effective collaborative teaching strategieshome.lagrange.edu/educate/advanced programs/m.ed....

EFFECTIVE COLLABORATIVE TEACHING STRATEGIES: A STUDY OF THREE SIXTH GRADE CLASSROOMS

A thesis submitted

by

Jarrod Paul Bingham

to

Lagrange College

in partial fulfillment of

the requirement for the

degree of

MASTER OF EDUCATION

in

Curriculum and Instruction

Lagrange, Georgia

May 10, 2011

Effective Collaborative Teaching iii

Abstract

This action research study was designed to test the effect of using parallel, station,

and alternative collaborative teaching structures in a sixth grade mathematics classroom.

The three focus questions that guided the research process were concerned with

discovering how to implement the structures, student learning outcomes, and teachers’

and students’ appreciation of and ability to adapt to working in a classroom in which the

three structures are being used. While the treatment group in this study did not show

significant academic gains when compared to the control groups, they did display an

improved outlook toward learning in a collaborative classroom.

Effective Collaborative Teaching iv

Table of Contents

Abstract................................................................................................................iii

Table of Contents.................................................................................................iv

List of Tables .......................................................................................................vi

Chapter 1: Introduction.........................................................................................1Statement of the Problem..........................................................................1Significance of the Problem.......................................................................2Theoretical and Conceptual Frameworks..................................................2Focus Questions........................................................................................6Overview of Methodology .........................................................................7Human as Researcher...............................................................................8

Chapter 2: Review of the Literature ....................................................................9Inclusive Education: Equal Access to all Students ...................................9Collaborative Teaching: A Definition and Brief Rationale ........................10General Guidelines for Effectively Implementing the Strategies ..............12Implementing Specific Collaborative Teaching Strategies .......................15Measuring Student Outcomes in Collaborative Teaching Classrooms.....18Student and Teacher Perception of Collaborative Teaching.....................20

Chapter 3: Methodology......................................................................................22Research Design......................................................................................22Setting......................................................................................................23Subjects and Participants.........................................................................23Procedures and Data Collection Methods ...............................................24Validity and Reliability Measures..............................................................27Analysis of Data .......................................................................................30

Chapter 4: Results...............................................................................................33

Chapter 5: Analysis and Discussion of Results...................................................55Analysis....................................................................................................55Discussion................................................................................................67Implications..............................................................................................70Impact on Student Learning.....................................................................73

Effective Collaborative Teaching v

Recommendations for Future Research.............................................................73

References.........................................................................................................74

Appendices………………………………………………………..…………………77

Appendix A Instructional Plan

Appendix B Instructional Plan Rubric

Appendix C Pre/Post Test

Appendix D Student Survey

Appendix E Reflective Journal Prompts

Appendix F Interview Questions

Effective Collaborative Teaching vi

List of Tables

Tables

Table 3.1 Data Shell.......................................................................................25

Table 4.1 Anova: Pretest................................................................................36

Table 4.2 Anova: Posttest..............................................................................37

Table 4.3 t-Test: Treatment Group.................................................................37

Table 4.4 t-Test: Control Group A..................................................................38

Table 4.5 t-Test: Control Group B..................................................................38

Table 4.6 Anova: Pretest, No Outliers............................................................39

Table 4.7 Anova: Posttest, No Outliers..........................................................40

Table 4.8 t-Test: Treatment, No Outliers........................................................40

Table 4.9 t-Test: Control Group A, No Outliers..............................................41

Table 4.10 t-Test: Control Group B, No Outliers..............................................41

Table 4.11 Anova: Pretest, Girls......................................................................42

Table 4.12 Anova: Pretest, Boys......................................................................42

Table 4.13 Anova: Posttest, Girls.....................................................................43

Table 4.14 Anova: Posttest, Boys....................................................................43

Table 4.15 t-Test: Treatment, Girls..................................................................44

Table 4.16 t-Test: Treatment, Boys..................................................................44

Table 4.17 t-Test: Control Group A, Girls.........................................................45

Effective Collaborative Teaching vii

Table 4.18 t-Test: Control Group A, Boys........................................................45

Table 4.19 t-Test: Control Group B, Girls........................................................46

Table 4.20 t-Test: Control Group B, Boys.......................................................46

Table 4.21 Chi Square....................................................................................48

Effective Collaborative Teaching 1

CHAPTER ONE: INTRODUCTION

Statement of the Problem

The purpose of this study is to examine the effect of established collaborative

teaching strategies upon the achievement of both general and special education students

in a sixth grade math inclusion classroom. Villa, Thousand, and Nevin define a

collaborative or co-teaching team as “a content teacher and a special education teacher

who teach the curriculum to all students” (as cited in Little & Dieker, 2009, p. 42). In an

effective co-teaching situation, each teacher has a partner with which to plan and share

ideas, and each student has an additional teacher to provide a fresh prospective and small

group instruction. It follows that co-teaching should greatly benefit both teachers and

students.

In many classrooms, however, the teaching partners do not work together as they

should. Patterson, Connolly, and Ritter (2009) describe a typical collaborative teaching

classroom. The general education teacher teaches as he/she has always taught; he/she

guides the students through a lesson, shows a few example problems, and gives an

independent assignment for the students to complete. During this time, the special

education teacher floats around the room and encourages the students to stay on task and

pay attention. The teachers in the previous example are employing a commonly used co-

teaching model that Friend (2008) dubs “one teach, one assist” (p. 92). Patterson et al.

(2009) explain that the teachers’ sole use of the one teach, one assist method leads to the

students becoming very dependent upon the teachers. These teachers, like many teachers

throughout the United States, collaborate in this way only because it seems like the most

natural way to work together—not because it is the best way to teach. Teachers often


receive very little co-teach training; they are thrown into the same classroom and told to

teach the content together. As a result, both teachers teach “as they had always taught”

(Patterson et al., 2009, p. 48).

Significance of the Problem

When co-teachers rely solely upon a single co-teaching strategy, particularly if it

is a strategy that fails to take full advantage of the skills of both teachers, the students do

not benefit as much as they should from having two teachers. Furthermore, an unhealthy

classroom dynamic often forms in which one educator appears to be the main teacher and

the other appears to be a helper or teacher’s assistant. After relying solely upon the one

teach, one assist approach, the teachers in Patterson, et al.’s (2009) example realize that

they are not meeting the needs of their students. In the most dramatic cases, co-teaching

in an inclusion classroom hurts the general education and/or the special education

students as the single lead teacher tailors instruction to a particular group of students.

When collaborative teachers employ a variety of co-teaching methods, particularly

methods that involve both teachers equally, they are better able to meet their students’

needs.

Theoretical and Conceptual Frameworks

When executed correctly, collaborative teaching enables classroom teachers to

more easily establish instruction that is built upon the principles of social constructivism.

Powell and Kalina (2009) explained that constructivism, or cognitive constructivism, was

construed by Piaget, who claimed that “humans cannot be given information, which they

immediately understand and use; instead, humans must construct their own knowledge”

(p. 242). Years later, Vygotsky built upon Piaget’s theory in creating social


constructivism (Powell & Kalina). Vygotsky theorized that “ideas are constructed

through interaction with the teacher and other students” (Powell & Kalina, p. 241).

The following example illustrates the difference between the practices of a traditional

teacher and a constructivist teacher. A father wants to teach his son how to build a tree

house. He buys the needed supplies, grabs his son, and begins to talk the boy through the

process. After a couple of hours of instruction, the father is sure that his son will be

successful—he did, after all, go over every little detail. The father points to the supplies,

wishes his son well, and walks back into his house confident that his son will not fail.

