Threats of Violence By Students in Special Education

By Kaplan, Sebastian G; Cornell, Dewey G

ABSTRACT:

We compared threats of violence made by K-12 students in special education (120 cases) or general education (136 cases) in schools that were implementing threat assessment guidelines for managing student threats of violence (Cornell, Sheras, Kaplan, McConville, Posey, Levy-Elkon, et al., 2004; Cornell & Sheras, in press). Students in special education made disproportionately more threats, as well as more severe threats, than peers in general education. Students classified as emotionally disturbed (ED) exhibited the highest threat rates. Nevertheless, use of school suspension as a disciplinary consequence for threats was consistent for students in special and general education, and few students were expelled. Our findings support the use of threat assessment to manage threats of violence by students in special education.

Based on its study of school shooting incidents, the Federal Bureau of Investigation (FBI;OToole,2000)advisedschoolsto abandon efforts to develop profiles of potentially violent students and instead to focus intervention efforts on students who communicate explicit threats of violence. Similarly, a joint report of the U.S. secret Service and the U.S. Department of Education recommended that schools establish threat assessment teams (Fein, Vossekuil, Pollack, Borum, Modzeleski, & Reddy, 2002). In response to these recommendations, the Virginia Youth Violence Project atthe University of Virginia developed detailed practice guidelines for schools to use in conducting threat assessments (Cornell & Sheras, in press). These guidelines were successfully field-tested in 35 schools for one academic year (Cornell, Sheras, Kaplan, McConville, Posey, Levy-Elkon, et al., 2004). The purpose of the present study is to compare the threat behaviors of students receiving special education services and students in general education programs in schools that use these threat assessment guidelines.

Students receiving special education services incur a disproportionate number of school discipline infractions (Skiba, Peterson, & Williams, 1997; Wright & Dusek, 1998). For example, Skiba et al. found that 38.6% of a middle school’s office referrals for discipline violations were for students in special education, even though only 15.6% of the school’s population received special education services. Wright and Dusek found that in a sample of 230 discipline referrals for aggression collected over a two-year period in an elementary school, an average of 26% of students in special education had at least one referral for aggression per year versus only 8% of students in general education.

Students with disabilities are also more likely to receive out- of-school suspensions for endangering others and bringing weapons to school than are nondisabled peers (Rose, 1988). The 1994 Gun-Free Schools Act mandated that schools implement a one-year expulsion, with provisions for a shortened exclusionary period on a case-by- case basis, for any student in possession of a weapon on school grounds. Although the mandated federal report does not include the number of students with disabilities who are expelled, it does provide data on the disabil ity status of those who receive a shortened expulsion. During the 2000-2001 school year, students with disabilities represented 28% of the shortened expulsions under the act (U.S. Department of Education, 2003).

Morrison and D’lncau (1997) found that students with disabilities received recommendations for expulsion at nearly twice the expected rate over a two-year period in a school district that implemented a zerotolerance policy. The authors attributed several risk factors to the 158 students recommended for expulsion, including below-average grades and achievement scores and truancy. In particular, records of students with disabilities often identified chronic emotional and family problems beginning at early ages. Although the authors did not conclude that the students with disabilities were unfairly treated under the school’s discipline policy, they stated that these students’ overrepresentation in the sample recommended for expulsion should not come as a surprise, given the academic and emotional challenges they faced.

Furthermore, McFadden, Marsh, Price, and Hwang (1992) reported differences in the administration of punishment to students with and without disabilities for equivalent offenses. A higher proportion of students with a disability (56%) received corporal punishment for fighting than students without a disability (36%). In fact, corporal punishment was the most common consequence for misbehavior by students with disabilities (40%). Students with disabilities (18%) were less likely to receive in-school suspension for defiance of school authority than were nondisabled peers (45%). The authors concluded, “Commission of the most common school offenses would more likely result in corporal punishment for the handicapped and internal suspension for the non-handicapped” (p. 247).

Zero-tolerance policies are intended as a means of protecting students from threatening or potentially dangerous behavior, but in practice such policies can result in harsh punishment for seemingly minor infractions such as accidentally bringing a plastic knife or toy gun to school (Skiba & Peterson, 1999). For instance, a second grader was suspended and sent to an alternative school for one month for bringing to “show and tell” a watch attached to a one-inch pocketknife (Skiba & Peterson, 1999). Another school division implemented a zero-tolerance policy for all use of threatening statements but had to repeal the policy when it led to 50 suspensions in a period of six weeks, largely of students from kindergarten through third grade (Zernike, 2001). The widespread use of zero-tolerance policies poses particular risk for students whose disabling condition might predispose them to engage in impulsive behavior or make rash statements that are interpreted as threats.

