The Impact of Disruptive Students in Wisconsin School Districts
In 2010-2011, more than 48,000 Wisconsin students were suspended. The disruptive behavior leading to these suspensions is detrimental to teachers, school cultures, and ultimately, student learning. Reducing suspension rates in Wisconsin school districts with high numbers of disruptive pupils can substantially increase achievement levels in those districts. An analysis of suspension rates in Wisconsin shows that decreasing those rates by five percentage points would yield an almost five percentage point increase in math proficiency, and a three and one-half percentage point increase in reading proficiency on the Wisconsin Knowledge and Concepts Exam.
In other words, reducing disruptive behavior can yield substantial achievement gains for Wisconsin pupils.
This report reviews existing research on the link between student disruption and academic achievement, reviews current Wisconsin statues and practices regarding student behavior, includes comments from a discussion with teachers from the state’s largest school district, and uses data from both the Department of Public Instruction and from the National Center for Education Statistics to test several hypotheses. The finding that student behavior affects student achievement at the school district level is both intuitive and well-supported by evidence.
The findings are particularly interesting because the other factors that significantly affect achievement in Wisconsin districts, such as the socioeconomic makeup of the student population, cannot be readily addressed in the ways that student behavior can.
Ultimately, this report concludes that Wisconsin must honor its commitment to make a public education available to all of its students, but must not do so at the expense of the vast majority of pupils who do not engage in disruptive behaviors. Similarly, teachers must be supported and allowed to teach in an environment where their focus can be on student learning, not discipline.
The formal recommendations of this report include supporting and strengthening ongoing efforts to instruct teachers on how to deal with problem students, and state efforts to bring evidence-supported strategies for disruptive students to Wisconsin schools. In addition, strategies should be pursued to ensure that chronically disruptive pupils are permanently removed from regular classrooms, perhaps with an increased use of virtual schools. Perhaps most important, Wisconsin must pay greater attention to this issue because doing so can improve student outcomes as well as the overall work and learning environment of teachers and students.
Disruptive students in Wisconsin classrooms make it difficult for other students to learn and difficult for teachers to teach. Addressing this problem can have a very real and positive effect on student performance.
Imagine being a teacher charged with the responsibility of ensuring that the students in your classroom receive an education that prepares them to become productive adults. Beyond that, imagine having to deal with a small number of students who continually inhibit your ability to fulfill that responsibility. What do you do? After all, according to the state constitution, every child -- disruptive or not -- has a right to a public education:
“The legislature shall provide by law for the establishment of district schools, which shall be as nearly uniform as practicable; and such schools shall be free and without charge for tuition to all children between the ages of 4 and 20 years”;1
But what happens when the behavior of one child infringes on the ability of other children to obtain an adequate education? Does the offending student forfeit his or her right to a public education, or must a school district do the best it can under difficult circumstances, even if it means well-behaved students receive instruction of a diminished quality? Clearly, it takes a skilled teacher to manage any classroom. Day-to-day discipline problems inevitably arise in every school. But what happens when disruptive behavior confounds the ability of even the most skilled teacher?
The following report seeks to answer the research question: What is the impact of disruptive students on academic outcomes in Wisconsin? The question will be answered by:
- Defining what is meant by a disruptive student;
- Explaining the data available on disruptive students in Wisconsin; and
- Using this data to test several hypotheses on the relationship between indicators of disruptive students and the academic performance of the broad student population
The reasons for suspecting that student behavior negatively impacts student achievement are grounded in existing research and logic. The connection between individual student behavior and lowered academic performance is well-established (Skiba & Rausch 2004).2 Basically, students who are suspended suffer academically because of missed class time and being ostracized. Logically, removing these students from the classroom for any period of time is unlikely to help their education because less instruction means less learning.
Generally, findings of the UCLA-based Civil Rights Project -- an organization dedicated to renewing the civil rights movement while examining the effects of school discipline policies -- and similar studies are used to justify scaling back zero-tolerance behavior policies that raise suspension rates in schools. The argument is that students most often suspended are disproportionately members of other sub-groups with well-documented educational struggles, including minority and socioeconomically disadvantaged groups.
More recently, attention has been given to the effect disruptive students have on other students. Education researchers Talisha Lee et. al. (2011) use data on almost 300 Virginia high schools to conclude that high suspension rates are correlated with increased high school dropout rates.3 Put another way, schools with behavioral problems as demonstrated by high suspension rates also have high dropout rates. The main weakness of the Lee study is the possibility that dropouts are simply the students who were suspended. In other words, disruptive behavior may not be affecting the school as a whole. One way to verify whether this is the case is to gauge the stability of suspension rates by school or district. If suspension rates are consistent annually, they are likely a good indicator of the general level of behavioral problems in a school.
The specific ways a disruptive student might hurt the learning ability of other students are fairly self-evident. One effect might be decreased teacher effectiveness. Studies and logic suggest that teacher quality is a major determinant of student learning.4 A teacher who spends large chunks of his or her time dealing with student discipline is not spending time on instruction. Inevitably, other students in a disruptive environment will suffer.