Two hours later, the man expectantly emerges from his dwelling only to see a hideous

construct that appears more like the work of crazed beavers than the tree house that he

knew his son was capable of building. Like the father in the story, many well intentioned

teachers believe that continuously telling students how to follow an algorithm will

eventually lead to their being able to solve problems correctly. This method of teaching,

which many call direct instruction, is a viable teaching strategy when used in extreme

moderation. Teachers who overuse direct instruction, however, fail to consider their

students’ background knowledge, preferred learning styles, and current level of

understanding. Furthermore, co-teachers who use too much direct instruction in an

inclusion classroom risk boring the more capable students and leaving the special needs

students behind as they aim instruction at the students in the middle. It follows that co-

teachers, more than any other educators, must not fall into the direct instruction trap.

What if the father in the previous example had approached the situation differently?

Instead of explaining how to build a tree house, he gives his son a hammer and shows the

boy how to get started. This time the father watches and assesses as his son works. As the


man watches his son, he provides support by telling the boy what he is doing wrong as

well as what he is doing right. When the boy doesn’t know how to do something, his

father helps him out. The boy learns how to build a tree house in the most obvious way:

by building a tree house. Every time the boy builds a new tree house, he needs a little less

help from his father. Eventually, the boy is able to show his friends how to build their

own tree houses. This time, the father taught his son from a constructivist viewpoint.

Rather than telling his son how to build a tree house, he provided a task that was within

his son’s zone of proximal development (ZPD), provided scaffolding, and allowed his

son to “construct” his own knowledge. ZPD, scaffolding, and social interaction are three

of social constructivism’s most important tenants (Powell & Kalina, 2009). Powell and

Kalina (2009) define ZPD as “a zone where learning occurs when a child is helped in

learning a concept in the classroom.” (p. 244). Scaffolding is “an assisted learning

process that supports the ZPD” (p. 244). Constructivist teachers seek to understand their

students’ current knowledge levels, set goals within their students’ ZPDs, and scaffold

students as they build new knowledge. Effective co-teaching strategies, like parallel and

station teaching, enable teachers to move past the traditional role of dispensers of

knowledge to the new role of facilitator. Once teachers implement these strategies, they

will better understand their students’ current level of understanding. This will enable

teachers to design appropriate tasks for each group of students. Because effective co-

teaching strategies lower the student-teacher ratio, practicing them will also allow

teachers to better provide the appropriate scaffolds to each student. Students also benefit

from the increased social interaction that many co-teaching models provide. Considering

the variety of student levels that are present in most inclusion classes, collaborative


teaching practices that do not approach student variance and learner profiles from a

constructivist viewpoint could hardly be considered effective.

The content of this thesis most closely relates to Tenet one of the Lagrange

College Education Department’s (2008) Conceptual Framework, which states that

learners should be enthusiastically engaged in the learning process. The third cluster, in

particular, states that teachers should not only understand how students learn and

develop, but should also be able to provide learning opportunities that “support students’

intellectual, social, and personal development based on students’ stages of development,

multiple intelligences, learning styles, and areas of exceptionality” (p. 4). The purpose of

intentionally implementing specific co-teaching strategies is to create a classroom

environment in which teachers better understand their students’ learning needs, which

enables them to provide instructional opportunities based upon those needs. Students who

are taught according to their developmental and learning needs are much more likely to

be enthusiastically engaged in learning.

The first Tenet of the Lagrange College Education Department’s (2008)

Conceptual Framework bears remarkable similarity to the second domain and third

proposition of the Georgia Framework for Teaching and the NBPTS Core Propositions

for Experienced teachers respectively. The first document reiterates that teachers should

consider students’ knowledge and development when planning and delivering instruction.

Proposition three of NBPTS specifically states that teachers should utilize a variety of

instructional techniques in order to keep students motivated and engaged. When

implemented correctly, co-teaching provides teachers with several methods to easily

reach all of the above goals.


Focus Questions

This research was guided by three focus questions. Before implementing specific

co-teaching strategies in the classroom, teachers must know the most effective way to do

so. The first focus question—What is the most effective way to implement parallel,

station, and alternative teaching structures in a co-teach classroom?—was designed to

address this.

The second focus question, which follows, quantitatively measured the study’s

results: “what effect do co-teaching strategies have upon student achievement?” In the

wake of No Child Left Behind’s test driven educational reforms, any study must focus, at

least in part, on student achievement on written assessments. In this study, I measured

student achievement by calculating both individual student growth using pre and post

tests and overall class performance by comparing the class that received the treatment to

a control class, which did not receive the treatment.

The last focus question was “how well do students and teachers adapt to this new

type of instruction?” Because analyzing test data alone cannot render a complete picture

of the effects of implementing specific co-teaching structures, I designed the third focus

question to gauge important qualitative factors including student engagement and

perception as well as the perceptions of the teachers involved in the study. The teachers’

perceptions of the co-teach structures that were studied were considered critical due to

the simple fact that teachers generally will not teach in a way that they hate to teach. In

order for strategies like these to achieve widespread usage, they must be presented in a

way that is pleasing to teachers regardless of their teaching style.


Overview of Methodology

This study, which most closely followed the principles of action research, used

both qualitative and quantitative data types as well as descriptive and inferential statistics

to examine the effects of implementing parallel, station, and alternative teaching in a

sixth grade mathematics inclusion class. To study these effects, I drew a convenience

sample from a rural middle school in west Georgia. The subjects were chosen based upon

their placement in one of three sixth grade mathematics classes. The control group was

the only class that did not have a co-teacher; the remaining two groups were in inclusion

classes and therefore had both a general and a special education teacher. I administered a

pretest to all three groups and found that there was no significant difference between the

classes at the onset of instruction. All three groups received quality instruction that was

differentiated according to student need, learning style, and/or interest. The major

difference between the three groups was the structure in which the information was

presented. For the first inclusion group, the special education teacher and I implemented

the following three team based co-teaching structures: parallel, station, and alternative

teaching. I have used the term “team based” to describe these particular structures

because all three of them fully utilize the skills of both teachers. The second inclusion

group received the more common, but often considered the least effective, co-teach

structures of one teach, one assist and one teach, one observe. The non-inclusive class

received no co-teaching. All classes received an identical posttest at the study’s

conclusion. I then used a dependent t-test to measure growth within each class and an

ANOVA to compare each of the groups.


In addition to the quantitative measurement enumerated in the preceding

paragraph, I recorded observational data in a reflective journal. The data included notes

on student engagement as well as objective observations concerning my perception of

each class. To obtain perception data from the special education teacher, I administered a

written survey. Student perception was gathered from surveys.

Human as Researcher

In the year and a half that I have been a classroom teacher, I have had the

opportunity to work with sixth, seventh, and eighth grade general and special education

students. The undergraduate program from which I graduated focused heavily upon the

principles of learner centered and constructivist education. As a result, I have endeavored

to incorporate differentiated learning strategies into classroom instruction. Going into this

study of collaborative teaching strategies, I believed that parallel, station, and alternative

teaching would lead to greater student engagement and achievement. I also believed that

using the one teach, one assist and one teach, one observe methods would produce similar

results to teaching without a co-teaching partner.