Special Education Classification

Among students in special education, several groups appear to experience higher rates of disciplinary violations. Students who suffer from an emotional disturbance are particularly at risk for discipline violations (McFadden et al., 1992; Conroy, Katsiyannis, Clark, Gable, & Fox, 2002) and often receive harsh punishments for their infractions (Skiba et al., 1997). Research has also shown that students who receive special education services for learning disabilities (LD) also have elevated rates of disciplinary infractions (McFadden et al., 1992; Skiba et al., 1997; Sprague, Walker, Stieber, Simonsen, Nishioka, & Wagner, 2001). Children diagnosed with attention deficit-hyperactivity disorder (ADHD), who may receive services under the other health impaired (OHI) classification, have significantly greater rates of disciplinary violations, suspensions, and expulsions than other students (Barkley, Fischer, Edelbrock, & Smallish, 1990; Murphy & Barkley, 1996). Although these studies identify special education classifications with a higher rate of disciplinary violations, no study has investigated the frequency of threats of violence and the associated disciplinary consequences for these threats.

Discipline and Special Education

What is an appropriate response to a student in special education who has made a threat of violence? According to the Individuals with Disabilities Education Act (IDEA Amendments, 1997) and the Office of Special Education Programs (OSEP), disciplinary decisions for students in special education require that “there must be a balanced approach to the issue of discipline of students with disabilities that reflects the need for orderly and safe schools and the need to protect the right of students with disabilities to a free and appropriate public education” (OSEP, 1997).

Difficulties in maintaining a balance between school safety and the student’s right to a free and appropriate public education have led to confusion and frustration for many school administrators in deciding on appropriate disciplinary consequences for students in special education (Skiba, 2002;Taylor& Baker, 2002). A student with a disability cannot receive standard disciplinary consequences for behavior that is a manifestation of his or her disability. Skiba cited the lack of available measures for the purpose of manifestation determination decisions as a flaw in the IDEA disciplinary requirements. Skiba observed that some school authorities feel that IDEA deprives “schools and school districts of tools-school suspension and expulsion-needed to ensure school discipline” (p. 87).

Threat Assessment

Threat assessment is an approach to violence prevention originally developed by the U.S. secret Service based on studies of persons who attacked or threatened to attack public officials (Fein, Vossekuil, & Holden, 1995). Threat assessment has evolved into a standard approach to analyze many different dangerous situations, such as threats of workplace violence, and more recently, school violence (Reddy, Borum, Berglund, Vossekuil, Fein, & Modzeleski, 2001 ). A threat assessment is conducted when a person (or persons) threatens to commit a violent act or engages in behavio\r that appears to threaten an act of violence. Threat assessment is a process of evaluating the threat and the circumstances surrounding the threat to uncover any facts or evidence that indicate the threat is likely to be carried out.

Threat assessment differs from a zerotolerance approach because of its emphasis on the context and meaning of the student’s behavior. For example, under a zero-tolerance approach, a student would be disciplined for any possession of a weapon, regardless of the reason or circumstances of the behavior. Under a threat- assessment approach, however, school authorities would consider the reason why the student had a weapon, the danger posed by the weapon, and what the student intended to do with it. Threat assessment distinguishes toy guns from real guns and unintentional possession of a weapon from use of a weapon to threaten or intimidate someone (Cornell & Sheras, in press). Threat assessment may be particularly relevant for students in special education who make threats of violence, because administrators are guided to make informed disciplinary decisions based on a careful review of the details and context in which the threat occurred. In contrast, a zero-tolerance policy would impose harsh penalties without consideration of the context and meaning of the behavior, and it thus runs the risk of punishing a disabled student for behavior that is a manifestation of his or her disability.

A literature search identified only one study reporting the frequency of threats made by students receiving special education services. Ryan-Arredondo, Renouf, Egyed, Doxey, Dobbins, Sanchez, et al. (2001 ) reported on the implementation of a risk assessment instrument in the Dallas Independent School District. The authors examined the results of 139 threats, of which 27% were made by students receiving special education services. Presumably, the proportion of threats made by students receiving special education services was higher than the proportion of such students in the school population, although this comparison was not reported. Of those students receiving special education services, 42% were classified ED and 34% LD.

