until more than 4 months after becoming eligible). The distribution of the ODR difference scores met the assumption of normality. The assumption of independent observations was also met. Given approximately equal cell sizes, ANOVA outcomes are largely robust to violations of homogeneity of variance (Howell, 2002). Because our cell sizes differed, outcomes need to be interpreted with caution. We again conducted separate analyses for the elementary and the middle school samples.

Results

The outcomes of our descriptive analyses to examine the extent to which students from Hispanic-American, African-American, and White backgrounds are disproportionately represented among students at low, medium, and high risk for behavioral failure (research question 1) and among students who receive CICO at each risk level (research question 2) are presented in Figures 1 and 2. The left side of Figure 1 depicts the magnitude of disproportionate discipline outcomes in elementary schools for Hispanic-American, African-American, and White students at low, medium, and high-risk levels. While Hispanic-American and White students were substantially underrepresented among students at all risk levels, African-American students were slightly under-represented at the low-risk level, and over-represented at the medium and high-risk level. The right side of Figure 1 depicts the magnitude of disproportionate access to CICO for Hispanic-American, African-American, and White elementary school students at low, medium, and high-risk levels. African-American students were substantially over-represented at each risk tier among students on CICO. The over-representation of Hispanic-American students among students on CICO decreased with increasing risk for behavioral failure. White students were slightly over-represented among students on CICO at the low and medium risk tier, and under-represented at the high-risk tier.

[FIGURE 1 OMITTED]

The left side of Figure 2 depicts the magnitude of disproportionate discipline outcomes in middle schools for Hispanic-American, African-American, and White students at low, medium, and high-risk levels. In general, the racial/ethnic discrepancy of discipline outcomes that was apparent in elementary schools was magnified in middle schools. White students were under-represented by more than 20 percentage points among students at each risk level. Hispanic students were less drastically under-represented, and African-American students were under-represented among students at low risk, and over-represented at medium and high risk for behavioral failure. The right side of Figure 2 again depicts the magnitude of disproportionate access to CICO for Hispanic-American, African-American, and White middle school students at low, medium, and high-risk levels. The over-representation of Hispanic-American students decreased with increasing risk level, while the over-representation of White students increased with increasing risk level. African-American students, however, were slightly under-represented among students on CICO at the low-risk level, approximately proportionately represented at the medium risk level, and somewhat over-represented at the high-risk level.

[FIGURE 2 OMITTED]

The logistic regression outcomes for research question 3 (Does the probability of a student gaining access to CICO vary with number of ODR received, with racial-ethnic background, or with both?) were as follows: For the elementary school sample, the test of the null model against the full model was statistically significant ([X.sup.2] (3, N =7323) = 221.585, p<.0005), indicating that both predictors (ODR received during semester 1 and race/ethnicity) together distinguished between receiving and not receiving CICO. The test of the full model showed that semester 1 ODR was a statistically significant predictor of receiving CICO (p<.0005) as was race/ethnicity (p<.0005). EXP (B) values indicated that with each added ODR, students were 1.247 as likely to receive CICO. Compared to White students, Hispanic-American students were 1.420 as likely to receive CICO, and African-American students were 1.624 as likely. The results for the middle school sample were similar. The test of the null model against the full model was statistically significant ([X.sup.2] (3, N =6024) = 116.709, p<.0005), indicating that both predictors together distinguished between receiving and not receiving CICO. The test of the full model indicated that semester 1 ODR was a statistically significant predictor of receiving CICO (p<.0005) as was race/ethnicity (p<.0005). EXP (B) values indicated that with each added ODR, students were 1.193 as likely to receive CICO. Compared to White students, Hispanic-American students were 1.469 as likely to receive CICO, and African-American students were .749 as likely.

Table 3 summarizes the results for the elementary school sample of the factorial ANOVA to answer research question 4 (Do reductions in ODR for students at medium-and high-risk levels who receive CICO vary with racial-ethnic background, with CICO start date, or with both?) Given that there was no significant interaction effect, we examined main effects for statistical significance. Only CICO start date emerged as a statistically significant main effect, F (2,366) = 13.351, r.0005, and accounted for 3.5% of the between subject variance. The main effect of race/ethnicity was non-significant, F (2, 366) = 32.146, p = .194. Students who received CICO during the first semester of the school year experienced statistically significant decreases in ODR (M = -.42, SD 3.04 for Hispanic-American students, M = -.88, SD 3.92 for African-American students, and M = -.35, SD = 4.28 for White students). Figure 3 provides an overview of the number of 1st and 2nd semester ODR by race/ethnicity for early (left panel) and late (right panel) CICO starters in elementary school.

