Reading Level and Office Referrals: Is a Student's Reading Level an Antecedent to Office Referrals?
By Ann Bailey
Discipline in schools has been an ongoing struggle for teachers and administrators throughout the history of education. While the issue is a very complex interaction of many factors in a student's life, it is very important to a student's success in school and in adult life. Many studies have looked at numerous factors influencing the inappropriate behaviors of students in the classroom. They have studied family structure, family income level or SES, social structure, race, sex, and special education needs. Although these factors have a strong relationship to the inappropriate behaviors in the classroom, it is very difficult to develop an intervention plan for these factors within a school building. Teachers and administrators devote a large amount of time trying to reduce the number of office referrals in order to increase the learning time of each student in the classroom. In order to create the most effective intervention plan, strategies should be centered on behaviors that can be developed within the school building. In this study I will look at one factor that may be increasing the number of office referrals from the classroom and could be a dynamic part of a school intervention program, the reading level of the student. I would suggest that many office referrals stem from a student's inability to successfully complete or even comprehend the materials and activities assigned to the student. The student who is below reading level does not want to appear academically lower or intellectually inferior to other students, so they change the atmosphere of the classroom and reflect this frustration back at the teacher in the form of inappropriate behaviors. These behaviors allow the student to avoid the work and reduce the chances that they will be seen as below level by other students in the class. While this will not fully encompass all the students who are referred to the office, I believe they will be a substantial number of students. These students could then be part of an intervention program to increase their success in the classroom and reduce the number of office referrals. This study will also show the importance of grade level reading and instruction that effectively produces good readers in the elementary schools as a precursor to the middle school and high school years when office referrals increase in number.
Middle school is a very difficult transition time for students. They are experiencing many physical and emotional changes as well as increased expectations in the classroom. Many of these students are unable to find success and this lack of success continues through the higher grades. This is often shown through a student's increase in office referrals.
Some studies have shown that there are certain student characteristics that have a strong relationship with a higher number of office referrals. Skiba and Peterson (1997) found that students within certain categories were more likely to be referred to the office. Students who were African-American, low SES as determined by free and reduced lunches, emotionally disturbed students, and boys are more likely to be referred than other students in the school. Intertwined with this issue of increased office referrals, these student characteristics often have a strong relationship with low academic performance as well.
The SES of a student's family can affect their academic performance. Sutton and Soderstrom (1999) examined the relationship between school demographic and socioeconomic variables and student achievement as measured by the Illinois Goal Assessment Program (IGAP). The statistics for this study were retrieved from the demographic and socioeconomic factors reported on the Illinois School Report Card and were compared to the IGAP results. The authors determined that there are many variables that schools cannot control that affect a student's success academically. They found that the strongest predictors of low achievement on the IGAP in grade three are low income (SES), percent of Caucasian students, and attendance. In grade ten the strongest predictor was SES. In a study that examined school achievement and family background, Willms and Somers (2001) found that the most successful Latin American country with positive schooling outcomes was Cuba, which has very little variance between SES. These studies support the relationship between SES and academic success. This category is also a predictor of increased office referrals.
Studies have shown that there are significant differences in academic success between the sexes. Halpern (2002) examined the testing scores of males and females and whether test design favors either sex. She identifies females within the classroom as having higher grades and standardized test scores than males in most subject areas, yet males score higher on SAT tests with an average 35 point difference between males and females. She goes on to identify females as much more successful in the elementary school years, particularly in reading and language skills. Males are much more likely to be diagnosed with a disability at this age level and are more spatially oriented. Based on this study one could speculate that this earlier success in school for females could decrease their office referrals in higher grades since they see themselves as successful in earlier years and have lower frustration levels in the classroom.
Birth weight of a child has also been examined as a factor in future success of a student. Jefferis, Power and Hertzman (2002) completed a longitudinal study that examined the combined effect of social class and birth weight on future educational outcomes. They followed 10,845 males and females who were born during March 3 – 9, 1958 in regards to birth weight, social class (SES), and cognitive tests. Although there was a relationship with social class and test scores, birth weight showed the strongest relationship. Since SES has been shown as a factor in increased office referrals, their lack of success academically may show a relationship between these factors.