The current study addressed four main questions: (1) Do students in special education make threats more frequently than students in general education? (2) Do students in special education make different kinds of threats than students in general education? (3) Do students in special education and general education receive different disciplinary consequences for their threats? (4) How do students in special and general education differ in their postthreat behavior?

Method

Participants

The core sample for this study was obtained from two school divisions participating in the original demonstration project to field-test threat assessment guidelines (Cornell, Sheras, Kaplan, Levy-Elkon, McConville, McKnight, et al., 2004). These school divisions consisted of 35 schools (22 elementary, 6 middle, 4 high, and 3 alternative schools) with a total student population of 16,273 students, of which 71% were Caucasian, 22% African American, and 7% other groups. Approximately 26% of the students were eligible for free or reducedprice meals. The two school divisions together served a small city and the surrounding county in central Virginia with a combined population of 129,000.

As described in detail elsewhere (Cornell, Sheras, Kaplan, Levy- Elkon etal., 2004; Cornell, Sheras, Kaplan, McConville et al., 2004), each school had a threat assessment team consisting of the school principal or assistant principal who led the team, the school’s resource officer or a liaison police officer, a school psychologist, and a school counselor. A threat was defined as any communication of intent to harm someone. Threats could be spoken, written, or expressed through gestures or possession of a weapon. Threats could be made directly to an individual or expressed to third parties. Whenever school authorities learned that a student had threatened to harm someone, the team leader began a threat assessment and documented the case on a standard form (described in a subsequent section).

The 35 team leaders reported 188 threats of violence during the school year. As shown in Table 1, boys made 78% (n = 146) of the threats and girls made 22% (n = 42). In addition, Caucasian (55%) and African American (43%) students made nearly all threats; a small number were made by students of Hispanic (1 %) and other (1 %) ethnic backgrounds.

To ensure a sufficiently large sample of threats from students receiving special education services, we supplemented the core sample with 68 cases collected over the next two years from the two original school divisions and two additional Virginia school divisions that received threat assessment training. These additional divisions consisted of 14 schools, of which 6 (1 high school, 3 middle schools, and 2 elementary schools) participated in reporting of threats. In these school divisions, 70% of the students were Caucasian, 28% were African American, and 2% were from other ethnic backgrounds. Thirty-nine percent of students received free and reduced-price lunch.

After the initial field-test year, schools were not required to document and report all student threats to the researchers, and as a result, some school principals did not participate in the reporting process. For this reason, the threats obtained in the supplemental sample were not included in analyses intended to measure the rate of threats relative to the general school population. The supplemental cases (also in Table 1 ) involved 53 boys and 15 girls, with an ethnic breakdown of Caucasian (68%), African American (28%), Hispanic (3%), and other (1%) ethnic backgrounds.

Measures

Threat ratings. Threats were coded for seriousness and content. On receiving a report of a student threat of violence, the team leader interviewed the student who made the threat and all available witnesses, guided by a standard set of questions. The team leader then prepared a written summary of the threat and completed a checklist of actions taken in response to the threat (Cornell & Sheras, in press). Based on this initial investigation, the team leader made an important distinction between threats that are serious, in the sense that they pose a continuing risk or danger to others, and those that are not serious, because they are readily resolved and do not pose a continuing risk. Threats that were not serious and were readily resolved were classified as transient threats. Serious threats were called substantive threats.

Transient threats are defined as behaviors that can be readily identified as expressions of anger or frustration-or perhaps inappropriate attempts at humor-but that dissipate quickly when the student has time to reflect on the meaning of what he or she has said. The most important feature of a transient threat is that the student does not have a sustained intention to harm someone. In contrast, substantive threats represent a sustained intent to harm someone beyond the immediate incident during which the threat was made. If there is doubt whether a threat is transient or substantive, the threat is regarded as substantive. Substantive threats were further classified as serious if they involved a threat to assault or beat up someone and very serious if they involved a threat to kill, use a weapon, severely injure, or rape someone.

Threat content was coded by the researchers with six categories: threats to kill, hit, shoot, or stab, vague threats in which the intended action was unclear (e.g., “you’d better watch your back”), and other threats (e.g., bomb threats) that did not fall into one of the first five categories.

Follow-up information. Discipline records were available for 184 out of 188 cases in the core sample, but not for the supplemental sample. Because schools used somewhat different discipline categories, we classified infractions into four categories: violence/ weapons (e.g., fighting, battery, weapon possession), disorderly conduct(e.g., disrespect, using obscene language or gestures), bullying (e.g., bullying, threats, harassment), and other (e.g., tardiness, truancy, vandalism, drug/ alcohol use). Schools also reported whether the student received a suspension (or expulsion) from school for the threat.