[FIGURE 3 OMITTED]

The ANOVA results for the middle school sample were slightly different. Table 4 provides a summary of the outcomes. Because the interaction effect was again non-significant, we examined main effects. Both race/ethnicity (F (2,153) = 3.887, p = .023) and CICO start date (F (1, 153) = 9.608, p = .002) emerged as statistically significant main effects, with CICO start date accounting for 5.9% and race/ethnicity accounting for 4.8% of the between-subject variance. Overall, all students experienced increases in ODR from semester 1 to semester 2; however students who received CICO within 4 months of becoming eligible experienced statistically significantly lesser increases than those who had to wait: M = 1.57, SD = 5.78 for Hispanic-American students, M = -2.4, SD = 4.88 for African-American students, and M = .86, SD = 5.63 for White students. It is interesting to note that only African-American students who received CICO experienced a mean decrease in ODRs. A post-hoc Tukey HSD test for the main effect of race indicated that the difference between African-American students and their peers was statistically non-significant. This atypical non-significant post-hoc result following a significant omnibus test is likely due to different cell sizes. However, the means suggest that African-American students appear to be benefitting from timely access to CICO in middle school. Figure 4 provides an overview of the number of 1st and 2nd semester ODR by race/ethnicity for early (left panel) and late (right panel) CICO starters in middle school.

[FIGURE 4 OMITTED]

Discussion

Our investigation was driven by the overall hypothesis that--given widely documented racial/ethnic disproportionality in discipline outcomes and the emerging evidence that universal behavior support alone does not appear to reduce the problem (Losen & Orfield, 2002; National Research Council, 2003)--schools might use secondary level support to "equalize" discipline outcomes across students from various ethnic backgrounds. Within this context, a number of important messages emerged. For both the elementary and the middle school sample, both number of ODR and race/ethnicity emerged as significant predictors of access to CICO. Number of ODR is expected to predict access to CICO; students with multiple ODR are assumed to be eligible for more than universal support (Hawken et al., 2009). At first blush, it seems surprising that student race/ethnicity should predict access to CICO; however, in conjunction with our other outcomes, various interpretations present themselves. In our elementary school sample, African-American students were over-represented among students with multiple ODR as well as among students on CICO. Among all the student groups included in our investigation, they were most likely to receive CICO. If they received CICO early in the academic year, they received fewer ODR in the remaining months of the school year. Given this positive outcome for African-American students, perhaps provision of CICO was intended to reduce the over-representation of African-American students in discipline incidents.

A different message emerged from our middle school sample. Although African-American students were similarly over-represented among students with multiple ODR, being African-American predicted a lesser likelihood to receive CICO. Those few African-American students who did receive CICO in middle school, and who received it promptly, however, were impressively successful in lowering their ODR. This combination of outcomes in our middle school sample suggests that schools did not use CICO simply in response to number of ODR a student accumulated, but that other considerations played a role. It appears difficult to imagine that schools would be reluctant to give a student access to CICO based on his or her racial/ethnic background. Perhaps the fact that adults tend to interpret inappropriate behavior African-American students engage in as non-compliance with or defiance to adult directions (Skiba et al., 2011) makes CICO, whose critical component is increased adult contact, appear less promising at the middle school level. Perhaps African-American students are also less willing to engage with more adult contact at the middle school level, when their trust in teachers tends to erode (Crosnoe, Johnson & Elder, 2004, Delpit, 1992). Given the behavioral success African-American middle school students experienced with CICO, it seems important that middle school behavioral support policies be implemented with greater cultural responsiveness to help prevent African-American students' lack of support and consequently--perhaps--growing alienation from school.

In our elementary school sample, Hispanic-American students as well as White students were under-represented among students with multiple ODR; Hispanic-American students, however, were over-represented among students receiving access to CICO. They were more likely than White students to receive CICO, and--if they received it promptly--they benefited from it at a rate similar to White students. A similar pattern emerged from the middle school sample. Although Hispanic-American students were less under-represented among students with ODR, they remained over-represented among students on CICO. In the case of Hispanic-American students, number of ODR appeared to be not the primary consideration of providing CICO; based on our analysis, student race appeared to be a consideration. Given evidence that Hispanic-American students tend to engage in internalizing behavior (Varela et al., 2008), CICO might be considered a useful intervention to provide more adult contact and encouragement to students who tend to be socially withdrawn based on their cultural background. Cheney et al. (2009) has found that students with internalizing disorders respond positively to CICO-type interventions. It might also be important to consider that CICO is intended for students with adult attention as the primary function of problem behavior. It might be easier to provide adult attention to a student engaging in internalizing behavior than a student engaging in externalizing behavior.