Race has continued to have a strong relationship with lower academic success. Kain and Singleton (1996) revisited a study that was completed almost 30 years ago, the Coleman Report. The Coleman Report concluded that African Americans are more likely to be in segregated schools and have lower academic scores on standardized achievement tests. Kain and Singleton reevaluated these conclusions as a repeated cross-sectional study of Texas students. They found that even though schools are more integrated, there is still an academic gap between minority achievements as identified through standardized tests in reading. It also concluded that as the years of education increase, the gap widens for most minorities. The status of minority in this test grew to include many more Hispanic students as the population of Texas has changed accordingly. The author also discussed the ramifications of English proficiency as a factor for Hispanic students. Both African American and Hispanic students tested at 94 percent of the statewide average while Asian and Anglo students tested at 107 and 106 percent. From this study, this difference in success in school, particularly in higher grades, could be identified by a student's frustration in the classroom. As students become less and less a part of the successful group in school, their interest in conforming to school rules may reduce considerably, resulting in a higher number of office referrals.
Attitude toward the referral process is also an important factor in office referrals. Romi and Freund (1999) found that teachers most consistently agree on the severity of the behaviors that involve office referrals, with verbal and physical violence toward other students being most severe. They also showed similar agreement among other behaviors. Parents and students did not show this same consistency in agreement, with students the least likely to agree on the severity of the behavior. This inconsistency may allow students to reduce their ownership of the behavior and place the blame on to the authority figure. They may not see themselves as affecting the learning process and school community to a great degree, which allows them to continue in their behavior without internal consequence. Tulley and Chiu (1998) also examined a student's perceptions toward classroom discipline. They surveyed 148 sixth grade students from three different schools in Indiana . They found that most of the reported discipline responses were disruptions, defiance, and aggression, with disruptions as the most frequent type of discipline problem. The authors then studied the most and least effective methods of strategies used by teachers. The two most effective strategies, as described by the students, was rote punishment or explanation. The removal from the class was seen as one of the least effective methods of punishment from the perspective of the student in the classroom. From these studies it might be concluded that the lack of student ownership for their behavior and their perspective of effectiveness of punishment may be a very important link in the successful intervention by a classroom teacher. This may also show that current methods of office referral may not be an effective method for changing a student's behavior, which is an integral part of an intervention plan.
When intervention plans are designed for students, you must first identify which patterns are predictors of these inappropriate behaviors. Tobin and Sugai (1996) studied the pattern of office referrals starting in grade six and whether there was an indicator pattern that would predict future repeat offenders in the following years. They found that if students received two or more referrals in sixth grade for any reason or one referral for harassment, they had a much higher chance of becoming repeatedly referred in higher grades. This study shows the need to begin interventions at an early stage of inappropriate behaviors to reduce the number of office referrals within a school.
The types of inappropriate behaviors students exhibit in the classroom are also of great importance when designing an intervention plan. Skiba and Peterson (1997) analyzed the discipline files of urban middle schools to detail the types of disciplinary offenses in the selected middle schools. They found that disobedience, conduct interference, and disrespect accounted for about 51% of the documented offenses. When they analyzed the types of disciplinary consequences, suspension (33%) was the most frequently used method of discipline. While there may be a variety of reasons for these classroom behaviors, an effective intervention plan must look at ways to change the behavior instead of merely removing the student.
Nelson, Martella and Marchand-Martella (2002) examined a comprehensive school-based program implemented at seven elementary schools for two years. Although the plan implemented many other aspects within the school, one element was one-on-one tutoring in early reading skills. Target students were identified as those who were most likely to express certain inappropriate behaviors. It was found that these target students increased academically and decreased their inappropriate behaviors. These findings support the theory that academic performance and inappropriate behaviors are inversely related.
Office referrals and low academic success may be identified as precursors to a future without success. In a study which examined behaviors of delinquent girls, Fejes-Mendoza and Darcy (1995) found that many of these students were below grade level and almost half had been retained at least one grade level (43% and 53%). Tobin and Sugai (1999) examined the relationship between sixth grade discipline referrals and referrals for violent behavior in eighth grade. They concluded that as the number of referrals increase for sixth grade students, they are more likely to be referred for violent behavior in eighth grade. They also followed the students into high school and found that they were also more likely to drop out of high school. These studies show the importance of identifying a pattern of behavior to implement a program to increase student success.
There have been studies that show success of reading programs within schools. Greenlee and Bruner (2001) show that many types of reading programs increase student performance on norm-referenced tests. It identified the importance of implementing reading programs to increase student achievement. Malmgren and Leone (2000) studied the effects of a short term reading program on incarcerated students. They found that within their population, 44.4% of their sample received special education services, compared to 10% nationwide. Additionally, in the pretest for reading, 61.2 % of all the students were below the first percentile as measured by the GORT-3. Within the short time the program was implemented, student generally showed significant gains within 3 of the 4 sub-tests, although a large number of students still tested under the first percentile. The authors concluded that even longer programs would produce even more significant progress in student reading achievement. Even more important would be early intervention with reading programs at primary grades.