For cases in the core sample, research assistants conducted follow-up interviews with principals at the end of the school year. Principals were asked to rate each student’s behavior as improved, about the same, or worse after the threat. Principals provided ratings on 94% of cases, omitting cases involving students who had moved, transferred, or for some other reason left school. Principals also provided ratings (improved, same, or worse) of the students’ relationship with their victims following the threat in 67% of cases.

TABLE 1

Demographic Information

Procedure

All schools received approximately six hours of training on the threat assessment guidelines (Cornell, Sheras, Kaplan, Levy-Elkon, et al., 2004; Cornell, Sheras, Kaplan, McConville, et al., 2004). A research assistant assigned to each school provided consultation on the guidelines throughout the school year. School principals reported cases by completing an electronic form at a secure Web site maintained by the researchers. This form served the dual purpose of providing schools with printed documentation of their response to a student threat and informing researchers of a new case so they could follow up with the principal. The Web site form collected demographic information (i.e., age, gender, grade, race) on the student who made the threat and the intended threat victim and provided space for a description of the threat incident and the classification of the threat as transient or substantive. The form also presented a checklist of actions taken in response to the threat (e.g., suspending the student, contacting parents).

For \the core sample, the research assistant conducted two follow- up interviews with the principals-one at the end of the school year and another the following fall. The average follow-up period from the date of the threat incident to the interview with the principal at the end of the school year was 148 days (range 3 to 282 days). The second follow-up interview occurred an average of 424 days after the threat incident. In the follow-up interviews, researchers asked the school principals to describe their response to the threat incident, whether the student carried out the threat, and whether the student’s relationship with the threat recipient was improved, about the same, or worse than prior to the threat. They were asked to rate the student’s overall behavior at school after the incident as improved, about the same, or worse than prior to the threat.

Results

As shown in Table 1, preliminary analyses found no significant differences between the core and supplemental samples in gender, race, special education status, special education classification, and school level (elementary, middle, and high). Table 1 also contains demographic information for the core, supplemental, and full samples. For the full sample, males committed 78% of threats versus 22% by females. Caucasian (58%) and African American (40%) students committed nearly all threats, with only 2% committed by Hispanic students and 1% by students of other ethnic backgrounds. Grade placement of students committing threats ranged from kindergarten to twelfth grade, with 44% occurring in elementary school, 34% in middle school, and 23% in high school.

Students in general education programs made 53% of threats versus 47% by students in special education. Of those students receiving special education services who committed threats, one-half were classified as ED and onefourth as LD, with the remaining students having OHI (16%) or other (8%) classifications. Those students with other classifications included six with mental retardation, two receiving solely speech and language services, one with autism, and one with developmental delay.

Threat Rates of Students in General and Special Education Programs

The first question investigated whether students in special education exhibited higher threat rates than general education students. These analyses were conducted on the core sample because it represented all of the threats reported to the participating schools for one school year and because data were available for the size of special education population in these schools. Students in special education committed 49.5% of threats versus 50.5% by students in general education.

Students receiving special education services did not make threats at similar rates. Students with an ED classification were most likely to make threats relative to other students in special education programs. Specifically, although students with an ED classification made up only 10% of the special education population in these schools, they accounted for 50% of the threats made by the special education population. In contrast, students with an LD classification made up 37% of the special education population, yet they only accounted for 23% of the threats made by students receiving special education services.

Students with OHI classifications made up 14% of the special education population and accounted for a similar percentage (17%) of the threats made by students in special education. Students with classifications other than ED, LD, and OHI committed disproportionately fewer threats than would be expected; they constituted 39% of all students receiving special education services, but committed only 11 % of the threats made by students in special education.

Comparisons with full school enrollment of the four special education classification groups also revealed disproportionate percentages, particularly for students receiving ED services. Students with an ED classification constituted only 2% of the student enrollment but made one-fourth (46 of 188) of the threats in the core sample. The percentage of students receiving LD services that made threats (21 of 188, 11%) was closer to their proportion within the student population (6%). Students classified as OHI made 9% (16 of 188) of threats and comprised 2% of the student enrollment. The number of students receiving services for other disabling conditions who made threats (10 of 188, 7%) was consistent with the proportion of these students in the core sample (5%).