It was surprising to see that Hispanic-American students' over-representation among students receiving CICO decreased with increasing risk level in both elementary and middle schools. This pattern is difficult to interpret. Perhaps Hispanic-American students who accumulated multiple ODR engaged in behavior that was deemed less responsive to the CICO intervention. White students' over-representation among students on CICO increased with increasing risk level in the middle school sample, suggesting that students who repeatedly engaged in inappropriate behavior were considered good candidates for CICO. Perhaps specific behavior students engaged in and their expected responsiveness drove this difference in Hispanic-American and White students' access to CICO or by different expectations for social success.

It was interesting to note that in elementary as well as middle schools, students at the low-risk tier who had not accumulated multiple ODR were also given access to CICO. Perhaps CICO was used as a preventative strategy for students who showed signs of behavioral difficulties without accumulating multiple ODR, This might have been especially the case for Hispanic-American students engaging in internalizing behavior not resulting in ODR. Although number of ODR is a common indicator of CICO eligibility, type of behavior and behavioral motivation are important other considerations (Crone et al., 2010).

Another important message emerging from our study was that timely access to CICO was critical for behavioral success, while delayed access to CICO resulted in further increases in ODR. The pattern for early CICO starters in middle school is especially interesting. African-American students, given access to CICO early in the academic year, were the only students who achieved significant behavioral gains as documented by their decreased number of ODR in semester 2. Hispanic-American and White students who received CICO early were able to slow their accumulation of ODR compared to Hispanic-American and White students who received CICO late. African-American students who received CICO late also experienced an increase in ODR. The overall increase in ODR in our middle school sample might be partially due to middle schools having generally higher ODR rates than elementary schools, and ODR rates generally increasing across the academic year as teacher tolerance declines.

Our hope was that the disproportionate representation of African American and Hispanic-American students in disciplinary incidents could be alleviated through provision of secondary level interventions. Unfortunately, although we found that elementary schools gave Hispanic-American and African-American students disproportionately greater access to CICO than White students, the overall data from this evaluation indicated that, in most cases, implementation of secondary level support did not decrease the discipline disparity.

The majority of evaluations of CICO have been conducted at an individual school or district level with relatively small samples sizes which do not allow for disaggregation of data by race/ethnicity (e.g., Filter et al., 2007; Hawken et al., 2007; Hawken & Horner, 2003). This study provided a large-scale evaluation of CICO across districts and states, which allowed us to examine the differential effectiveness of CICO by race/ethnicity. Schumann and Burrow-Sanchez (2010) provide specific recommendations on adaptations to CICO to support students from diverse backgrounds and future research on CICO should evaluate these recommendations to determine if they increase effectiveness across racial groups.

Limitations

Interpretation of our study outcomes is limited by a number of considerations. Most importantly, the large number of students whose race/ethnicity was coded as "not listed/unknown" in SWIS created a sizeable error source that might have impacted our disproportionality calculations. Because we focused our analyses on African-American, Hispanic-American, and White students only, results may not generalize to students from other racial groups or those of mixed racial backgrounds. An equally important limitation was our lack of knowledge about how our sample schools implemented SWPBS and CICO. Although the SWIS-CICO readiness checklist stipulates that only schools that implement universal support to criterion gain access to SWIS-CICO, there is little information about how this requirement is met, i.e., what criterion is applied. Similarly, the SWIS-CICO readiness checklist stipulates that schools must have a process for identifying students eligible for CICO and use CICO data for decision-making. However, there is also little information about what the identification process looks like or how data are used. Further inquiry into how specific CICO implementation practices impact student outcomes across racial/ethnic groups is warranted.

Interpretation of our study's results was also limited due to sample characteristics and size. Although we had an impressive number of schools, disaggregating our data by risk level and by student race/ethnicity within risk level led to increasingly smaller sample sizes that might have yielded unstable outcomes. A larger sample might have produced more readily interpretable outcomes.

Finally, it is important to consider that our dataset contained information only on student behavioral outcomes. We did not have access to data on intervention fidelity (e.g. CICO use), or teacher and administrator characteristics. Therefore, we cannot assume a causal relationship between CICO use and observed student outcomes. Interpretation of outcomes is also limited by the presence of confounding variables. ODR occur at the intersection of student behavior, teacher interpretation of that behavior, and administrative interpretation of the teacher report. Thus, ODR data potentially contain multiple confounding variables, including teacher race/ethnicity, experience, or possible adult bias. Given the limitations of our dataset, we were unable to include potential confounding variables in our analyses.