For example, education researcher Catherine Ennis (1996) interviewed several urban teachers and found that they regularly chose to eliminate from the curriculum or avoid discussion on topics that caused students to be disruptive in the classroom.5 While conflict may be prevented, non-disruptive students miss out on the potential lesson.
The findings of Philip Garner (1993) lend credibility to Ennis’ conclusions. Garner interviewed disruptive students who told him the reasons for their behavior were generally tied to disliking a teacher’s instructional style, personality or the subject that was being taught.6
Another way that disruptive students might hurt the learning ability of their classmates is through peer pressure. Students exposed to poor behavior by their peers may be more likely to engage in such behavior themselves. In other words, a student who might otherwise be entirely interested in schoolwork is actively engaged in disrupting a class simply because a peer is doing the same.
A less extreme factor is the effect of a distracting presence in a classroom. Even if a teacher is able to continue to teach and even if a student does not join the disruptive behavior, one or both may be distracted from the lesson because of the misbehaving student.
The most important and most abstract consequence of misbehaving students is their potential to serve as a roadblock to the building of a successful school and school district culture. Short & Greer (2002) define school culture broadly as the norms within a school that can be influenced by a school’s teachers and principal.7 Importantly, education researchers have found that a positive school culture is closely related to school performance (Alsbury 2008).8 Generally, if the accepted norms of a school include things such as order, respect and accountability, then academic achievement can be expected to increase.
Misbehaving students undermine the basic norms that are conducive to school success. A building or district that is plagued with disorder, significant teacher and principal time-demands for activities unrelated to student learning, and constant stress caused by numerous disruptions will struggle to build a successful school culture.
The introduction to this report establishes a reasonable basis for thinking that disruptive students negatively affect student learning. Before exploring data to test specific hypotheses, it is necessary to understand what is meant by a disruptive student, and just as important, what Wisconsin schools and districts are empowered to do about them.
What is meant by “disruptive” student?
A simple intuitively satisfying definition of a disruptive student is any K-12 student who violates the rules of a school or classroom. It is presumed that school districts, schools and teachers make rules of behavior for the purpose of creating a quality learning environment. Simply, classroom rules are not punitive.
There is no shortage of academic research examining why students engage in disruptive behavior. Research blames misbehavior on such factors as boredom, mental health issues, medical conditions such as Attention Deficit Disorder, stressful home lives and peer pressure.9 No doubt all of these factors and more contribute to why an individual student violates the rules of a classroom or school. But for purposes of this paper, disruptive students can be divided into two broad categories:
- Those with special needs.
- Those without special needs.
Students with special needs specifically means public school students with an Individual Educational Program (IEP). According to the Department of Public Instruction (DPI), about 14% of all public school students in Wisconsin have an IEP. Table One lists the breakdown of disabilities afflicting Wisconsin students.
Table 1 – Wisconsin Students with Disabilities – 2011-12
The distinction between disruptive students with special needs and those without special needs is important because schools and districts already have strict guidelines for dealing with special needs pupils. If a classroom is negatively affected by a disruptive student suspected of having special needs, there is a course of action that teachers are expected to take.10 First, a teacher makes a referral to their principal or special education director explaining why he or she suspects a student is in need of special education. The principal then has 15 days to decide if more tests are needed to determine the special needs status of a pupil. If more tests are conducted and a special need is diagnosed, an IEP is created and the child is sent to a school or classroom where the needs explained in the IEP can be satisfied.
The disruptive students whose behavioral problems are addressed via suspension rather than a referral for special services are the focus of this report. In practice, there is significant overlap between the two groups since special needs students certainly do engage in behavior that results in suspension. In 2011, in fact, about 41% of all suspension-creating incidents were committed by students with special needs. So, it is important to be clear that there are many Wisconsin pupils with discipline problems that can be and are addressed through the IEP process.
In 2010-11, more than 95,000 incidents in Wisconsin public schools resulted in suspensions. The vast majority of those incidents, 91,550, were not weapon or drug related. Presumably then, they were the result of misbehavior that fell below the threshold of needing law enforcement. The 5,284 incidents that were weapon or drug related presumably involved law enforcement.
For the purposes of this paper, suspension incidents will be used as a proxy for gauging the level of student disruption in a school district. While it is possible that suspension rates do not paint a complete portrait of the level of disruption in a district, state statutes give an idea of the types of incidents that result in suspensions.
Under Wisconsin state statute 120.13 (1)(b) pupils can be suspended for:
- Violating rules of conduct developed and approved by the school board.
- Threats against school property.
- Behavior that threatens the “property, health, or safety of others.”
- Possession of a firearm.