CHAPTER TWO: REVIEW OF THE LITERATURE

Inclusive Education: Equal Access to all Students

The passage of Public Law 94-142, or The Individuals with Disabilities Education

Act (IDEA), and subsequent amendments fundamentally altered the delivery method of

special education services in the United States of America by mandating that all

handicapped children be educated in the least restrictive environment (LRE) that is

possible given their disability (Falvey & Givner, 2005; Little & Dieker, 2009; Patterson

et al., 2009; Thousand, Villa, & Nevin, 2006). Inclusive classes, which are classes that

include both general and special education students, fulfill the LRE requirement of IDEA

for many students with disabilities (Patterson et al., 2009). According to Falvey et al.

(2005), denying inclusive education is very difficult in the post IDEA education system.

Indeed, the 2008 National Center for Education Statistics claims that “54 percent [of

students with disabilities] receive instruction in general education classrooms at least 80

percent of the school day and another 25 percent of those students are in general

education classrooms between 40 percent and 80 percent of the day (as cited in Treahy &

Gurganus, 2010, p. 485).”

Current research has suggested a number of rationales for the widespread use of

inclusive education in the United States. The reasons most often given include providing

special education students with access to the general education curriculum, embracing

diversity and meeting every student’s need to belong, and more effectively facilitating the

non-curricular lessons that students generally receive in a general education classroom.

Falvey (2004) described the pre IDEA special education system as being one in which

students were given materials that were not age-appropriate. The students were being


treated as young children, and many of them were not being challenged. In 2001,

however, the No Child Left Behind Act promoted inclusive settings for more students

with disabilities by holding them accountable to the same standard as their non-disabled

peers (Dingle, Falvey, Givner, & Haager, 2004; Thousand et al., 2006; Treahy &

Gurganus, 2010). As early as 2000, the National Council of Teachers of Mathematics

issued the following statement regarding the importance of giving all students access to a

rigorous math curriculum:

All students, regardless of their personal characteristics, backgrounds, or

physical challenges, must have opportunities to study--and support to learn--

mathematics. This does not mean that every student should be treated the same.

But all students need access each year they are in school to a coherent,

challenging mathematics curriculum that is taught by competent and well-

supported mathematics teachers (National Council for Teachers of Mathematics

(NCTM), 2000, p.12).

Falvey and Givner (2005) indicated that inclusion is the ideal special education delivery

method for most students because it ensures that they will have access to the same

rigorous standards as every other student.

Collaborative Teaching: A Definition and Brief Rationale

Treahy and Gurganus (2010) briefly discussed the following challenges faced by

the inclusion movement: students placed in classes without essential supports,

insufficient accommodations given to students, and the lack of adequate professional

training given to general education teachers in how to teach students with disabilities.

When effectively implemented, collaborative teaching or co-teaching seems to


adequately address the challenges enumerated above (Cook & Friend, 1995; Little &

Dieker, 2009). Gurgur and Uzuner (2010) described co-teaching as the sharing of

planning, teaching, and assessing responsibilities by a general and special education

teacher in the same classroom environment. Patterson et al. (2009) and Little and Dieker

(2009) extended this definition by insisting that both teachers work to meet the

instructional needs of all students--not just those with special needs. In general, a co-

teaching team consists of a general education teacher, who specializes in content and

teaching methods, and a special education teacher, who specializes in making

accommodations (Little & Dieker, 2009). Little and Dieker continued by explaining that

a true co-teaching situation is one which both teachers are equally involved in all the

curricular and non-curricular aspects of running a classroom including planning and

delivering instruction, assessment, and classroom management. Special education

teachers, therefore, should not be treated like a glorified and well paid teacher’s assistant,

but instead should be considered an integral and irreplaceable part of a teaching team.

Research has suggested that co-teaching can be an effective medium to deliver

special education services. Villa and Thousand (2005) stated that delivering the twenty-

first century curriculum requires teachers to try new ways of teaching like cooperative

group learning, differentiated instruction, active student-centered learning, detracking,

and focusing on social skills. Co-teaching, stated Cook and Friend (1995), helps teachers

expand their instructional approaches and increase opportunities for student success.

Cook and Friend continued by noting that co-teaching lowers the student-teacher ratio,

which allows for greater student participation, increases student engagement, and makes

it easier for teachers to facilitates many of the instructional strategies listed above.


Besides providing more instructional options and increasing engagement for all students,

co-teaching promises to “reduce the stigma for students with special needs” (Cook &

Friend, 1995, p. 3).

General Guidelines for Effectively Implementing Collaborative Teaching

Strategies

Research has suggested several general teacher, administrator, and environmental

guidelines for implementing a successful collaborative teaching program. To lead an

effective collaborative teaching program, teachers should be willing to take the following

two steps: frequently communicate with teaching partner and differentiate instruction

based upon students’ needs, learning styles, and interests. Perhaps the single most often

cited key to establishing an effective co-teaching program is frequent communication and

planning. Sileo and Garderen (2010) insisted that co-teaching requires a considerable

amount of planning. Teachers, suggested Murawski and Dieker (2004), should plan

together, using the same plan book, at least twice a week. Writing specifically to special

education teachers, Hammeken (2000) suggested that co-teaching partners schedule time

to meet with each other and plan daily. When they meet, the two should bring each of

their individual perspectives together to develop a plan that will benefit all students. Cook

and Friend (1995) contended that frequent collaborative planning, although ideal, is not

always possible due to large caseloads, scheduling, and other professional

responsibilities. Still, because lack of common planning often leads to poor collaborative

teaching practices, co-teaching partners should endeavor to plan together as often as

possible (Gurgur & Uzuner, 2010).


Most research has agreed that using the collaborative setting to differentiate and

modify instruction according to student need is another important adjustment that

teachers should make to develop a successful collaborative teaching program. According

to Dingle et al (2004), “general and special educators must have the skills necessary to

modify educational programs to effectively meet the needs of individual students” (p.

37). Patterson et al. (2009) illuminated the importance of differentiating instruction in the

collaborative teaching classroom. They described two teachers who were not taking full

advantage of a two teacher classroom. Nearly every day, the general education teacher

stood in front of the class and taught in a very traditional manner while the special

education teacher mostly observed and occasionally assisted one of the special needs

students. The two teachers decided to place students into heterogeneous flex groups and

differentiate their instruction. For the first time, the teachers began to realize the full

benefit of a two teacher classroom.

School administration must also take certain measures to ensure the success of a

collaborative teaching program. One of the most significant administrative measures is an

obvious extension of the importance of communication between co-teaching partners. If

co-teachers need to communicate and plan together to make a collaborative program

work, then they must have a designated time in which to plan. Little and Dieker (2009)

claimed that “a common planning time was essential because as the needs of the students

and co-teachers changed throughout the year, the team needed to revise its instructional

plans” (p. 44). Thousand et al. (2006) also listed common planning time as one of the key

administrative supports of collaborative teaching. In fact, because they have so many


non-teaching duties to fulfill, special education teachers may require more planning time

than other teachers (Gurgur & Uzuner, 2010).

Administration should also model and provide training in effective collaborative

teaching strategies (Cook & Friend, 1995; Thousand et al., 2006). Cook and Friend stated

that attempting to implement co-teaching as a remedy for a bad teacher is a misuse of the

approach. Co-teaching clearly requires competent teachers who are willing to grow

professionally and learn how to meet the needs of all of their students. School

administration should provide teachers with a rationale for collaborative teaching and

consistently provide teachers with practical techniques that will maximize the

effectiveness of both teachers (Thousand et al., 2006).