Recommendations for Future Research

Given the increasing diversity of the U.S. student population, reducing racial/ethnic disproportionality in discipline outcomes seems critical. Overall, our study showed the importance of taking student race/ethnicity into consideration when designing and evaluating discipline practices. Regular review of discipline data disaggregated by student race could (a) direct school teams' attention to discrepancies in the percentage of students by racial/ethnic group who receive multiple ODR and who gain access to secondary level support and (b) encourage teams to investigate why discrepancies occur and how they could be alleviated.

The outcomes of our study suggest that secondary level support can work to reduce growth in ODRs for students from diverse racial/ethnic backgrounds. However, the timely provision of these supports is critical to ensuring their success. Regular review of discipline data disaggregated by student race might encourage prompt provision of support for students who accumulate multiple ODR early in the year. Equitable access to and culturally responsive provision of secondary level support could increase the success of students from diverse racial/ethnic backgrounds. To deliver culturally responsive behavior support, the literature recommends (a) collection and use of data to determine a student's responsiveness to existing support mechanisms, (b) open discussions of race in data interpretations, and (c) early and culturally specific intervention in the form of additional support for students who are at risk of behavioral failure (Gresham et al., 2000; Hawken et al., 2008; Lane et al., 2009; Skiba et al., 2008). Our study outcomes appear to underscore the urgency of these recommendations to decrease racial/ethnic disproportionality in discipline ou tcomes.

Too often research focuses on data aggregated across all racial/ethnic groups. Disaggregating data by student race/ethnicity appears imperative to identify what works and what does not work for students who have been shown to be disproportionately represented among students with behavioral problems. True effectiveness of any behavioral support practice hinges on its documented effectiveness for students from all racial/ethnic backgrounds. Taken together, the outcomes of our study clearly suggest a need for continuous review of data disaggregated by student race/ethnicity to evaluate the effectiveness of discipline policies as well as intervention outcomes, especially at the middle school level where students develop into young adults with distinct racial/ethnic identities. To help students succeed, schools need to make an effort to create environments where these racial/ethnic identities can grow and be positively acknowledged.

Finally, extant data represent a rich source of information to guide our efforts to create culturally responsive school environments. Future research might avail itself of EDA strategies to identify key variables of interest in establishing linkages between disciplinary practices and racially disproportionate outcomes. For example, it might be valuable to examine ODR patterns across referring teachers, across administrators with varying levels of experience, or across teacher-student dyads of same and different racial/ethnic backgrounds. Similarly, CICO access across types of behavioral violations (e.g. disrespect, fighting, truancy) could be examined to explore for which inappropriate behavior CICO is deemed to be an adequate response. These potential approaches to extant data analysis might help untangle the relationships between students of given racial/ethnic backgrounds engaging in certain behavior, teachers with given characteristics reporting that behavior, and administrators with given characteristics making decisions that affect students' educational success.

References

Aud, S., Fox, M. A., & KewalRamani, A. (2010). Status and trends in the education of racial and ethnic groups. (NCES 2010-015). U.S. Department of Education, National Center for Education Statistics. Washington, DC: U.S. Government Printing Office.

Bohanon, H., Fenning, P., Carney, K. L., Minnis-Kim, M. J., Anderson-Harriss, S., Moroz, K. B., ... Pigott, T. D. (2006). Schoolwide application of positive behavior support in an urban high school: A case study journal of Positive Behavior Interventions, 8, 131-145.

Bradshaw, C. P., Mitchell, M. M., & Leaf, P. J. (2010). Examining the effects of schoolwide positive behavioral interventions and supports on student outcomes. Journal of Positive Behavior Interventions, 12, 133-148..

Bradshaw, C. P., Mitchell, M. M., O'Brennan, L. M., & Leaf, P. J. (2010). Multilevel exploration of factors contributing to the overrepresentation of black students in office discipline referrals. Journal of Educational Psychology, 102, 508-520.

Bradshaw, C., Koth, C., Thornton, L., & Leaf, P. (2009). Altering school climate through school-wide positive behavioral interventions and supports: Findings from a group-randomized effectiveness trial. Prevention Science, 10, 100-115.

Bradshaw, C., Reinke, W., Brown, L., Bevans, K., & Leaf, P. (2008). Implementation of school-wide positive behavioral interventions and supports (PBIS) in elementary schools: Observations from a randomized trial. Education and Treatment of Children, 31(1), 1-26.