Past studies have consistently shown a lack of achievement within certain student categories, as well as an increase in the number of office referrals within that same population. Many factors that influence the number of office referrals may not be under the control of the school and therefore would not be a very effective aspect of an intervention plan. Schools must focus on what they can change. A student's ability to read on grade level can be facilitated through reading intervention programs within the school. When a student finds success in the classroom they will take responsibility for their behaviors and become part of the educational group process. Students who see themselves as out of the educational loop may not find purpose in classroom activities and are more likely to act out in these circumstances. This study will examine the relationship between a student's reading level and the number of office referrals that student receives.
Methods
Hypothesis and Variables
The purpose of this study is to identify the relationship between the independent variable and the dependent variable; a student's reading level and the number of office referrals respectfully. It is hypothesized that students who are reading below grade level will receive more office referrals. This is a result of a student's frustration level in the classroom becoming an inappropriate behavior. The independent variable will be examined at several categorical levels. Students will be randomly selected from specified reading levels as assessed by school testing. The CTBS will be one test that will be used to identify the student's reading level. The CTBS is a multiple assessment test that has been tested by McGraw-Hill with assessed validity and reliability. The above-level reading students will be assessed by the CTBS reading score and their placement in a gifted and talented group for reading through the appropriate methods designated in the school system. On-grade reading students will be assessed by CTBS reading scores that indicate these students as reading on grade level and placement in an on-grade reading class. Students who are reading below grade level will be assessed by the CTBS reading score and additional testing by the school reading specialist as reading below their current grade level. These students will be separated into two categories, one year below grade level and more than one grade below grade level. A tally of written referrals will assess the number of office referrals, the dependent variable, for the participants in the study. The type of referral or the action by the administrator will not be assessed.
Participants
In this study, students will be selected from 6 th , 7 th and 8 th graders in several public school systems on the East Coast. Three school systems will be chosen from three specific demographic areas. The first system will be chosen in an urban area. In this urban area, the school system will have certain general characteristics. There will be a relatively even distribution of male and female students. The minority status should be at least 25%, which includes any non-Caucasian categories. The number of students who are eligible for free or reduced lunch will assess the economic status of the population. This category will be referred to as SES and the school should have a population of at least 25% low SES. The second and third school system will be chosen from a suburban and rural school system with similar category qualifications. It is preferable to identify school systems that are closely matched with their minority and SES levels.
Data Collection
The data for this study will be supplied by the above mentioned school systems.
Randomly generated identification numbers will track students so they will not be able to be identified in any way. Students who withdraw from the school system for any reason will be dropped from the study. The students will be selected in a stratified random survey strategy, to select an even distribution within the independent variable categories and within the three grade levels. Permission from the school system will be required before students will be selected from school system rosters. Individual notification of students and their families will not be necessary since students will not be able to be identified in this study nor will students be contacted at anytime. Office referrals will be assessed by information entered into the school system database. The type of referral or the resulting action will not be assessed.
Data Analysis
The data will be analyzed through an ANOVA test to assess the relationship between the ordinal independent variable categories and the dependent ratio variable. The ANOVA test is a preferable test with multiple group variables. This allows the test of each factor while controlling for all the other factors within the study. This test will also allow an analysis of the interaction between the groups assessed in this study. The statistical relationship between the groups will be assessed through a p-level test, which is considered statistically significant if p<.05. Additionally a Newman-Keuls test will be completed after the statistics are assessed to show that the difference between the groups is not just due to chance. The results of this study will be summarized in both table and graph representation.
Validity and Reliability
The methods used to complete this study will provide validity for this research project. Measurement validity is reached through several techniques in this experiment. The subjects are chosen using a stratified random sampling process. This way the students chosen for the sample will have an equal representation from sixth, seventh and eighth grade and within the identified reading levels. The selected subjects will also be as heterogeneous as possible by choosing samples from schools with similar SES, sex, and minority distributions. The largest sample possible will be chosen to allow for the four comparisons made between the independent variable categories. It is through this natural setting, the minimal influence from the researchers gathering data, and the random population samples chosen that the external validity is increased and the results are more readily generalized to the larger population.