When a student is suspended, the student and parent are notified of the length and reason for the suspension. A suspension can generally last up to five days and is set at the discretion of the school. Parents can appeal a suspension to a school administrator, and if the appeal is won, the suspension is expunged from that child’s record. In 2011, a total of 48,276 pupils, or 5.54% of the total state student public school population, were suspended in Wisconsin. Given that the total number of suspensions in 2011 was more than 95,000, many students were suspended more than once.
What can Wisconsin districts do about disruptive students?
The most direct and extreme response to a disruptive student is expulsion. According to the DPI, in the 2010-2011 school year a total of 1,123 students were expelled from school. Under Wisconsin law, a student can be expelled for making threats, endangering the safety of others, or, if over the age of 16, by being chronically disruptive.11 However, expulsion is an extremely involved process for a school district. The district must give written notice of an expulsion hearing to parents, actually hold the hearing, and the parent and student have the right to contest the hearing. Further, the student and family may appeal the decision first to the State Superintendent, and second to county circuit court.
Notably, the due process provisions available to expelled students did not exist before the U.S. Supreme Court’s 1975 Goss v. Lopez decisions. That lawsuit, which originated from a group of Ohio students seeking to have expulsions expunged from their records, concluded that under the 14th amendment suspended students have a right to due process.12
However, there are other actions short of expulsion and suspension that a school district can take to try to minimize the negative impact of a disruptive student on student learning. Most importantly, teachers are empowered to remove a pupil from a classroom for misconduct.
Under Wisconsin state statute 120.13 (1)(a) each school district develops a specific code of classroom conduct. If a teacher determines that a student is violating the code of conduct, he or she can immediately remove that student from the classroom. In addition under Wisconsin statute 118.164 (2) a teacher may remove a student who is:
“…dangerous, unruly or disruptive or exhibits behavior that interferes with the ability of the teacher to teach effectively…”
On the surface, this suggests that teachers already have complete control over the behavior of pupils in their classroom. If a student is misbehaving, a teacher must simply remove that student from the classroom. Under state law, after a student is removed from the classroom the student may either be sent to an alternative school, a different classroom within the school, or returned to the classroom if:
“after weighing the interests of the removed pupil, the other pupils and the teacher, the school principal or his/her designee determines that re-admission to the class is the best or only alternative.”13
In practice, this means that if an alternative school does not exist, it is entirely possible that disruptive students are simply moved to another classroom where they can continue to be disruptive, or returned to their original classroom for lack of a better alternative.
For example, a 2007 report on disruptive behavior in the Milwaukee Journal Sentineldescribed how disruptive students in the Milwaukee Public Schools (MPS) are moved annually from school-to-school; a process one MPS principal called “the dance of the lemons.”14 In other words, the disruption caused by problem pupils often just moves from school-to-school; the actual problem is never addressed.
In Madison, however, the local teachers union has taken specific steps to try to remove certain disruptive students from the classroom permanently. It is the policy of Madison Teachers Inc. to seek a restraining order against students who either threaten to physically harm, or actually do physically harm teachers.15 16 Actual legal action against threatening students seems to be the exception, however, and not the norm. According to the Wisconsin Education Association Council (WEAC), restraining orders against students are rarely granted, and teachers are generally advised to do their best under difficult circumstances.17
Focus groups of MPS teachers conducted by the Wisconsin Policy Research Institute (WPRI) in the summer of 2012 suggest that discipline is a major issue in many schools, and one that is not adequately addressed by existing structures such as those described.
One high school teacher expressed extreme frustration with the behavior of his students:
“you're frustrated with all kinds of pieces that don't have anything to do with education. Telling somebody to be quiet, telling somebody to be polite, telling someone not to text during class, I think frustration is the biggest thing, you see a lot of frustration.”
Another teacher described how student behavior forced her to alter her teaching approach by putting increased focus on disruptive pupils:
“your naughty kids kind of take precedence, so as a teacher you're just kind of managing behaviors on a day-to-day versus putting them out – the ones that don't want to conform to being in a classroom and stay focused and it just kind of depends on where you are. And the child that comes will determine how smoothly your room is able to run and what materials you have to work with.”
Perhaps most telling, both the teachers quoted were dissatisfied with the existing procedures for dealing with disruptive students. The male teacher stated plainly:
“And we just don't get the backing of the discipline that we need.”
More disturbing was a description by the female teacher of a time where she was physically threatened and intimidated into not taking action to remove a problem student from a classroom. Specifically, she refers to the disabling of a classroom phone that is designed to give teachers a way to call for help:
“But when you start making physical threats, and you're invading my space, the telephones don't always work. I've tried to call before, and the kid has yanked the phone out of the wall.”
These answers were from MPS teachers and may not be typical of other Wisconsin schools. However, as illustrated in Table Two, districts across the state deal with high suspension rates, suggesting that the problem of disruptive students is not isolated to urban or large school districts. While school districts serving cities have the highest rates of suspension, rates in districts serving suburbs and towns are far from inconsequential. For example, 21 rural districts and 12 suburban districts have suspension rates above 5%.