Cartledge, G., Singh, A., & Gibson, L. (2008). Practical behavior-management techniques to close the accessibility gap for students who are culturally and linguistically diverse. Preventing School Failure 52(3),29-38.

Cheney, D., Lynass, L., Flower, A., Waugh, M., Iwaszuk, W., Mielenz, C., & Hawken, L. (2010). The Check, Connect, and Expect program: A targeted, tier 2 intervention in the schoolwide positive behavior support model. Preventing School Failure, 54, 152-158.

Cheney, D., Stage, S. A., Hawken, L., Lynass, L., Mielenz, C., & Waugh, M. (2009). A two-year outcome study of the Check, Connect, and Expect intervention for students at-risk for severe behavior problems. Journal of Emotional and Behavioral Disorders, 17, 226-243.

Christenson, S. L., Thurlow, M. L., Sinclair, M. F., Lehr, C. A., Kaibel, C. M., Reschly, A. L., Mavis, A., & Pohl, A. (2008). Check & Connect: A Comprehensive Student Engagement Intervention Manual. Minneapolis, MN: University of Minnesota.

Committee for Children (2010). Second Step: Skills for social and academic success. Seattle, WA: Author.

Coutinho, M. J. & Oswald, D. P. (2000). Disproportionate representation in special education: A synthesis and recommendations journal of Child and Family Studies 9, 135-156.

Crone, D. A., Hawken, L. S., & Horner, R. H. (2010). Responding to problem behavior in schools: The Behavior Education Program (2nd Ed.). New York, NY: Guilford Press.

Crosnoe, R., Johnson, M. K., & Elder, G. H. (2004). Intergenerational bonding in schools: The behavioral and contextual correlates of student-teacher relationships. Sociology of Education, 77, 60-81.

Delpit, L.D. (1992). Education in a multicultural society: Our future's greatest challenge. Journal of Negro Education, 61, 237-249.

Fairbanks, S., Sugai, G., Guardino, D., & Lathrop, M. (2007). Response to intervention: Examining classroom behavior support in second grade. Exceptional Children, 73, 288-310.

Filter, K., Benedict, E. A., Horner, R. H., Todd, A. W., & Watson, J. (2007). Check-in/Check out: A post hoc evaluation of an efficient secondary level intervention for reducing problem behaviors in schools. Education and Treatment of Children, 30, 69-84.

Fletcher, J.M. (2008). Adolescent depression: Diagnosis, treatment and educational attainment. Health Economics 17, 1215-1235.

Glackman, T., Martin, R., Hyman, I., McDowell, E., Bent, V., & Spino, P. (1978). Corporal punishment, school suspension, and the civil rights of students: An analysis of Office for Civil Rights school surveys. Inequality in Education, 23, 61-65.

Gregory, J.F. (1995). The crime of punishment. Racial and gender disparities in the use of corporal punishment in the U.S. Public Schools. Journal of Negro Education, 64, 454-462.

Gresham, F. M., Lane, K. L., & Lambros, K. M. (2000). Comorbidity of conduct problems and ADHD: Identification of 'fledgling psychopaths.' journal of Emotional and Behavioral Disorders, 8, 83-94. doi: 10.1177/106342660000800204.

Harry, B., & Klingner, J. (2006). Why are so many minority students in special education? New York, NY: Teachers College.

Hawken, L. S. (2006). School psychologists as leaders in the implementation of a targeted intervention: The Behavior Education Program (BEP). School Psychology Quarterly, 21, 91-111.

Hawken, L. S., 8z Horner, R. H. (2003). Evaluation of a targeted intervention within a schoolwide system of behavior support. Journal of Behavioral Education, 12, 225-240.

Hawken, L. S., 8z Johnson, S. S. (2007). Preventing severe problem behaviors in young children: The Behavior Education Program. Journal of Early and Intensive Behavior Intervention, 4, 599-613.

Hawken, L. S., Adolphson, S. L., MacLeod, K. S., & Schumann, J. (2009). Secondary-tier interventions and supports. In W. Sailor, G. Dunlap, G., Sugai, & R. Horner, (Eds.). Handbook of Positive Behavior Support (pp. 395-420). New York, NY: Springer.

Hawken, L. S., Pettersson, H., Mootz, J., & Anderson, C. (2005). The Behavior Education Program: A check-in, check-out intervention for students at risk. [DVD]. New York, NY: Guilford Press.

Hawken, L. S., Vincent, C. G., & Schumann, J. (2008). Response to intervention for social behavior: Challenges and opportunities. Journal of Behavioral and Emotional Disorders, 16, 213-225.