The variables will be analyzed through an ANOVA test to assess the relationship between the ordinal independent variable categories and the dependent ratio variable. This type of test will accurately measure the relationship between more than two variable categories. An additional p-level test will be completed to identify statistical significance and a Newman-Keuls test will determine that the relationship between groups is not due to chance. Since the statistical measurements are qualitative in nature, there will be very little subjectivity when interpreting the data. This is also true when all office referrals are counted without making a judgment on the type of referral made. This will increase the content validity of this statistical measure.
The criterion validity of the measurement of the independent variable is strengthened by not just looking at the student's placement in a class, but also at CTBS scores on the reading test, which has been shown over time to be a reliable test. To establish a strong causal relationship between the variables, past research has identified a strong relationship between many factors involving office referrals and academic success. If reading level is shown to have a strong relationship with office referrals, teachers and administrators will have specific intervention strategies to reduce office referrals. A causal direction of the variables are also supported, since a low reading level could cause office referrals, but office referrals have not been shown in any past studies to cause a student to develop a lower reading level. Based on past research, spurious variables will be controlled for in selection of population and sampling strategy so all population sampled can be compared, in turn, to the larger population. The research has shown a strong relationship between certain population categories and office referrals and these same categories and academic success. There has also been a relationship between a lack of student success and higher drop out rates and incarceration. With all these conditions in use, this experiment will accurately measure the relationship between a student's reading level and the number of office referrals.
There may be some difficulty with the populations chosen for the study. Although the three school systems may be similar, it may not represent a school system that differs in an extreme statistic, such as a much larger or smaller minority or SES population. It has been shown in past studies that students in these categories are often referred to the office more frequently than other students. It has also been shown that students who have disabilities are more likely to be referred to the office, but since I have chosen to look at reading level I did not want to include this as a factor. This may be seen as a fault in the controls. I have not looked at the relationship between reading level and SES and minority category. I believe there could be a strong relationship between these variables, but I think this would be another experiment entirely and would add too many variables to this study. It may be a good follow-up experiment to this study.
It also may be of significance that I did not choose to look at the type of referral made for the students involved in the study. It was shown in other studies that disruptive and disrespectful behaviors are most likely reasons to be referred to the office while violent offenses are not as frequent. (Skiba & Peterson 1997) Since students who would be frustrated with their schoolwork may express this frustration in a variety of behaviors, I did not want to limit these choices. Other studies have shown that social and family environments have a strong relationship with office referrals. This is not being controlled for in this experiment and could be seen as a spurious variable, but this issue could be found across all levels of reading.
The structure of this study is as ethical as possible. Since the study is merely a collection of data that is readily available, there is little harm possible for the student participants in this study. Additional measures are taken so no student or school system will be able to be identified through this study. Permission from the school system will be required and a general notice to all school system members will be made prior to any data acquisition. All measures and findings will be presented to a review board to assure accurate findings and additional reviews by fellow colleagues will be supported as well. If a strong relationship is found between a student's reading level and the number of office referrals, these findings will be a valuable tool for both educators and parents to find a possible solution for inappropriate student behaviors.
Through these studies you can see that there are many factors that influence the number of office referrals a student will receive. The student's reading level may not be the most influential factor for all students, but for those students who are frustrated in the classroom this is a factor that can be pursued in the school building. Many of the other factors identified in studies; SES, minority, and sex, are not able to be controlled the school staff. If a student's reading level has a strong relationship with the number of office referrals a student receives, then school staff can implement interventions that can be controlled within the confines of the school building, in turn reducing the number of office referrals within the school. The increase in reading levels and reduction in inappropriate behaviors will also increase student success levels, and reduce high school drop-out rates and rates of incarceration.
Demographic: Percent:
Caucasian
African-American
Asian American
Hispanic
Native American
Middle-Eastern
Pacific Islander
Other
Male
Female
Students qualifying
for free or reduced lunch
(out of 100%)
Table 2: Multi-Factor ANOVA Test for Reading Level and Office Referrals Collapsing All Geographic Areas.