Table 2 – Suspension Rates by District Type
The qualitative data from the WPRI focus groups gives reason to suspect that:
1) Disruptive students are a problem in some Wisconsin classroom, and
2) Existing policies for dealing with disruptive students are at times inadequate.
The fact that there are twice as many suspension incidents than students involved in suspensions shows that many students are multiple offenders. Teachers may be reluctant to remove pupils from the classroom if previous experience shows it to be a temporary fix to a long-term problem.
The next section of the report will use Wisconsin-specific data to test several hypotheses to determine the nature of and extent to which disruptive student behavior is affecting academic performance in Wisconsin.
The optimal way to test the impact of disruptive students in Wisconsin classrooms would be to run a test in actual classrooms.
If the classrooms with disruptive students scored significantly lower than the group without disruptive students, it would be possible to conclude that disruptive students negatively affect test scores in Wisconsin. Unfortunately, such an experiment is impossible to conduct because of ethical considerations and data limitations.
However, a multivariate regression analysis can be used to isolate the impact of suspension rates, which serve as a proxy for the level of disruptive student behavior in a school district. The creation of predictive models also can be used to make inferences regarding what would happen if suspension levels were reduced.
Unit of Analysis
The unit of analysis in this report is the school district. The impact of suspension rates on school district outcomes will be the focus of the hypotheses testing to follow. The reasons for focusing on the school district level are numerous.
First, there is data available for district level variables from both the DPI and the National Center for Education Statistics. Second, Wisconsin school districts are large enough to presume that suspension rates across school districts are generally measuring the same phenomenon. If we were to measure at the school level, in contrast, schools with low enrollment may produce suspension rates that vary widely year-to-year and lack a reasonable presumption of comparability to large schools. Third, and most important, the policies governing disruptive students, in accordance with Wisconsin state statute 120.13 (1)(b), are made at the school district level. While the actions that teachers might take to deal with disruptive students may be of interest, the focus of this report is to inform public policy. Given that these policies are district-level creations, an analysis focused on the school district level is logical.
The Independent Variable
If suspension rates are to be used as an indicator of classroom disruption, an understanding of what DPI-reported suspension rates measure is necessary. Suspension rates are simply the number of students suspended in a given year divided by the number of students enrolled in the same year. Statewide in 2011, a total of 48,726 of Wisconsin’s 871,550 public school students were responsible for more than 90,000 incidents, resulting in a suspension rate of 5.5% (48,726/871,550 = .055).
As shown in Figure One, suspension rates differ across school types. High schools account for 44% of all suspended students in Wisconsin (21,402), but elementary schools (13,391) and middle schools/junior highs (10,773) have a significant number of suspensions, too.
A possible weakness of using suspension rates as a proxy for student disruption is that a suspension may actually be a sign that student misbehavior is being addressed. It is possible, for example, that disruptive students’ behavior is being corrected through suspension. If this were the case, suspension rates might be a poor indicator of the overall level of disruption in a school district.
A simple histogram graphing the distribution of changes in district-level suspension rates will be used to check this possible measurement problem. If suspension rates are an indicator of a problem being solved, wide variation in changes in district suspension rates should be expected over time. However, if changes in district-level suspension rates generally mirror changes in state-level suspension rates, a certain consistent level of disruptive behavior in a school district can be assumed.
As illustrated in Figure Two, district changes in suspension rates are clustered around the state average decline of 1.7%, meaning districts in general are mirroring state trends. Given this, it is reasonable to conclude that suspension rates are a good district-level indicator of the level of disruptive behavior (as defined by volume of suspension worthy offenses) at the school district level.
Table Three lists the summary statistics for the 2011 suspension rate of Wisconsin public school districts.18 With 424 school districts reporting data in 2011, the average suspension rate was 2.43% per district, and the range of suspension rates varied from 0% to 23%. A review of the distributions of suspension rates indicates that four school districts can safely be called outliers in the data set. Bayfield, Beloit, Racine and Milwaukee all had suspension rates above 12%, far greater than the state average. The presence of outliers will be considered in the development of predictive models.
Table 3 – Summary Statistics for Suspension Rate
Three dependent variables will be used to test hypotheses: (See summary statistics in Table Four.)
- The percentage of district pupils scoring advanced proficient on the math Wisconsin Knowledge and Concepts Exam (WKCE) in all tested grades (3-8, and 10);
- The percentage of district pupils scoring advanced proficient on the reading WKCE in all tested grades (3-8, and 10); and
- The district average composite ACT score, which measures the college readiness of the 61% of Wisconsin juniors who take the test.
Table 4 – Summary Statistics for Dependent Variables
Both of the WKCE variables are imperfect, but popularly accepted (though widely criticized) indicators of academic achievement in Wisconsin. Justifying the use of WKCE scores as an indicator of academic success requires a fuller discussion of what the WKCE is, and what it is not.