Of the 188 reported threats, school officials judged 70% to be transient and 30% substantive. Table 2 shows the threat rate for general and special education students, as well as for ED, LD, and OHI classifications, in threats per 1,000 students. Too few students with other classifications were available to enable calculation of a threat rate. We compared threat rates for students in special education using chi-square and z-test formulas from Glass and Hopkins (1996). Students in special education exhibited a significantly higher total threat rate than general education peers, z (1, n = 188) = 4.33, p

There were no statistical differences in threat content between students in general and special education, X^sup 2^ (5, n = 169) = 2.81, p > .70. Overall, the most frequent threat was a threat to hit (41 %), followed by an unspecified, vague threat (15%, e.g., “I’m going to hurt you”) and a threat to kill (15%).

TABLE 2

Threat Rates

Students receiving ED services exhibited significantly higher total threat rates than students receiving LD or OHI services, X^sup 2^ (2, n = 83) = 17.30, p.

In follow-up to the threat rate differences between students in general and special education, we conducted a hierarchical logistic regression analysis to examine whether special education status improved the prediction of principal threat ratings beyond demographics and threat content (see Table 3). At Step 1 we entered student age, gender, race, and threat content, which accounted for 19% of the variance in principal ratings. Age and gender were both statistically significant predictors in this model. At Step 2, special education status accounted for an additional 4% of the variance.

Disciplinary Infractions

The next question considered whether students in special education who made threats committed more disciplinary infractions throughout the school year than students in general education. These analyses were limited to the core sample because disciplinary data were not available for the supplemental sample.

A multivariate analysis of variance (MANOVA) comparing students in general and special education on the four disciplinary categories was statistically significant, f (3) = 5.59, p

TABLE 3

Hierarchical Logistic Regression Analysis for Principal Threat Ratings

A follow-up multivariate analysis of covariance (MANCOVA) tested differences between general and special education students on disciplinary infractions for the 2001-2002 school year using age, threat type, gender, and race as covariates. As shown in table 5, special education students exhibited significantly higher violent and disorderly conduct infractions. General and special education students no longer differed in bullying infractions.

The next analyses considered whether schools administered more severe disciplinary consequences to students in special education than students in general education within the full sample. No statistical differences existed in school use of suspension between the core and supplemental samples. School expulsions were not analyzed because there were only three cases. Students receiving special education services were about as likely to be suspended from school for makinga threat (36%) as students in general education (31%). For those students who received a suspension, we conducted a two-step hierarchical regression analysis to determine whether special education status would predict length of suspension beyond age, gender, race, and seriousness of threat (transient or substantive). As shown in Table 6, age and principal ratings of threat severity were statistically significant predictors of length of suspension at Step 1. Mean length of suspension for substantive threats was 4.7 days (SD = 3.35) versus 2.5 days (SD = 2.45) for transient threats. However, special education status did not significantly con\tribute to the prediction of number of days suspended.

Postthreat Behavior

The next analysis compared behavioral changes of threat perpetrators after the threat incident. School officials were more likely to rate students in general education (53%) than in special education (33%) as exhibiting improved behavior, and more likely to rate special education (28%) than general education (8%) students as displaying worse behavior following the threat, X^sup 2^ (2, n = 188) = 13.6, p

We then conducted a two-step hierarchical regression analysis to determine whether special education status was a significant predictor of postthreat behavior beyond student age and threat severity. At Step 1 we entered age and threat severity, which together significantly predicted principal ratings of postthreat behavior, F (2) = 3.87, p

TABLE 4

Comparison of Discipline Infractions Between Students in General and Special Education

TABLE 5

Comparison of Discipline Infractions Between Students in General and Special Education Controlling for Age, Gender, Race, and Threat Type

Another important question related to postthreat behavior concerns whether any of the threats were actually carried out. For the core sample, we conducted interviews with school principals during the final weeks of the school year and during the following fall. According to the school principals, none of the 188 threats of violence were carried out (see Cornell, Sheras, Kaplan, McConville, et al., 2004 for further information).

TABLE 6

Hierarchical Regression Analysis for Comparison of School Use of Suspension for Threats Made by Students in General Versus Special Education

Discussion

Threats of violence appear to be more prevalent among students in special education than general education. Students in special education made nearly half of all threats reported to school principals, even though students in special education represented only 17% of school enrollment. It should be noted that these were threats that came to the attention of school authorities and do not represent all threats that might have occurred. Although no previous research has compared students in special and general education on threats of violence, our findings are consistent with those of previous reports that find elevated rates of disciplinary infractions among students in special education (Skiba et al., 1997; Wright & Dusek, 1998). These results support further attention to the issue of threats of violence by students who receive special education services.