While admitting that environmental considerations depend largely upon the

individual teaching styles and grade level of co-teaching partners, Hammeken (2000)

enumerated several tips for setting up a classroom environment that is conducive to co-

teaching. First, the two teachers should decide which specific co-teaching strategies they

will use and arrange desks or tables accordingly. Furthermore, students who are receiving

support service should be placed in an area where the special education teacher can reach

them without disturbing the rest of the class. If the teachers plan to implement supportive,

supplemental, or alternative teaching, they should designate an area in or outside of the

classroom for such teaching to take place. Regarding the classroom environment, Cook

and Friend (1995) also suggested that co-teaching partners discuss the classroom

routines, discipline, and the level of classroom noise with which they are comfortable.

Implementing Specific Collaborative Teaching Strategies


The five most commonly used co-teaching models are one teaching, one

observing or assisting, station teaching, parallel teaching, alternative teaching, and team

teaching (Cook & Friend, 1995; Gurgur & Uzuner, 2010; Treahy & Gurganus, 2010).

Thousand et al. (2006), on the other hand, listed the following four broader categories to

surmise the commonly used co-teaching models: supportive teaching, parallel teaching,

complementary teaching, and team teaching. Still, the broader categories contain the

more specific models. Furthermore, teachers seem to develop the strategies in the order

listed above (Cook & Friend, 1995). For example, most teachers begin their career as

collaborative teachers by using the one teach, one observe or assist model. As they

become more comfortable with co-teaching, teachers begin integrating other models into

their co-teaching repertoire, sometimes using two to three structures within a single

lesson. Depending upon the age of the students and the subject matter being taught, each

strategy has a place and a purpose in the classroom (Cook & Friend, 1995).

The first two strategies that new co-teachers most often develop are one teach,

one assist and one teach, one observe. Thousand et al (2006) called these strategies

“supportive teaching (p. 243).” When following this model, one teacher leads the class

while the other teacher observes specific students or provides independent support.

Supportive teaching is useful for collecting data, monitoring student behavior, writing

and evaluating IEPs, checking for individual student understanding, providing individual

instruction (Sileo & Garderen, 2010). One of the major drawbacks of this model is that it

minimizes the role of the supporting teacher and can make him or her feel like a glorified

teaching assistant (Cook & Friend, 1995). To offset this attitude, Cook and Friend

suggested that the teachers alternate roles when using this model. Thousand et al. (2006)


also warned that the supportive teacher should not call attention to special needs students

by constantly hovering around them. According to Gurgur and Uzuner (2010), many

teachers rely on the supportive models as their sole collaborative teaching strategies.

This, they claimed, is often due to a lack of training and planning time. After observing

this phenomenon, Patterson et al. (2009) warned that teachers must rely less on the

traditional, lecture-heavy style of teaching that is conducive to the supportive model and

incorporate other models into their classroom instruction.

Cook and Friend (1995) described station teaching as a co-teaching model in

which teachers divide the students into two or more groups. Each teacher teaches his or

her part of the material to every group of students. When using this method, teachers

often create an independent station along with the two teacher stations. Cook and Friend

also noted that either the teachers or the students can rotate. Thousand et al. (2006)

lumped station teaching together with parallel teaching because of the similarity between

the two models. The major distinction between station teaching and parallel teaching is

that the station teaching model requires teachers to plan distinct lessons and teach them to

two or more groups of students while the parallel model entails teachers to both teach the

same lesson only once to their half of the class (Cook & Friend, 1995; Thousand et al.,

2006; Treahy & Gurganus, 2010). According to Cook and Friend, the major advantage of

both station and parallel teaching is that both models lower the student-teacher ratio. Both

models also give both teachers well defined roles and involve both teachers equally

(Cook & Friend 1995 ). Thousand et al. (2006) warned that teachers using this model

should be careful not to create a “special class within a class by routinely grouping the

same students in the same group with the same co-teacher (p. 244).” This, they warned,


could stigmatize certain students, particularly if the special needs students are always

placed in the same group. Instead, the students should be grouped heterogeneously and

rotated among the co-teachers. Cook and Friend (1995) pointed out that both teachers

must be able cope with a more a moderate to high level of noise and activity when using

either of these approaches.

The fifth co-teaching strategy that teachers develop is alternative teaching.

According to Sileo and Garderen (2010), alternative teaching involves “one teacher

[teaching] a small group of three to eight students while the other teacher teaches the

whole class (p. 16).” Alternative teaching benefits students by providing small groups

with intense and individualized instruction in a specific area (Sileo & Garderen, 2010).

After observing this particular benefit, Treahy and Gurganus (2010) claimed that

mathematics content is particularly conducive to alternative teaching. Cook and Friend

(1995) noted that alternative teaching is well suited for both pre-teaching and re-teaching

before noting a possible drawback of this model. The greatest risk of alternative teaching,

they claimed, is that it could call attention to the students who are pulled aside. To

alleviate this risk, they suggest that teachers vary the small group and purposefully

include every student in the group at some point. Gurgur and Uzuner (2010) suggested

that co-teachers do not attempt to implement alternative teaching until they are

comfortable working together.

The final co-teaching model that teachers tend to develop is team teaching.

When using this structure, both teachers lead the entire class simultaneously (Cook &

Friend, 1995). The teachers may alternate teaching or one may teach while the other

teacher models questioning or note taking techniques. The team teaching model provides


teachers with the opportunity to teach the areas of the content with which they feel more

competent (Thousand et al., 2006). Allowing teachers to teach to their strengths in this

manner benefits both the students and the teachers. Teachers tend to use the team

teaching model only after they are very comfortable with each other and the idea of co-

teaching in general (Cook & Friend).

Measuring Student Outcomes in Collaborative Teaching Classrooms

Perhaps the strongest consensus amongst the research is that the students are the

ones who benefit from an effectively run collaborative teaching classroom (Thousand et

al., 2006; Little & Dieker, 2009; Cook & Friend, 1995). Research measuring the exact

degree of benefit, however, is limited (Cook & Friend). Regarding this Thousand et al.

asserted that “collaborative planning and teaching can result in a variety of positive

outcomes for the kids we have today as well as the educators who teach these children

and youth (p. 246).” Patterson et al. (2009) explained that students benefit more from

collaborative teaching when teachers implement a variety of co-teaching models and use

research based teaching methods such as collaborative learning and differentiated

instruction. Sileo and Garderen (2010) also noted the importance of teachers using a

variety of co-teaching models and effective teaching strategies.

Thousand et al. (2006) discussed several studies that document student outcomes

in collaborative classroom environments. A 2003 study conducted in sixteen California

elementary, middle, and high schools showed that collaborative teaching resulted in

increased overall student achievement, fewer referrals to intensive special education

services, and fewer behavior referrals. Two other studies, one conducted in 1997 and one

conducted in 2000, demonstrated positive improvements in academics and social skills


for special needs students. Another study cited by Thousand et al. suggested that co-

teaching can also increase performance on high-stakes assessments. Several high schools

in Shelby County, Tennessee implemented collaborative teaching, opting to include

seventy percent of the special needs population in general education classrooms. After

only a year of co-teaching, the percentage of special needs students passing a statewide

standardized test increased from twenty percent to forty percent.