The WKCE does not measure student growth. In order to understand the impact a specific school or district has on student test scores it would be necessary to test the pupil before he or she enters a school or district, account for all the non-classroom factors known to affect academic achievement, and test the student again after a given time period. Unfortunately, this type of data does not yet exist in Wisconsin in a way that can be used in this analysis. However, there are many reasons why the WKCE is still an acceptable marker of academic achievement for purposes of this study.
First, because the district is the unit of analysis the WKCE proficiency rates for students in grades 3-8 and 10 are pooled. In other words, changes in achievement levels over time in a district are likely to be somewhat accounted for because data for pupils who have been in a district for many years is included. It is theoretically possible that some districts will have an overrepresentation of pupils who just entered the district, but there is no reason to suspect this to be the case.
Second, the regression analyses that will be used include control variables for non-classroom factors known to influence academic achievement. Though not a perfect solution, the use of control variables will ensure that these known outside factors are taken into account when measuring the relationship between suspension rates and WKCE proficiency.
The third dependent variable, composite ACT score, is a measure of the average ACT score for district juniors. This measure, too, is imperfect because not all students in Wisconsin take the ACT. However, the inclusion of a control variable for the percentage of district juniors taking the ACT will ensure that the impact of low or high participation rates is included when measuring the relationship between suspension rates and ACT scores.
Hypotheses and Control Variables
Three hypotheses will be tested:
- High suspension rates negatively impact math WKCE proficiency levels.
- High suspension rates negatively impact reading WKCE proficiency levels.
- High suspension rates negatively impact ACT scores.
All three of these hypotheses originate from the original research question: What is the impact of disruptive students on academic outcomes in Wisconsin? As mentioned, testing these hypotheses requires the use of control variables known to affect academic outcomes, and those typically used in education research (Hanushek 1997).19 Such variables include:
- Socioeconomic demographics;
- Racial demographics;
- Student/teacher ratios;
- Teacher experience;
- School district expenditure;
- The percentage of special needs pupils in a district;
- School district size; and
- School district location.
Table Five lists the summary statistics of the continuous control variables used to test the previously listed hypotheses. All data comes from the DPI and the National Center for Education Statistics.
Table 5 – Summary Statistics of Control Variables20
Questions to Consider
A key point to consider before conducting any regression analyses is the nature of the relationship between district suspension rates and WKCE proficiency. First, is the relationship linear? A predictive model that allows for inferences to be made regarding the potential for increasing achievement by reducing suspension rates requires a linear relationship between the dependent and independent variables. However, a visual examination (see Figures Three and Four) suggests that the relationship is not linear. (Note that graphs using math WKCE proficiency rates look similar.) Figure Four demonstrates that the slope of the non-linear line of best fit increases dramatically after district suspension rates reach about two percent.
This suggests a spillover effect from disruptive behavior on student proficiency rates occurs once a certain level of disruption is reached. But what is meant by the phrase spillover effect?
It was established previously that existing research indicates that being suspended corresponds with lower individual achievement. Given that, overall district achievement should be expected to decrease at a predictable rate as the suspension rate increases due to the poor performance of suspended students. However, if the negative impact on test proficiency increases as suspension rates climb, there is good reason to suspect that proficiency levels of non-suspended students are being affected as well. This is the spillover effect.
Another way to visually test for the presence of a spillover effect is to divide the data sample in half and draw two lines of best fit for districts with suspension rates above and below the median suspension rate of 1.72%. As illustrated in Figure Five, the relationship between suspension rates and reading proficiency barely exists at suspension rates below 1.72%. There is a slight negative slope to the line, suggesting a slightly negative relationship most likely attributable to the lower performance of suspended students.
Figure Six, however, shows a strong negative relationship between districts with suspension rates above 1.72% and reading proficiency scores. Though a somewhat arbitrary cutoff point, using the median suspension rate is useful and logical for purposes of testing our three hypotheses. Accordingly, the regression analyses that follow are restricted to school districts with suspension rates above the state median of 1.72%, and the results of these analyses are only meaningful for Wisconsin school districts with suspension rates above the state average.
Results and Discussion
In this section, the following hypotheses limited to districts with suspension rates above 1.72% are tested using three Ordinary Least Squares (OLS) regression models:
- High suspension rates negatively impact math WKCE proficiency levels.
- High suspension rates negatively impact reading WKCE proficiency levels.
- High suspension rates negatively impact ACT scores.
The results listed in Table Six seem to support hypotheses one and two. There is evidence that suspension rates negatively impact the percentage of pupils that score advanced and proficient on the WKCE in reading and math. The models, as indicated by the R-Squared statistics, explain about 57% and 55% of the variation in reading and math WKCE proficiency, respectively.*
* The R-Squared statistic, or coefficient of determination, is a goodness of fit measure showing how much variation in the dependent variable is uniquely explained by the independent variables in the model. In other words, subtracting the R-squared statistics from 1 shows how much of the variation in the dependent variable is due to variables not included in the model.