Hawken, L., MacLeod, K., & Rawlings, L. (2007). Effects of the Behavior Education Program (BEP) on problem behavior with elementary school students. Journal of Positive Behavior Interventions, 9, 94-101.

Horner, R. H., Sugai, G., Smolkowski, K., Eber, L., Nakasato, J., Todd, A., & Esperanza, J. (2009). A randomized, waitlist-controlled effectiveness trial assessing school-wide positive behavior support in elementary schools. Journal of Positive Behavior Interventions 11, 133-144.

Horner, R. H., Sugai, G., Todd, A.W., & Lewis-Palmer, T. (2005). School-wide positive behavior support. In L. Bambara & L. Kern (Eds.) Individualized supports for students with problem behaviors: Designing positive behavior plans. (pp. 359-390) New York, NY: Guilford Press.

Howell, D. (2002). Statistical methods for psychology. (5th ed.) Pacific Grove, CA: Duxbury.

Kaufman, J. S., Jaser, S. S., Vaughan, E. L., Reynolds, J. S., Di Donato, J., Bernard, S. N., & Hernandez-Brereton, M. (2010). Patterns in office discipline referral data by grade, race/ethnicity, and gender. Journal of Positive Behavior Interventions, 12, 44-54.

KewelRamani, A., Gilbertson, L., Fox, M., & Provasnik, S. (2007). Status and trends in the education of racial and ethnic minorities (NCES 2007-039). Washington, DC: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education.

Klingner, J.K., Artiles, A.J., Kozleski, E., Harry, B., Zion, S., Tate, W., Duran, G.Z., & Riley, D. (2005). Addressing the disproportionate representation of culturally and linguistically diverse students in special education through culturally responsive educational systems. Education Policy Analysis Archives, 13(38), 1-42. Retrieved from http://epaa.asu.edu/epaa/v13n38/.

Krezmien, M. P., Leone, P. E., & Achilles, G. M. (2006). Suspension, race, and disability: Analysis of statewide practices and reporting. Journal of Emotional and Behavioral Disorders, 14, 217-226.

Lane, K. L., Kalberg, J. R., & Menzies, H. M. (2009). Developing school-wide programs to prevent and manage problem behaviors. New York, NY: Guilford.

Lee, J. (2002). Racial and ethnic achievement gap trends: Reversing the progress toward equity? Educational Researcher 31, 3-12.

Lee, J., Grigg, W., & Donahue, P. (2007). The Nation's Report Card: Reading 2007 (NCES 2007-496). National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education, Washington, D.C.

Leinhardt, G., & Leinhardt, S. (1980). Exploratory data analysis: New tools for the analysis of empirical data. Review of Research in Education, 8, 85-157.

Losen, D. J., & Orfield, G. (Eds.). (2002). Racial inequality in special education. Cambridge, MA: Harvard University Press.

March, R. E., & Horner, R. H. (2002). Feasibility and contributions of functional behavioral assessment in schools. Journal for Emotional and Behavioral Disorders, 10, 158-170.

May, S., Ard, W., Todd, A., Horner, R., Glasgow, A., Sugai, G., & Sprague, J. (2005). School-wide Information System (SWIS(c)), University of Oregon, Educational and Community Supports.

McCurdy, B. L., Kunsch, C., & Reibstein, S. (2007). Secondary prevention in the urban school: Implementing the Behavior Education Program. Preventing School Failure, 51, 12-19.

McIntosh, K., & Av-Gay, H. (2007). Implications of current research on the use of functional behavior assessment and behavior support planning in school systems. International Journal of Behavior Consultation and Therapy, 3, 38-52.

McIntosh, K., Campbell, A., Carter, D., & Dickey, C. (2009). Differential effects of a tier 2 behavioral intervention based on function of problem behavior. Journal of Positive Behavior Interventions, 11, 82-93.

McLaughlin, K. A., Hilt, L. M., & Nolen-Hoeksema, S. (2007). Racial/ethnic differences in internalizing and externalizing symptoms in adolescents. Journal of Abnormal Psychology, 35, 801-816.

Muscott, H. Mann, E., & LeBrun, M. (2008). Positive behavioral interventions and supports in New Hampshire: Effects of large-scale implementation of schoolwide positive behavior support on student discipline and academic achievement. Journal of Positive Behavior Interventions, 10, 190-205.

National Research Council (U.S.). Committee on Improving Measures of Access to Equal Educational Opportunity. (2003). Measuring access to learning opportunities. Washington, D.C.: National Academies Press.