Research also addresses the common concern that including special needs

students in the general education classroom hurts the students without disabilities. Villa

and Thousand (2005) emphatically stated that including students with disabilities does

not hinder the performance of general education students in standardized testing or report

cards. Collaborative teaching increases success for all students by expanding instructional

approaches (Cook & Friend, 1995). As Cook and Friend explained, each teacher brings

his or her own unique perspectives and teaching styles, which helps the classroom

instruction appeal to a wider variety of learning modalities. At risk students, in particular,

benefit from the increased instructional time and individualized attention that the lowered

student-teacher ratio provides (Sileo & Garderen, 2010). Time on task for all students

also increases in an effectively run co-teaching classroom (Thousand et al., 2006). Even

gifted and talented students may benefit from the increased options and opportunities for

individualized learning that co-teaching provides (Cook & Friend).

Student and Teacher Perception of Collaborative Teaching

As should be the case with any educational initiative, the primary goal of

collaborative teaching is to benefit the students on either an academic or emotional level.

Co-teaching promises to benefit students on the emotional level as well as the academic


level. To illuminate the emotional need for an alternative to the traditional pullout

method, Villa and Thousand (2005) asserted that sending students to special classes sends

them the message that they do not belong or that belonging must be earned. Children’s

need to belong, they continue, is a critical, if not necessary, factor in motivating them to

learn. A collaborative teaching program, therefore, must significantly reduce the stigma

associated with traditional special education services.

Cook and Friend (1995) cited evidence that suggests that students prefer to

receive special education services in the classroom with their peers rather than being

pulled out of the classroom. According to Sileo and Garderen (2010), collaborative

teaching leads not only to improved academic performance, but also to increased self-

esteem and confidence and better developed peer relationships. Mahony (1997) reported

that “for special education students, being part of the large class meant making new

friends” (p. 59). Cook and Friend noted that all of these potential benefits could be

undone should teachers create a classroom within a classroom in which students with

disabilities are constantly pulled to the side to receive their instruction. According to

Cook and Friend (1995), “The stigmatizing of students using this approach can be as

great if not greater than in traditional special education pullout services, and few of the

other benefits of co-teaching accrue to the student” (p. 3).

Teachers’ perceptions of collaborative teaching seem to vary much more than

those of students. Hammeken (2000) described the classroom as an “animated stage” (p.

39) where one adult performs and the students actively participate. When a special

education teacher joins the group, the dynamics of the whole group may change. Not all

teachers, continued Hammeken, are comfortable with another adult in their classroom.


When this apprehension leads, as it often does, to the marginalization of the special

education teacher, he or she may begin to feel like a glorified teaching assistant (Cook &

Friend, 1995). This attitude, Cook and Friend (1995) stated, can easily develop when

teachers rely solely upon the one teach, one observe and one teach, one assist methods.

As teachers become more comfortable working together, “trust will develop”

(Hammeken, 2000, p. 39). Thousand et al., (2006) described studies of collaborative

teaching in which co-teaching partners report being happier and not feeling so isolated,

enhanced professional growth, and a greater since of community within the classroom.

Although general education teachers often enter into co-teaching apprehensive about

working with special needs students (Dingle et al., 2004), Thousand et al. listed more

tolerance for special needs students as a benefit of collaborative teaching.


CHAPTER THREE: METHODOLOGY

Research Design

This study followed the conventions of action research. According to Gillis,

Wilson, and Elias (2010), “action research encourages school personnel to systematically

develop a question, gather data, and then analyze that data to improve their practice” (p.

91). Hendricks (2009) agreed that action research model enables school personnel to

conduct research that leads to school improvement. Action research is ideal for studies,

like mine, that are designed to improve classroom practice because it emphasizes using

data to drive improvements.

To measure the effect of the treatment, I gathered data from three groups: a

treatment group and two control groups. To increase the study’s validity, I used both a

collaboratively taught and a non-collaboratively taught control group. The following

methods were used to collect data: an instructional plan, reflective journal, special

education teacher interview, pre and post tests, and student interviews. The instructional

plan guided classroom instruction and insured that the treatment was correctly

implemented. A colleague analyzed the plan using a rubric. The reflective journal was

coded for themes and used to determine which structures seemed to be working best in

order to guide daily decision making. The special education interview and student

surveys were used to solicit the special education teacher’s and students’ perception of

the treatment. The pre and post tests were analyzed with dependent t-tests and an Anova.

They were used to determine whether positive student outcomes resulted from the study.


Setting

The location of this study was a rural middle school in the West Georgia area. The

school is seventy-four percent white and seventeen percent black. The remaining students

are predominantly Hispanic or multiracial. Twenty-four percent of the students are

economically disadvantaged, and ten percent of the students are considered students with

disabilities. I chose this location because it is the school in which I currently teach.

Before conducting this study, I acquired permission from the county in which the school

is located and filed an IRB to receive institutional permission from Lagrange College.

Subjects and Participant

Over the course of the study, I monitored subjects in three different classes. The

first class comprised the treatment group. This class contained twelve general education

students, seven gifted students, and nine students with disabilities. Six of the students in

the treatment group were black, seventeen were white, and five were Hispanic or other.

The first control class consisted of sixteen general education students and eleven students

with disabilities. Ten of the students in this class were black, fourteen were white, and

three were Hispanic or other. The second control group contained twenty-nine general

education students and one gifted student. Two of the general education students were

classified as English Language Learners. Three of the students in this class were black,

six of the students were Hispanic, twenty of the students were white, and one student was

multiracial. The three groups were chosen primarily because they are classes that I

already teach. Out of the five classes that I teach, I chose these three classes in particular

because they were the easiest to compare. The first two classes were my only two co-


taught classes. The non-inclusion class was chosen because the students in that group

generally performed at a level comparable to the co-taught classes.

The special education teacher also volunteered to be a participant in this study.

She has worked in inclusion classes for several years. This teacher was chosen because

she already worked in the two co-taught classes that were being studied.

Procedures and Data Collection Methods

When conducting this study, I gathered and examined both quantitative and

qualitative data. Table 3.1 displays the data collection and analysis methods and relates

each set of data to the focus question to which it applies.

The purpose of this study was to determine whether using the parallel, station, and

alternative teaching models described by Cook and Friend (1995) would lead to a higher

level of student engagement, higher test scores, or both. The treatment group and both

control groups were studying the Georgia Performance Standards on proportion and scale

factor. Before implementing the co-teaching structures being tested, I administered a

pretest to all three groups of students. The pretest established a baseline, which I used as

a comparison point to determine the level of student improvement at the conclusion of the

study, and provided the data that I needed to group students and more effectively

differentiate instruction. Because the literature recommended using an integrated

approach when implementing collaborative teaching models, the special education

teacher and I alternated between parallel, station, and alternative depending upon the type

of instruction taking place while working with the treatment group. Parallel teaching was


Table 3.1. Data Shell

Focus Question

Literature Sources Data Sources Why do these data answer the question?

How are data analyzed?

What is the most effective way to implement parallel, station, and alternative teaching structures in a co-teach classroom?

Gurgur & Uzuner, (2010).

Hammeken, (2000).

Cook & Friend, (1995).

Thousand, Villa, and Nevin, (2006).

Method:Reflective journal

Special education teacher interview

Instructional plan

Data:Qualitative and nominative

As the co-teacher and I make observations, we will be able to determine which implementation strategies work best for our students

Observational notes will be compared to the notes taken in other classes in order to determine which structures and implementation methods seem to work best.

Qualitative data will be coded for themes

Descriptive statistics

What effect do co-teaching strategies have upon student achievement?

Patterson, Connolly, & Ritter, (2009).