This means that other factors not measured by the independent variables are accounting for some of the variation in test scores in Wisconsin districts. The unexplained variables might include parental involvement, outside learning opportunities, and other variables that are not directly measurable. The dependent variable, suspension rate, is statistically significant in both models.
More specifically, the coefficients for suspension rates indicate that with each single percentage point increase in the suspension rate, the percentage of students scoring advanced or proficient on the WKCE in reading decreases by 0.69 percentage points. Conversely, each percentage point decrease in the suspension rate, which is about 26 fewer students suspended in the average school district, increases the percentage of students scoring advanced or proficient on the WKCE in reading by 0.69 percentage points. So, if a district decreased its suspension rate by 5 percent, its reading proficiency rate would be expected to increase by 3.5%. Meaning, 3.5% more of the district’s students would, as defined by DPI, “[Demonstrate] a solid understanding of challenging subject matter and solves a wide variety of problems.” 21
Another way of looking at it is if the 100 Wisconsin school districts with the highest suspension rates reduced their number of suspended students by an average of 195 pupils, reading proficiency rates would increase by 3.5%.
Model Two shows that the percentage of district pupils scoring advanced and proficient on the math WKCE decreases by 0.94 percentage points with each single percentage point increase in the suspension rate. According to this model, a five-point reduction in the suspension rate would yield a 4.69 percentage point increase in math proficiency.
Several of the control variables in Models One and Two are statistically significant, including the percentage of minority pupils in a district, the percentage of pupils eligible for free and reduced price lunch in a district, and suburban location of a district. All of these variables have larger coefficients than the coefficient for suspension rates, meaning their manipulation by one unit (a percentage point or being or not being a suburban district) would yield greater variation in the math and reading proficiency than a one-unit manipulation of suspension rates. However, as will be discussed later, the major substantive difference between the significant control variables and the independent variable is that districts can take steps to change student behavior, but they cannot change the socioeconomic makeup of the students or locations of their school district.
Table 6 OLS Regression Results for WKCE Models 1 and 2
The results for Model Three, listed in Table Seven, support the hypothesis that high suspension rates negatively impact ACT scores in Wisconsin school districts with suspension rates above 1.72%. Though statistically significant, the relationship is less significant than the relationship between WKCE proficiency and suspension rates. With each percentage point increase in the suspension rate, district ACT scores can be expected to decrease by 0.12 points. In other words, increasing suspension rates by five percentage points would decrease the average district ACT score by 0.59 percentage points.
There are speculative reasons the relationship between suspension rates and ACT scores may be less substantive. First, the ACT is a self-selecting test geared toward college-bound pupils. Though a control variable for ACT participation rates is included in the model, it is possible that ACT scores are less affected because students taking it are less exposed to disruptive students based on the selection of higher-level courses. Second, college-bound pupils may be more likely to overcome the negative effect of disruptive pupils through non-classroom activities like independent reading.
Table 7 - OLS Regression Results for ACT Model 3
Overall, Models One, Two and Three demonstrate that there indeed exists a significant relationship between disruptive students and the overall level of academic achievement in Wisconsin school districts with suspension rates above 1.72%. Simply, high levels of student disruption are lowering test-score proficiency for Wisconsin pupils at an identifiable rate.
This finding is important because suspension rates are a variable that school districts can theoretically address in ways that result in higher test scores. As mentioned, other factors impacting student achievement in Wisconsin, such as socioeconomic status and location, are beyond the control of a school district. A district cannot take action to make their students come from more advantaged and involved families, but districts can address student behavior in ways that can be expected to increase achievement.
What Can Be Done: Conclusions and Recommendations
There is strong evidence that reducing suspension rates would increase the number of Wisconsin pupils with strong math and reading skills. The state can improve its students’ math and reading test scores without substantial cost by:
- Working to reverse disruptive behavior; and
- Utilizing virtual education as a tool for removing disruptive kids from traditional classrooms.
As state policymakers focus on reforms with substantial price tags, like small class sizes, the ability to raise achievement simply by making a more hospitable teaching and learning environment in Wisconsin should not be discounted. According to the analysis, decreasing suspension rates by five percentage points would yield an almost five percentage point increase in math proficiency and a three and one-half percentage point increase in reading proficiency in districts with high suspension rates.
To put those gains in perspective, the percentage of Wisconsin students scoring at least proficient in reading has remained almost stagnant since 2005, increasing by just half a percentage point.22 In math, the percentage of Wisconsin students scoring proficient on the WKCE has increased by about five percentage points since 2005.23 Taking aggressive action to eliminate the negative impact of disruptive students can ensure that even more Wisconsin students are where they need to be in math and reading without any additional state or local funds.
There are already some efforts under way in Wisconsin to address disruptive behavior. Many future teachers are being trained in how to deal with disruptive pupils. The UW-Stout education curriculum, for example, includes teaching of the work of educational and behavioral consultant Geoff Colvin (2009).24 Colvin identifies problem behavior as a significant detriment to student learning and provides tips for teachers to assess and manage problem behavior. Colvin (2010) identifies what he calls “The Three Main Goals in Correcting Problem Behavior:” 25
- Interrupt the problem behavior, and engage the student or students in the expected behavior.