Students receiving ED services made more threats than any other group. The high rate of threatening behavior by these students is not surprising in light of the criteria used to identify a student as eligible for services under this category. The federal definition of an emotional disturbance recognized by IDEA involves a condition that is present over a long period of time, adversely affects academic performance, and involves one of the following:

* An inability to learn that cannot be explained by intellectual, sensory, or health factors.

* An inability to build or maintain satisfactory relationships with peers and teachers.

* Inappropriate types of behavior or feelings under normal circumstances.

* A general pervasive mood of unhappiness or depression.

* A tendency to develop physical symptoms or fears associated with personal or school problems (Hallahan & Kauffman, 2000, p. 250).

Students receiving services for an ED classification likely experience relationship difficulties and interpersonal conflicts, and they may use inappropriate strategies for dealing with conflicts such as threatening others. As a result, IEP teams could interpret a threat of violence as a symptom of the child’s emotional disturbance.

However, schools must balance the rights of a student classified as ED with school safety. An IEP team has additional options beyond standard disciplinary techniques if a threat is determined to be a manifestation of a student’s disability. A school can place a student with a disability in an interim alternative educational setting (IAES) for up to 45 days if the student possesses weapons or drugs or if “substantial evidence” exists that the student is a danger to self or others (Skiba, 2002). Schools may place the student in an IAES without parental consent in response to weapons or drug violations. However, the school must obtain parental consent for an IAES placement for dangerousness, initiate an expedited due process hearing, and petition a hearing officer (Skiba, 2002). Parents also have the right to appeal any decision that constitutes a change of placement for students with disabilities, although the change in placement can proceed during the appeals process. Furthermore, the school may procure a court injunction to remove a student who is deemed dangerous and whose parents refuse to comply with the removal process, regardless of IDEA protection (see Skiba, 2002, for review).

The high rate of reported threats by students with ED classifications could be a result of their comparatively high rate of disciplinary violations. Students may be more concerned about a threat from a student with an ED classification than a student in general education without a similar history of infractions or peer conflict. Threat recipients or witnesses might perceive a student with an ED classification as more likely to carry out a threat and therefore might be more likely to report the threat to school authorities. Similarly, teachers may feel less able to manage a threat from a student with a history of disciplinary problems.

Students in special and general education exhibited no significant differences in what they threatened to do. This finding is important in light of the significantly higher rates of substantive threats for students in special education. The threat assessment guidelines direct school authorities to place more weight on the context and meaning of the threat than the content of the threat (Cornell & Sheras, in press). For example, threats to kill or shoot someone were frequently judged to be transient threats (Cornell, Sheras, Kaplan, McConville et al., 2004), despite the extreme content of the threat, if it was clear from the context that the student did not mean to carry out the threat.

The guidelines indicate that if the context or meaning of the threat is not clear, a threat should be classified as substantive. Cornell, Sheras, Kaplan, McConville et al. (2004) found that 30% of the 188 threats collected during the 2001-2002 school year were substantive and required more extensive intervention and follow-up. The higher rate of substantive threats by students in special education suggests that school authorities took these threats more seriously. Perhaps students receiving special education services had a more extensive history of violent behavior that increased the likelihood of a principal judging their threats to be substantive. For a more detailed discussion on how schools responded to both transient and substantive threats, please see Cornell, Sheras, Kaplan, McConville et al. (2004).

Students in special education who made threats committed more disciplinary infractions over the course of the 2001-2002 school year than peers in general education. Differences between the two groups also existed for infractions involving violence and weapons, disorderly conduct, and bullying. These results are again consistent with the Skiba et al. (1997) study involving office referrals. However, future research could compare disciplinary infractions between students in special education who do and do not make threats to assess whether the threat group represents a more challenging cohort within the student population.

In light of previous studies, it was surprising that students receiving special education services were not more likely to incur an external suspension than students in general education, or that the lengths of suspensions did not differ between groups. Several studies have identified a disproportionate number of students in special education receiving suspensions (Cooley, 1995; Leone, Mayer, Malmgren, & Meisel, 2000) and other harsh consequences such as corporal punishment (McFadden et al., 1992), even for relatively similar infractions such as endangering others and weapons violations (Rose, 1988). Principals in our study did not appear to apply disproportionately harsh consequences to students in special education for behavior comparable to that of general education students. It is possible that with the threat assessment guidelines, the principals were able to avoid the disproportionate disciplinary consequences found in other studies.