Little & Dieker, (2009).


Method:Teacher created pre and post tests

Data:Interval

Pre and post test will serve as both a baseline to determine students’ current knowledge and a benchmark to show what students have learned.

Dependent t-tests, an ANOVA, test-retest correlation, and effect size, and descriptive statistics

How well do students and teachers adapt to this new type of instruction?

Villa & Thousand, (2005).

Treahy & Gurganus, (2010)


Method:Student survey

Teacher interview and reflective journal

Data:Nominal and qualitative

These data collection methods are the ideal sources for gathering data related to perception, engagement, and enjoyment

These data will be combined and synthesized.A chi square and descriptive statistics may be used to compare the quantitative part of the student surveys.

Qualitative data will be coded for themes

mostly used to introduce completely new content, particularly the content with which

students seemed the least familiar. When using this method, I divided the students into

two equally sized heterogeneous groups, both of which received nearly identical

instruction. Station teaching was mostly used to teach more familiar new material,


reinforce previously taught content, or allow students to practice working with the

content. I generally led a station that, depending upon the group, reinforced or enriched

the material while the special education teacher monitored two or more student practice

stations. Alternative teaching was used to review material with small groups of students

after assessment data had demonstrated that they still lacked understanding. While one

teacher worked with small groups in the back of the room, the other teacher reviewed the

warm up with the rest of the class. The special education teacher and I alternated roles

when using alternative teaching. At the conclusion of the study, I administered a post test

that was identical to the pretest.

I compared the treatment group to the two control groups. The first control group

was another inclusion class. In this class, the co-teacher and I relied on the following two

structures that the literature claims develop first in a new co-teaching partnership: one

teach, one observe and one teach, one assist. I structured the first control group in this

way because I wanted to determine whether working to develop additional strategies

would actually lead to improved student outcomes. The second control group was not an

inclusion class and had only one teacher. I monitored this additional group because I

wanted to be able to compare the treatment group to a group where co-teaching was not

an option. Specifically, I wanted to be able to compare the performance of the subgroups

of on-level students and advanced students in the collaborative class with their

counterparts in the non-collaborative class.

To address the first focus question in this study, I mapped out the treatment in an

instruction plan. Before implementing the plan, I asked a colleague to evaluate the plan


and make suggestions. The revised plan and evaluation rubric are located in Appendixes

A and B respectively.

Focus question two was concerned with student outcomes. The pre/post test is

included in Appendix C.

The third focus question was concerned with student and teacher attitudes. To

better understand the students’ perceptions, I administered a survey to both co-taught

classes. The survey consisted of two parts; The first of which contained questions on the

Likert Scale to generate nominal data, and the second contained short answer questions

that generated qualitative data. Teacher perception was solicited by a reflective journal

and an interview. The student survey is attached in Appendix D; Reflective journal

prompts and interview questions are included in Appendixes E and F respectively.

Validity, Reliability, Dependability and Bias

To determine the best implementation method for the co-teaching strategies being

tested--focus question one--I gathered and analyzed data from the following sources: an

instruction plan and accompanying rubric and interviews, a reflective journal, and an

interview with the special education teacher. The reflective journal and special education

teacher interview were both used to generate data to address both the first and the third

focus questions. For the first focus question, I gathered evidence through a written

instructional plan. After constructing the plan, I asked two colleagues, another math

teacher and an assistant principal, to analyze it via a rubric and discuss the plan’s merits

and needed improvements in a personal interview. I also referred to comments and notes

from both my experience and that of the accompanying teacher that specifically referred


to the implementation elements that seemed to be effective and those that did not seem to

be effective within the study.

The data collections methods described above generated both qualitative and

quantitative data. Hendricks (2009) describes qualitative data as being very conducive to

action research because it allows researchers to study natural phenomena by “spend[ing]

time in the field observing, talking to people, and analyzing artifacts and products of the

setting under study” (p. 2). Because they generally addressed observations regarding

successful implementation, most of focus question one’s data were qualitative. A minute

amount of nominal data were gathered in response to this focus question in the form of

checklists used in the reflective journal. Meijer and Oosterioo (2008) infer that nominal

data are data that can be separated into distinct categories.

To ensure the content and construct validity of focus question one’s data, I asked

two colleagues to examine my instructional plan prior to its presentation and solicited the

special education teacher’s thoughts after instruction. Popham (2008) describes a similar

external review process as an ideal method for measuring content validity. To measure

content validity, I asked my colleagues to ensure that the learning activities and

assessments within the plan measured the specific learning objectives. To measure

construct validity, which according to Popham (2008), refers to whether the ideas being

tested are actually going to be measured, I asked my colleagues to ensure that the

learning activities would indeed measure the effectiveness of the co-teaching strategies

being tested. To ensure that the qualitative data were dependable, I followed the

procedures given by Hendricks (2009) for collecting and analyzing qualitative data

including coding for the data for themes, presenting interview transcriptions to


interviewees for verification, and selecting test groups that were similar enough to

exclude extraneous variables. Steps were also taken to ensure that the implementation of

the plan was not biased. Popham (2008) explains that bias occurs when a group of

students is unfairly penalized because of the students’ gender, race, ethnicity,

socioeconomic status, religion, or other group-defining characteristics. Following

Popham’s advise, I endeavored to exclude all offensiveness and unfairness and checked

results for any disparate impact.

Data for focus question two, the effect of the treatment upon student outcomes,

were drawn from teacher made pre and post tests. According to Kosheleva, Kreinovich,

Longpre, Tchoshanov, and Xiang (2007), data that are derived from numerical grades

constitutes interval data. To ensure content validity within the tests, I started my search

for appropriate test questions by carefully studying the standards being measured. After

feeling that I really understood the standards, I carefully chose several test questions that

I felt measured each objective. Popham (2008) terms this process of careful test creation

“developmental care” (p. 56). To ensure my pre/post post test would accurately predict

student performance on the Georgia CRCT, I gathered my test questions from a bank of

CRCT example questions. Popham uses the term “criterion validity” to refer to this

process (p. 60). The two final factors that I considered when collecting data for focus

question two were reliability and bias. Popham explains that reliability refers to the

consistency of a testing instrument. To ensure the reliability of the test results, I

administered identical pre and post tests. Having used identical tests, I could be more sure

that any perceived changes in student outcomes were due to the treatment and not to

variations within the pre and post tests. Another measure that I took to ensure reliability


was to run a test-retest correlation between the pre and post tests for each group. To

address issues of bias, I examined the testing instrument and determined that it was fair

and was not offensive. After grading the tests, I examined the results for evidence of

disparate impact.

The third focus question, which was concerned with teacher and student ability to

adapt to this type of instruction, was examined using the previously mentioned interview

and reflective journal as well as a student survey. The journal and interview both

generated qualitative data, and the survey generated qualitative and nominal data.

Hendricks (2009) describes data that results from observations such as the journal,

interview, and written portion of the survey, as qualitative. According to Sturges (1978),

survey questions that are based upon the Likert Scale generate nominal data. To ensure

construct validity of the survey instrument, I modified survey questions that had been

used in a published study. I used Cronbach’s alpha to maintain the reliability of the

nominal data. To address dependability issues, I presented the special education teacher

with a transcription of our interview. All survey and interview questions were checked to

rule out issues of bias including unfairness and offensiveness.