- Ensure the student or students exhibit the expected behavior in future occurrences of similar situations.
- Avoid escalating the situation to more serious behavior.
The DPI is also working to improve student behavior through its use of the national Positive Behavioral Interventions and Supports (PBIS) program. PBIS, currently implemented in 916 Wisconsin public schools, is an effort to give teachers and school leaders access to evidence-backed strategies to create positive school environments that ultimately lead to increased academic achievement.26 An example of an evidence-based technique that schools can use is the use of portfolio grading for problem students. Education researchers Tracy Carpenter-Aeby and P. David Kurtz (2000) find evidence that portfolio work can help build student confidence and increase the prospects for at-risk pupils. Such an approach has been used with some success in Milwaukee’s Shalom High Partnership High School.27
Many of these initiatives fit under the broad spectrum of what is known as character education. Character education was broadly described in a 2010 study published by the National Center for Education Statistics as “school-based programs designed to help elementary schools foster positive student behaviors [and] reduce negative behaviors,” and as programs which “ encourage the development of desirable traits, values, and ethics (e.g., respect, responsibility, honesty, fairness, and caring).” 28
The basic concept of character education is simple: Schools should be teaching pupils to be responsible human beings as well as how to read, do math, etc.29 Character education is a tool worth pursuing in Wisconsin because, as this analysis shows, reductions in the number of disruptive student can yield real academic gains.30
The second and likely more controversial way for Wisconsin to reduce the negative effects of disruptive students is to remove them from the classroom. Of course, this solution leaves the problem of what to do with students who are removed from the classroom.
Under current law, disruptive Wisconsin students can be sent to alternative programs, defined in section 115.28(7)(e)1 of the state statutes as:
“an instructional program, approved by the school board, that utilizes successful alternative or adaptive school structures and teaching techniques and that is incorporated into existing, traditional classrooms or regularly scheduled curricular programs or that is offered in place of regularly scheduled curricular programs. ‘Alternative educational program’ does not include a private school, a tribal school, or a home-based private educational program.”
An obvious barrier to these types of programs, in particular in rural areas where comparatively few numbers of pupils might be served, are the significant resources needed to pay for specialty teachers, facilities and curricula. An alternative approach is the increased use of innovations such as virtual schools to meet the state’s constitutional requirement to make education available to even disruptive students. The wording of that requirement from the Wisconsin State Constitution follows:
“The legislature shall provide by law for the establishment of district schools, which shall be as nearly uniform as practicable; and such schools shall be free and without charge for tuition to all children between the ages of 4 and 20 years.”
Certainly, the growth of virtual district schools, non-district charter schools, and even publicly funded private school voucher programs in Wisconsin suggest that the state’s constitutional requirement to provide an education in schools “as nearly uniform as practicable” shows that non-traditional public schools can be used to meet the state’s constitutional obligation.
In practice, applying this lesson to the problem of disruptive students might be as simple as automatically enrolling students deemed chronically disruptive in a statewide virtual school created specifically for disruptive students. Specifically, school districts would be given the option of assigning students to a statewide virtual school as an alternative to expulsion. The virtual school could be hosted by a school district (or districts) willing to enroll these pupils via the state open-enrollment program. Under that set-up, resident school districts would still be able to count these students in their enrollment, while sending a transfer payment to the district(s) housing the virtual school.
Disruptive students enrolled in a virtual school could be given both a computer and Internet connection so that they are able to access their academic programming. This solution would ensure that disruptive students have access to a public education but are also removed from the physical classrooms where they are a detriment to the education of their classmates.
The link between student disruption and academic achievement is real, and if addressed, could improve academic performance in Wisconsin. Efforts such as PBIS and character education should be continued and strengthened. However, these programs must be accompanied by increased emphasis on finding fair ways to permanently remove chronically disruptive students from traditional Wisconsin classrooms. Also, the link between special needs and suspension deserves further scrutiny. If mainstreaming special needs children is bringing disruptive students into classrooms to the detriment of teaching quality, a better understanding of the correlation between specific special needs and disruptive behavior is warranted.
Wisconsin has a commitment to provide a public education for all of its pupils. However, that commitment comes with the responsibility of giving teachers an instructional environment in which they can succeed and protecting students from the negative effects of misbehaving classmates. The preceding analysis serves as evidence of the very real academic gains that can be made by reducing student disruption, the need to draw attention to existing efforts, and the need to spur new efforts to improve the environment in which Wisconsin pupils learn.
Wisconsin State Constitution
Skiba, R., & Rausch, M. K. (2004). The Relationship between Achievement, Discipline, and Race: An Analysis of Factors Predicting ISTEP Scores. Children Left Behind Policy Briefs. Supplementary Analysis 2-D. Center for Evaluation and Education Policy, Indiana University.