Nelson, J. R., Gonzales, J. E., Epstein, M. H., & Benner, G. J. (2003). Administrative discipline contacts: A review of the literature. Behavioral Disorders, 8, 27-37.

Sailor, W., Dunlap, G., Sugai, G., & Horner, R. (Eds.). (2009). Handbook of positive behavior support. In M.C. Roberts (Series Ed) Issues in Clinical Child Psychology Vol. 26 New York, NY: Springer.

Schumann, J., & Burrow-Sanchez, J. (2010). Cultural considerations and adaptations for the BEP. In Crone, D., Horner, R., & Hawken, L. (Eds.), Responding to Problem Behavior in Schools: The Behavior Education Program. New York, NY: Guilford Press.

Shaw, S. R., & Braden, J. P. (1990). Race and gender bias in the administration of corporal punishment. School Psychology Review, 19, 378-383.

Shinn, M., Walker, H., & Stoner, G. (2002). Interventions for academic and behavior problem: Preventive and remedial approaches. Silver Spring, MD: National Association of School.

Skiba, R. J., & Peterson, R. L. (2000). School discipline at a crossroads: From zero tolerance to early response. Exceptional Children, 66, 335-347.

Skiba, R. J., Horner, R. H., Chung, C., Rausch, M. K., May, S. L., & Tobin, T. J. (2011). Race is not neutral: A national investigation of African American and Latino disproportionality in school discipline. School Psychology Review, 40, 85-107.

Skiba, R. J., Michael, R. S., Nardo, A. C., & Peterson, R. (2002). The color of discipline: Sources of racial and gender disproportionality in school punishment. Urban Review, 34, 317-342.

Skiba, R. J., Peterson, R. L., & Williams, T. (1997). Office referrals and suspension: Disciplinary intervention in middle schools. Education & Treatment of Children, 20, 295-316.

Skiba, R. J., Simmons, A. B., Ritter, S., Gibb, A. C., Rausch, M. K., Cuadrado, J., & Chung, C. (2008). Achieving equity in special education: History, status, and current challenges. Exceptional Children, 74, 264-288.

Stage, S., Cheney, D., Flower, A., Templeton, T., & Waugh, M. (in review). A concurrent validity study for a targeted group intervention using an internet-based daily performance report and chart review process using four student behavior constructs: Externalizing problem behavior, internalizing problem behavior, social skills, and academic skills journal of School Psychology.

Stillwell, R. (2010). Public school graduates and dropouts from the common core of data: School year 2007-2008. Washington, DC: National Center for Education Statistics. Available at http://nces.ed.gov/pubs2010/2010341.pdf

Sugai, G., & Lewis, T. J. (1999). Developing positive behavioral support systems. In G. Sugai & T. J. Lewis (Eds.), Developing positive behavioral support for students with challenging behaviors (pp. 15-23). Arlington, VA: Council for Children with Behavioral Disorders.

Sugai, G., Sprague, J. R., Horner, R. H., & Walker, H. M. (2000). Preventing school violence: The use of office discipline referrals to assess and monitor school-wide discipline interventions journal of Emotional and Behavioral Disorders, 8, 94-101.

Swain-Bradway, J., & Horner, R. H. (2010). High school behavior education program (HS-BEP). In D. Crone, R. Horner & L. Hawken (Eds), Responding to problem behavior in schools: The behavior education program. New York, NY: Guilford Press.

Technical Assistance Center on Positive Behavioral Interventions and Supports. (2010). School-wide positive behavior support: Implementers' blueprint and self-assessment. Eugene: University of Oregon. Retrieved from http://www.pbis.org/common/pbisresources/publications/SWPBS_ImplementationBlueprint_vSep_23_2010.pdf

Tobin, T. J. (2006). Use of the Team Implementation Checklist in regular and alternative high schools. Retrieved from http://uoregon.edu/~ttobin/alt_tic.pdf

Todd, A. W., Horner, R. H., & Rossetto Dickey, C. (2010). CICO/SWIS Readiness Checklist. Eugene: University of Oregon, Educational and Community Supports. Retrieved from http://www.swis.org/index.php?page=resources;rid=10111)

Todd, A.W., Kaufman, A., Meyer, G., & Horner, R. H. (2008). The effects of a targeted intervention to reduce problem behaviors: Elementary school implementation of Check-In-Check-Out. Journal of Positive Behavior Intervention, 10, 46-55.

Varela, R. E., Sanchez-Sosa, J. J., Biggs, B. K., & Luis, T. M. (2008). Anxiety symptoms and fears in Hispanic and European American children: Cross-cultural measurement equivalence. Journal of Psychopathology and Behavioral Assessment, 30, 132-145.