Analysis of Data

The qualitative data collected for focus question one, which comprised most of

that question’s data, were coded for themes. I analyzed the feedback for my instructional

plan, my reflective journal, and the teacher interview for any recurring, dominant, and

emerging characteristics. In doing so, I noted all categorical and repeating data that

formed patterns of behavior. I used descriptive statistics to examine the minimal amount

of quantitative data generated by the reflective journal.


The data related to focus question two were solely quantitative. Prior to the study,

I used an ANOVA to analyze the pretests to determine whether the classes were similar

enough to validate the study’s outcomes as well as to provide a benchmark to help

determine student growth. At the study’s conclusion, I ran a second ANOVA to

determine if a significant difference had developed between the three groups that could

be attributed to the treatment. I also ran dependent t-tests between the pre and post tests

of each group and found the effect size of each group’s growth. Pearson’s correlation was

used to add to the reliability of the data. I also ran each statistical test without the outliers

and disaggregated all data by sex to determine whether the structures had a

disproportionate effect upon any particular group. Finally, I used descriptive statistics to

further process the data.

The qualitative data that related to focus question three were coded for themes. I

searched the reflective journal, interview transcription, and short answer survey questions

for categorical and repeating data that demonstrated patterns of behavior. The nominal

data generated by the survey questions based upon the Likert Scale were analyzed using a

chi square

From the study’s inception, through its planning stages, and to its conclusion,

great care was taken to ensure its validity, credibility, and transferability. The Education

Department faculty review process at Lagrange College gave the study what Eisner

(1991) calls consensual validation. To maintain epistemological validation, a term that

Denzin and Lincoln (1998) use to describe the process of cycling back to the literature,

the findings from the study will be compared to the scholarly research written about in

chapter 2. The three facets of credibility that I considered were structural corroboration,


fairness, and rightness of fit. Eisner (1991) used the term structural corroboration to

describe the process of triangulating data to increase the study’s credibility. To ensure the

study’s structural corroboration, I drew data from a variety of sources. While reviewing

the literature, I endeavored to find scholarly work that represented a variety of

perspectives in order to maintain a degree of fairness. Having been careful to follow the

proper research techniques and achieve rightness of fit, I felt that I could build a strong

case to support the study’s conclusions. The aspect of transferability developed within

this paper was referential adequacy, a term Eisner (1991) used to describe the process of

one’s research increasing the perception and understanding of others. To ensure that the

study could be easily replicated by future researchers, I used great detail when explained

the procedures and methods used with the treatment and control groups. A final

consideration associated with the development of this research was ensuring that it was

transformational. The study had catalytic validity, as described by Kinchloe and McLaren

(1998), in that it affected positive change in my teaching practices as well as those of the

special education teacher involved in the study


CHAPTER FOUR: RESULTS

Both quantitative and qualitative data were collected in response to focus question one,

which was “what is the most effective way to implement parallel, station, and alternative

teaching structures in a co-teach classroom?” The quantitative data, which was drawn

from a checklist within the reflective journal, measured the on task behavior of the

students with disabilities and was analyzed with descriptive statistics. The qualitative data

that were collected from the unit plan and accompanying interviews, the special

education teacher interview, and the reflective journal. All qualitative data were coded for

themes. Throughout this section, “control group A” will be used to designate the

collaborative control group, and “control group B” will designate the control group in

which collaborative teaching did not take place.

The minimal amount of quantitative data that was gathered for focus question one

was nominal in nature. After monitoring the time on task of the special education

students in the treatment group and control group A over the course of several days of

instruction, I have found that the students in the treatment group were on task 85% of the

time. The special education students in control group A, however, were only on task 51%

of the time. For the purposes of this study, time on task was considered a state in which

the students being observed were actively working or listening and taking notes. Data of

this type were not collected on control group B due to its lack of special needs students.

Most of the data collected for focus question one was qualitative. The primary

sources of qualitative data were the peer teacher and assistant principal interviews. Two

themes clearly emerged from both interviews. When asked about the shortcomings of the

instructional plan, both my peer teacher and the assistant principal expressed concern


about the lack of “real world application” in the essential questions and learning

activities. “You could alter the essential questions just a little,” said the teacher, “and they

would help the students apply what you are studying to their lives.” The assistant

principal expressed a similar sentiment, “The questions work, but how are they going to

apply [the content] to their real life?” Both interviewees suggested that special needs

students need to know how the material applies to them. They also suggested that I find

more ways to include higher order thinking in the plan. “The only thing that you are

missing are the higher order thinking skills,” remarked the assistant principal, “you

probably do it in your classroom without even thinking about it, but you should put it in

your plan.” She suggested that I look back over my performance assessments and find

evidence of higher order thinking. On the positive side, both interviewees indicated that

the plan was well balanced with teacher led and student directed instruction. They also

remarked that they felt that the skills of both teachers would be fully utilized over the

course of the unit. When asked about what worked well in the plan, the teacher said “It is

well thought out.” “You are using different methods and using both teachers,” she

continued, “you are meeting the needs of each student: the high, the low, and the ones in

the middle aren’t getting lost either.” The assistant principal praised the way that the plan

relied upon preassessment data to determine which strategies to use and how to group

students. “If you are focusing on the preassessment to make decisions about which

strategies to use and how to group students,” she said, “then you are doing it correctly:

just make sure that you continuously cycle back to assessments as you make decisions.”

Several themes emerged from the reflective journal prompts that applied to focus

question one. Excessive transition times were an issue in the treatment group. This


seemed to be especially true on the days in which the special education teacher and I had

less time to plan together. After using the treatment strategies for a couple of weeks,

however, fewer comments concerning transition times were recorded in the reflective

journal. The specific strategies chosen seem to have been appropriate to each situation. A

final note on the implementation of the collaborative teaching strategies was the size of

the classroom. Considering the number of students, the classroom seemed to be a little

too small to effectively group students and provide teaching space for two teachers. There

were many days in which it was difficult to differentiate between the students who were

in my group and the students who were in the special education teacher’s group.

During her interview, the special education teacher also referred to focus question

one. “I feel like the implementation of the co-teaching plan was ideal,” she said,

“especially for that particular class.” Further on in the interview, the special education

teacher expressed that she felt that both the treatment group and control group A had the

best plans for the students in each group. “When planning this type of instruction,” she

explained, “you have to consider the group that you are working with.” “The [treatment

group] can work better in different situations, but [control group A] really needs more

structure.” When asked if she felt like we had adequate common planning time, the

special education teacher said “no, I feel like the majority of our planning time is

absorbed in meetings or trying to get other things done for the day.” She also said that she

has a class that ran through half of what would have otherwise been common planning

time.

All the data that was gathered for focus question two--What effect do co-teaching

strategies have upon student achievement?--were quantitative. Prior to the treatment, I


ran an ANOVA on the pretests. Upon completing the study, I ran another ANOVA as

well as dependent t-tests between the pretests and posttests of the treatment and control

groups. To determine whether the treatment had a disproportionate effect on any one

particular group, I repeated each test with the following groups: without the outliers, boys

only, and girls only. The alpha for each statistical test was set at .05. I also used

descriptive statistics to further analyze the data.

Table 4.1 displays the results of the pretest ANOVA. I found that F(2,80)=1.79,

which meant that the p>.05 and there was no significant difference among the three

groups.