Study Limitations

This study was limited to a sample of schools that were implementing a new procedure for managing student threats of violence. There was no comparison group of schools using a different procedure, so it is not possible to conclude that the outcomes observed in this study were attributable to the use of threat assessment guidelines. The original study was a demonstration project to field-test threat assessment guidelines and show that this approach was a viable procedure and therefore should be followed up with a controlled study (Cornell, Sheras, Kaplan, McConville et al., 2004). Because the schools were participating in the field-test project, it was possible to gather data that otherwise would not be available on threats of violence by students in special education.

Any study of student threats is limited by the nature of threat reporting. This study only examined threats that were reported to school authorities, and undoubtedly there are student threats that never come to the attention of school personnel. Cornell and Loper (1996) reported results from a survey of 10,909 students (g\rades 7, 9, and 11 ) in which more than one-fourth of students replied “yes” to the statement “Someone threatened to hurt you at school in the past 30 days.” Singer and Flannery (2000) found that more than onequarter of elementary school students and more than one-third of high school students reported threatening someone within the past year. It would be useful in future studies to study the incidence of student threats and the distinguishing characteristics of threats that are reported to school authorities. It would also be useful to gather more information on teacher and student perceptions of threats and how they judge the seriousness of a threat.

Conclusion

Can a threat assessment approach reduce the incidence of exclusionary discipline practices and disproportionately harsh discipline among students receiving special education services? This study cannot provide a conclusive answer to this question, but the results suggest that threat assessment is worthy of further study. It is possible that structured guidelines that emphasize the context and meaning of a threat over the content of the threat may be helpful to school authorities in responding to threats by students in special education.

It may be particularly useful to compare schools employing threat assessment with schools following a zero-tolerance approach. Skiba and Peterson (2000) cautione’d that a zero-tolerance environment in schools would clash with IDEA principles and provisions. In particular, IDEA emphasizes positive behavioral interventions for disruptive behavior and increased instructional inclusion of special education students, whereas a zerotolerance policy takes the opposite approach. Skiba and Peterson argued that “without general reform of school discipline practice, increased instructional inclusion for students with emotional and behavioral problems may lead to increased exclusion when those students engage in disruptive behavior in general education settings” (p. 340).

REFERENCES

Barkley, R. A., Fischer, M., Edelbrock, C. S., & Smallish, L. (1990). The adolescent outcome of hyperactive children diagnosed by research criteria: I. An 8-year prospective follow-up study. Journal of the American Academy of Child and Adolescent Psychiatry, 29, 546- 557.

Conroy, M. A., Katsiyannis, A., Clark, D., Cable, R. A., & Fox, J. J. (2002). State office of education practices implementing the IDEA disciplinary provisions. Behavioral Disorders, 27, 98-108.

Cooley, S. (1995). Suspension/expulsion of general and special education students in Kansas: A report to theKansas state board of education. Topeka: Kansas State Board of Education.

Cornell, D. C., & Loper, A. B. (1996). High risk behavior in Virginia schools: A survey of middle and high school students. Charlottesville: Curry School of Education, University of Virginia.

Cornell, D. C., & Sheras, P. L. (in press). Guidelines for responding to student threats of violence. Longmont, CO: Sopris West.

Cornell, D., Sheras, P., Kaplan, S., Levy-Elkon, A., McConville, D. McKnight, L., & Posey, j. (2004). Guidelines for responding to student threats of violence: Field test of a threat assessment approach. In M. J. Furlong, P. M. Bates, D. C. Smith, & P. M. Kingery (Eds.), Appraisal and prediction of school violence: Methods, issues, and contexts (pp. 11-36). Hauppauge, NY: Nova Science.

Cornell, D. G., Sheras, P. L., Kaplan, S. C., McConville, D., Posey, J., Levy-Elkon, A., et al. (2004). Guidelines for student threat assessment: Field-test findings. School Psychology Review, 33, 527-546.

Fein, R. A., Vossekuil, F., & Holden, G. A. (1995). Threat assessment: An approach to prevent targeted violence. National Institute of Justice: Research in Action, 1-7 (NCJ 155000). Retrieved October 6, 2005, from http:///www.ncjrs.org/ txtfiles/ threattxt

Fein, R. A., Vossekuil, F., Pollack, W. S., Borum, R., Modzeleski, W., & Reddy, M. (2002). Threat assessment in schools: A guide to managing threatening situations and to creating safe school climates. Washington, DC: U.S. secret Service and U.S. Department of Education.