Reading Level Mean Number of Office Referrals
Above Grade Level
On Grade Level
One Year Below
More Than One Year Below
|
Sum of Squares |
df |
Mean Square |
F |
Sig. |
Between Groups Within Groups Total |
|
|
|
|
|
Reading Level Mean Number of Office Referrals
Above Grade Level
On Grade Level
One Year Below
More Than One Year Below
|
Sum of Squares |
df |
Mean Square |
F |
Sig. |
Between Groups Within Groups Total |
|
|
|
|
|
Reading Level Mean Number of Office Referrals
Above Grade Level
On Grade Level
One Year Below
More Than One Year Below
|
Sum of Squares |
df |
Mean Square |
F |
Sig. |
Between Groups Within Groups Total |
|
|
|
|
|
Reading Level Mean Number of Office Referrals
Above Grade Level
On Grade Level
One Year Below
More Than One Year Below
|
Sum of Squares |
df |
Mean Square |
F |
Sig. |
Between Groups Within Groups Total |
|
|
|
|
|
Fejes-Mendoza, Kathy and Others. (1995) Portraits of Dysfunction: Criminal, Education, and Family Profiles of Juvenile Female Offenders. Education and Treatment of Children v18 n3 p309-21. Retrieved July 17, 2003 , from ERIC database.
Greenlee, Bobbie J. & Bruner, Darlene Y. (2001) Effects Of Success For All Reading Programs On Reading Achievement In Title I Schools. Education; Fall2001, Vol. 122 Issue 1, p177. Retrieved July 28, 2003 , from MasterFILE Premier.
Halpern, Diane F. (2002) Sex Differences in Achievement Scores: Can We Design Assessments that are Fair, Meaningful, and Valid for Girls and Boys? Issues in Education; 2002, Vol. 8 Issue 1, p2. Retrieved July 28, 2003 , from MasterFILE Premier.
Jefferis, Barbara J. Power, Chris Hertzman, Clyde . (2002) Birth weight, childhood socioeconomic environment, and cognitive development in the 1958 British birth cohort study. British Medical Journal; 8/10/2002 , Vol. 325 Issue 7359, p305. Retrieved July 28, 2003 , from MasterFILE Premier.
Kain, John F. & Singleton, Kraig. (1996) Equality of educational opportunity revisited.
New England Economic Review; May/Jun96, p87, 25p Retrieved July 28, 2003 , from MasterFILE Premier.
Malmgren, Kimber W. & Leone, Peter E. (2000) Effects of a Short-Term Auxiliary Reading Program on the Reading Skills of Incarcerated Youth. Education & Treatment of Children; Aug2000, Vol. 23 Issue 3, p239. Retrieved July 28, 2003 from MasterFILE Premier.
Nelson, J. Ron Martella, Ronald M. Marchand-Martella, Nancy . (2002) Maximizing Student Learning: The Effects of a Comprehensive School-Based Program for Preventing Problem Behaviors. Journal of Emotional & Behavioral Disorders; Fall2002, Vol. 10 Issue 3, p136. Retrieved July 28, 2003 , from MasterFILE Premier.
Romi, Shlomo & Freund, Mira. (1999). Teachers', Students' and Parents' Attitudes towards Disruptive Behavior Problems in High School: A Case Study. Educational Psychology: An International Journal of Experimental Educational Psychology v19 n1 p53-70. Retrieved July 17, 2003 , from ERIC database.
Skiba, Russell J. & Peterson, Reece L. (1997) Office referrals and suspension: Disciplinary intervention in Middle Schools. Education & Treatment of Children; Aug97, Vol. 20 Issue 3, p295. Retrieved July 28, 2003 , from MasterFILE Premier.
Sutton, Alice & Soderstrom, Irina. (1999) Predicting elementary and secondary school achievement with school-related and demographic factors. Journal of Educational Research; Jul/Aug99, Vol. 92 Issue 6, p330. Retrieved July 28, 2003 , from MasterFILE Premier.
Tobin, Tary and Others. (1996). Patterns in Middle School Discipline Records.
Journal of Emotional and Behavioral Disorders v4 n2 p82-94. Retrieved July 17, 2003 , from ERIC database.
Tobin, Tary J. & Sugai, George M. (1999) Using sixth-grade school records to predict school violence, chronic discipline problems, and high school outcomes. Journal of Emotional & Behavioral Disorders; Spring99, Vol. 7 Issue 1, p41. Retrieved July 28, 2003 , from MasterFILE Premier.
Tulley, Michael & Chiu, Lian Hwang (1998) Children's perceptions of the effectiveness of classroom discipline techniques. Journal of Instructional Psychology; Sep98, Vol. 25 Issue 3, p189. Retrieved July 17, 2003 , from Academic Search Elite
Willms, Douglas J. & Somer, Marie-Andrée. (2001) Family, Classroom, and School Effects on Children's Educational Outcomes in Latin America . School Effectiveness & School Improvement; Dec2001, Vol. 12 Issue 4, p409, 37p. Retrieved July 17, 2003 , from Academic Search Elite.