Lee, T., Cornell, D., Gregory, A., & Fan, X. (2011). High suspension schools and dropout rates for black and white students. Education and Treatment of Children, 34(2), 167-192.
Hanushek, E. A., Kain, J. F., O'Brien, D. M., & Rivkin, S. G. (2005). The market for teacher quality (No. w11154). National Bureau of Economic Research.
Darling-Hammond, L. (1999). Teacher quality and student achievement: A review of state policy evidence. Seattle, WA: Center for the Study of Teaching and Policy, University of Washington.
Rockoff, J. E. (2004). The impact of individual teachers on student achievement: Evidence from panel data. American Economic Review, 247-252.
Rothstein, J. (2010). Teacher quality in educational production: Tracking, decay, and student achievement. The Quarterly Journal of Economics, 125(1), 175-214.
Ennis, C. (1996). When avoiding confrontation leads to avoiding content: Disruptive students' impact on Curriculum. Journal of Curriculum & Supervision, 11(2), 145-162.
Garner, P. (1993). What disruptive students say about the school curriculum and the way it is taught. Therapeutic Care & Education.
Short, P. M., & Greer, J. T. (2002). Leadership in empowered schools: Themes from innovative efforts. Merrill Prentice Hall, One Lake Street, Upper Saddle River, NJ 07458.
Alsbury, T. L. (2008). The Future of School Board Governance: Relevancy and Revelation. Rowman & Littlefield Education. 15200 NBN Way, PO Box 191, Blue Ridge Summit, PA 17214-0191.
Buck, G. H. (1992). Classroom management and the disruptive child. Music Educators Journal, 36-42.
Adams, C., & Fruge, E. (1996). Why children misbehave and what to do about it. New Harbinger Publications, Incorporated.
Charles, C. M., Senter, G. W., & Barr, K. B. (1989). Building classroom discipline (pp. 143-148). New York: Longman.
Dornbush, M. P., & Pruitt, S. K. (1995). Teaching the tiger: a handbook for individuals involved in the education of students with attention deficit disorders, tourette syndrome, or obsessive-compulsive disorder. Hope Pr.
Barbetta, P. M., Norona, K. L., & Bicard, D. F. (2005). Classroom behavior management: A dozen common mistakes and what to do instead. Preventing School Failure: Alternative Education for Children and Youth, 49(3), 11-19.
DPI, An Introduction to Special Education.
Answers to Frequently Asked School Discipline Questions (2012), Wisconsin Department of Public Instruction.
Kratochvil, Jennifer A. "In Wisconsin Expulsions, We Don't Have to Leave Children Behind." Marquette Law Review 91, no. 4 (2008): 1213-31.
Wisconsin Department of Public Instruction, “Suspension Frequently Asked Questions.”
Sarah Carr, Disruptive Students Bounced from School to School, Milwaukee Journal Sentinel, May 8, 2007.
Madison Teachers Inc., Solidarity Newsletter, November 23, 2009.
16 Madison Teachers Inc., Solidarity Newsletter, November 23, 2009.
Molly Thompson, Taking Punches Not Part of the Job in Wisconsin Schools, The Wisconsin Education Association Council, February 6, 2006.
Data from the Wisconsin Department of Public Instruction.
Hanushek, E. (1997). Assessing the effects of school resources on student performance: An update. Educational evaluation and policy analysis, 19(2), 141.
Wisconsin Department of Public Instruction, WINSS Successful School Guide, available at:http://data.dpi.state.wi.us, and National Center for Education Statistics Common Core Database, U.S. Department of Education, available at: http://nces.ed.gov/ccd/bat/
DPI. (2012). Administrator’s Interpretive Guide Wisconsin Student Assessment System.
Madison, WI: Wisconsin Department of Public Instruction.
Data from the Wisconsin Department of Public Instruction.
Data from the Wisconsin Department of Public Instruction.
Colvin, G. (2009). Managing Noncompliance and Defiance in the Classroom. Thousand Oaks, CA: Corwin Press.
Colvin, G. (2010). Defusing Disruptive Behavior in the Classroom. Thousand Oaks, CA: Corwin Press.
Ryder, K., Thuli, H., Poulos, J., Schnebly, A., Marks, M., & Venit, K. (2012). 2011-2012 Evaluation Report. Madison, WI: Wisconsin RtI Center.
Carpenter-Aeby, T, & Kurtz, P. D. (2000). The portfolio as a strengths-based intervention to empower chronically disruptive students in an alternative school. Children & Schools,22(4), 217-231.
Consortium, Social and Character Development Research. "Efficacy of Schoolwide Programs to Promote Social and Character Development and Reduce Problem Behavior in Elementary School Children." Washington D.C.: National Center for Education Statistics, 2010.
Adams, Caralee J. "Character Education Seen as Student-Achievement Tool." Education Week, February 27, 2013.