Glass, G. V, & Hopkins, K. D. (1996). Statistical methods in education and psychology. Boston: Allyn and Bacon.

Hallahan, D. P., & Kauffman, J. M. (2000). Exceptional learners: Introduction to special education (8th ed.). Boston: Allyn and Bacon.

Individuals with Disabilities Education Act (IDEA) Amendments of 1997. Public Law 105-17.

Leone, P. E., Mayer, M. J., Malmgren, K., & Meisel, S. M. (2000). School violence and disruption: Rhetoric, reality, and reasonable balance, focus on Exceptional Children, 33, 2-20.

McFadden, A. C., Marsh, G. E., Price, B. J., & Hwang, Y. (1992). A study of race and gender bias in the punishment of handicapped children. Urban Review, 24, 239-251.

Morrison, G. M., & D’lncau, B. (1997). The web of zero- tolerance: Characteristics of students who are recommended for expulsion from school. Education and Treatment of Children, 20, 316335.

Murphy, K. & Barkley, R.A. (1996). Attention deficit hyperactivity disorder in adults: Comorbidities and adaptive impairments. Comprehensive Psychiatry, 37, 393-401.

Office of Special Education Programs (OSEP). (1997). Initial disciplinary guidance related to removal of children with disabilities from their current educational placement for ten school days or less. (OSEP Memorandum 97-7). Washington, DC: Author.

O’Toole, M. E. (2000). The school shooter: A threat assessment perspective. Quantico, Virginia: National Center for the Analysis of Violent Crime, Federal Bureau of Investigation.

Reddy, M., Borum, R., Berglund, J., Vossekuil, B., Fein, R., & Modzeleski, W. (2001). Evaluating risk for targeted violence in schools: Comparing risk assessment, threat assessment, and other approaches. Psychology in the Schools, 38, 157-172.

Rose, T. L. (1988). Current disciplinary practices with handicapped students: Suspensions and expulsions. Exceptional Children, 55, 230-239.

Ryan-Arredondo, K., Renouf, K., Egyed, C., Doxey, M., Dobbins, M., Sanchez, S., et al. (2001). Threats of violence in schools: The Dallas Independent School District’s Response. Psychology in the Schools, 38, 185-196.

Skiba, R. J. (2002). Special education and school discipline: A precarious balance. Behavioral Disorders, 27, 81-97.

Skiba, R. J., & Peterson, R. L. (1999). The dark side of zero tolerance: Can punishment lead to safe schools? Phi Delta Kappan, 80, 372-382.

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., Peterson, R. L., & Williams, T. (1997). Office referrals and suspension: Disciplinary intervention in middle schools. Education and Treatment of Children, 20, 295-315.

Singer, M. I., & Flannery, D.J, (2000).The relationship between children’s threats of violence and violent behaviors. Archives of Pediatrie and Adolescent Medicine, 154, 785-790.

Sprague, J., Walker, H. M., Stieber, S., Simonsen, B., Nishioka, V, & Wagner, L. (2001). Exploring the relationship between school discipline referrals and delinquency. Psychology in the Schools, 38, 197-206.

Taylor, J. A. & Baker, R. A. (2002). Discipline and the special education student. Educational Leadership, 59(4), 28-30.

U.S. Department of Education, Office of Elementary and secondary Education and Planning and Evaluation Service. (2003). Report on state/ territory implementation of the Gun-Free Schools Act: School year 2000-2001. Author.

Wright, J. A., & Dusek, J. B. (1998). Compiling school base rates for disruptive behaviors from student disciplinary referral data. School Psychology Review, 27, 138-147.

Zernike, K. (2001, May 17). Crackdown on threats in schools fails a test. New York Times. Available from http://www.nytimes.com/ref/ membercenter/nytarchive.htm

Sebastian G. Kaplan & Dewey C. Cornell

University of Virginia

NOTES

The authors, with the Programs in Clinical and School Psychology, Curry School of Education, University of Virginia, thank Amherst County, Albemarle County, Appomattox County, and Charlottesville City schools for their participation in this study and colleagues on the Virginia Youth Violence Project for their assistance throughout this project. This study was supported by a grant from the Jessie Ball DuPontfund.

Address inquiries concerning this article to Sebastian G. Kaplan, Programs in Clinical and School Psychology, Curry School of Education, University of Virginia, 405 Emmet Street, Charlottesville, Virginia 22904-4270. Electronic mail may be sent to sgk4j@virginia. edu.

Copyright Council for Children with Behavioral Disorders Nov 2005