Vincent, C. G., Cartledge, G., May, S., & Tobin, T. (2009). Do elementary schools that document reductions in overall office discipline referrals document reductions across all student races and ethnicities? Retrieved from http://pbis.org/evaluation/evaluation_briefs/oct_09.aspx

Vincent, C. G., Tobin, T. J., Swain-Bradway, J., & May, S. (2011). Disciplinary referrals for culturally and linguistically diverse students with and without disabilities: Patterns resulting from school-wide positive behavior support. Exceptionality, 19, 175-190.

Waitoller, F. R., Artiles, A. J., & Cheney, D. A. (2010). The miner's canary: A review of overrepresentation research and explanations. Journal of Special Education, 44, 29-49.

Wallace, J. M., Goodkind, S., Wallace, C. M., & Bachman, J. G. (2008). Racial, ethnic, and gender differences in school discipline among U.S. high school students: 1991-2005. Negro Educational Review, 59, 47-62.

Zayas, L. H., Lester, R. J., Cabassa, L. J., & Fortuna, L. R. (2005). Why do many Latina teens attempt suicide? A conceptual model. American journal of Orthopsychiatry, 75, 275-287.

Zhang, D., Katsiyannis, A., & Herbst, M. (2004). Disciplinary exclusions in special education: A four-year analysis. Behavioral Disorders, 29, 337-347.

Correspondence to Claudia G. Vincent, Educational and Community Supports (ECS) 1761 Alder Street, 1235 University of Oregon, Eugene, OR 97403-123.5; e-mail: clavin@uoregon.edu.

This research was supported in part by U.S. Department of Education Grant No. F1326S030002. Opinions expressed herein do not necessarily reflect the policy of the Department of Education, and no official endorsement by the Department should be inferred.

Claudia G. Vincent and Tary J. Tobin

University of Oregon

Leanne S. Hawken

University of Utah

Jennifer L. Frank

Pennsylvania State University
Table 1

Sample overview by school level

                          Elem (n = 155)  Middle (n = 46)

Locations                      IL: 43.3%        IL: 43.5%
                                MD: 9.0%        MD: 19.6%
                                OR: 9.0%        OR: 17.4%
                                CO: 8.4%         MI: 6.5%
Avg Enrollment                       484              700
Avg ODR/100 Students/Day             .65              .98
Total Students Enrolled           72,778           30,365
Total Students on CICO      1051 (1.44%)      346 (1.14%)


Table 2
Students with 1, 2-5, and 6+ ORD by race and school level

                   Elementary
                    (n = 155)
                            1          2-5           6+           1
                            #      %     #      %     #      %    #

NativeAm                   27    .42    21    .46     9    .65   29
Asian                      89   1.38    64   1.39    14   1.02   43
PaIslander                 22    .34    17    .37     5    .36    5
HispAm                    795  12.29   544  11.82   166  12.07  453
AfrAm                    1104  17.07   926  20.13   302  21.96  715
White                    1770  27.37  1302  28.30   414  30.11  212
NotListed/Unknown        2661  41.14  1727  37.54   465  33.82  724
Total                    6468    100  4601    100  1375    100  U81

                            Middle
                          (n = 46)
                               2-5           6+
                       %         #      %     #      %

NativeAm             .69        27    .69    12    .69
Asian               1.03        53   1.35    11    .63
PaIslander           .12         6    .15     3    .17
HispAm             10.83       452  11.52   Z58  14.83
AfrAm              17.10       851  21.68   124  24.37
White              28.99      1142  29.10   517  29.71
NotListed/Unknown  41.23      1394  35.52   515  29.60
Total                100      3925    100  1740    100


Table 3
Anova Summary Table-Elementary Schools

Source            SS        df   MS       F       P       [eta]2

Between subjects

Race/Ethnicity      64.291   2    32.146   1.648    .194    .009

C1C0 start date    260.394   1   260.394  13.351  <.0005    .035

Racefethnicity x    27.136   2    13.351    .696    .499    .004

CICO start date

Error             7138.528  366   19.504


Table 4
ANOVA Summary Table-Middle Schools

Source            SS        df   MS       F      P     [eta]2

Between subjects

Race/Ethnicity     226.881  2    113.441  3.887  .023    .048

CICO start date    280.370  1    280.370  9.608  .002    .059

Race/ethnicity x   131.389  2     65.694  2.251  .109    .059

CICO start date

Error             4464.849  153   29.182
Gale Copyright:
Copyright 2012 Gale, Cengage Learning. All rights reserved.