Table 4.1 ANOVA Between Pre-Tests for Control and Treatment Groups

SUMMARY

Groups Count Sum AverageVarianc

e3rd 26 1,105.6 42.523 479.7674th 28 1,501.6 53.629 639.844

5th 29 1,435.3 49.493 305.174

ANOVASource of Variation SS df MS F P-value F crit

Between Groups 1,690.396 2 845.198 1.788 0.174 3.111Within Groups 37,814.822 80 472.685

Total 39,505.218 82

At the study’s conclusion, I ran a posttest ANOVA and a dependent t-test on each

class in its entirety. On the posttest ANOVA (Table 4.2), I found that F(2,80)=.32 and the

probability was greater than .05. There was no significant difference among the three

groups.


Table 4.2 ANOVA Between Post-Tests for Control and Treatment Groups

SUMMARYGroups Count Sum Average Variance

3rd 26 1,747 67.192 523.8614th 28 2,017.7 72.061 691.2625th 29 2,034.8 70.166 280.99


Between Groups 323.331 2 161.665 0.326 0.722 3.111Within Groups 39,628.331 80 495.354

Total 39,951.662 82

The results of the dependent t-test for the treatment group (Table 4.3) are listed

below. Pearson’s correlation, the reliability statistic for this test, was r = .87. After

running the test, I found that t(27)=-7.4 and p<.05. There was a significant difference

between the pretests and posttests of the treatment group. At .34, The effect size was

medium.

Table 4.3 Dependent t-Test between the pretest and posttest of the treatment group

Pretest PosttestMean 53.629 72.061Variance 639.844 691.262Observations 28 28Pearson Correlation 0.87Hypothesized Mean Difference 0df 27t Stat -7.403P(T<=t) one-tail 0t Critical one-tail 1.703P(T<=t) two-tail 0t Critical two-tail 2.052


The dependent t-test for control group A (Table 4.4) yielded the following results:

t(25)=-6.93 and p<.05. There was a significant difference between the pretests and

posttests of this group. At .48, the effect size for this group was large. Pearson’s

correlation for this test was r = .67.

Table 4.4 Dependent t-Test between the pretest and posttest of control group A


Control group B’s t-test (Table 4.5) showed that t(28)=-6.92 and p<.05. There was

a significant difference between the pre and post tests of this group. The effect size for

this group was also large at .52. Pearson’s correlation for this test was r = .56.

Table 4.5 Dependent t-Test between the pretest and posttest of control

group B



After running the above tests, I repeated each of them, this time excluding

students whose pretest grades constituted outlying data. Because the pretests tended to be

low, I considered any grade above an 85 to be an outlier, which meant that any student

who missed more than two pretest questions was included in this data. With the outliers

excluded, the pretest ANOVA (Table 4.6) yielded F(2,74)=3.38 with a probability greater

than .05. There was no significant difference among the groups. The post test ANOVA

for this group (Table 4.7) resulted in F(2,74)=.36 with p>.05 and no significant difference

among the groups.

Table 4.6 ANOVA between the pretests of the control and treatment

groups without outliers


3rd 24 923.3 38.471 297.984th 25 1,225.2 49.008 511.2915th 28 1,347.1 48.111 259.005



Total 27,801.139 76


Table 4.7 ANOVA between the pretests of the control and treatment groups without

outliers


3rd 24 1,552.9 64.704 484.6774th 25 1,723.6 68.944 682.2635th 28 1,946.6 69.521 278.921



Total 35,398.369 76

Without the outliers, the treatment group’s dependent t-test’s results (Table 4.8)

were t(23)=-7.26, which meant that the probability was less than .05 and there was a

significant difference between the pre and post tests. At .38, the effect size for this group

was large. Pearson’s Correlation was r = .87.

Table 4.8 Dependent t-Tests between the pretest and posttest of the

treatment group without outliers

47.1 76.5Mean 49.088 68.629Variance 533.356 709.34Observations 24 24Pearson Correlation 0.869Hypothesized Mean Difference 0df 23t Stat -7.263P(T<=t) one-tail 0t Critical one-tail 1.714P(T<=t) two-tail 0t Critical two-tail 2.069


Control group A’s dependent t-tests (Table 4.9) also showed that t(23)=-7.26 with

p< .05. Pearson’s correlation was r = .61. There was a significant difference between the

two tests. The effect size for this control group was .55. This was a large effect size.

Table 4.9 Dependent t-Tests between the pretest and posttest of

control group A without outliers


Control group B had only one outlier. Pearson’s correlation for this test was r

= .53. With the outlier excluded, control group B’s dependent t-test (Table 4.10) yielded

the following results: t(27)=-7.13, p<.05, and a large effect size at .55.

Table 4.10 Dependent t-Tests between the pretest and posttest of

control group B without outliers

pre postMean 48.111 69.521Variance 259.005 278.921Observations 28 28Pearson Correlation 0.532Hypothesized Mean Difference 0df 27t Stat -7.135P(T<=t) one-tail 0t Critical one-tail 1.703P(T<=t) two-tail 0t Critical two-tail 2.052


I also disaggregated the data by sex to see if boys and girls responded differently

to this type of instruction. The pretest ANOVA for the girls (Table 4.11) showed that

there was no significant difference among the three groups. On this test, F(2,31)=1.07 and

p>.05. The boys’ pretest ANOVA (Table 4.12) showed that F(2,47)=1.55 with a probability

greater than .05, which meant that there was no significant difference among the groups.

Table 4.11 ANOVA between the pretests of the girls in the control and

treatment groups


3rd 10 423.4 42.34 306.2784th 9 447.1 49.678 511.4375th 15 829.2 55.28 547.19



Total 15,514.527 33

Table 4.12 ANOVA between the pretests of the boys in the control and treatment groups


3rd 16 1,129.3 70.581 553.744th 19 1,370.6 72.137 883.9255th 15 1,135.5 75.7 137.63



Total 26,356.417 49


Neither of the posttest ANOVAs showed a significant difference among the

groups. The girls posttest ANOVA (Table 4.13) resulted in F(2,30)=.65; p>.05, and the

boys posttest ANOVA (Table 4.14) resulted in F(2,47)=.19; p>.05.

Table 4.13 ANOVA between the postests of the girls in the control and

treatment groups


3rd 10 617.7 61.77 479.1824th 9 647.1 71.9 344.1375th 14 899.3 64.236 383.782



Total 12,584.855 32Table 4.14 ANOVA between the posttests of the boys in the control and

treatment groups


3rd 16 1,129.3 70.581 553.744th 19 1,370.6 72.137 883.9255th 15 1,135.5 75.7 137.63



Total 26,356.417 49

I also ran dependent t-tests that were disaggregated by sex. Pearson’s correlation

was r = .67 for the treatment group’s girls and r = .93 for the treatment group’s boys,


r = .61 for control group A’s girls and r = .72 for this group’s boys, and r = .32 for control

group B’s girls and r = .13 for this group’s boys. The dependent t-tests for the treatment

group girls (Table 4.15) showed t(8)=-3.9 with p<.05 and a large effect size of .47. There

was a significant difference in this group. The treatment group’s boys’ t-test (Table 4.16)

resulted in t(18)=-6.63, p<.05, and a medium effect size of .28. There was also a significant

difference between the boys’ pre and post tests.


girls in the treatment group

pre postMean 49.678 71.9Variance 511.437 344.137Observations 9 9Pearson Correlation 0.671Hypothesized Mean Difference 0df 8t Stat -3.897P(T<=t) one-tail 0.002t Critical one-tail 1.86P(T<=t) two-tail 0.005t Critical two-tail 2.306Table 4.16 Dependent t-Tests between the pretest and posttest of the boys in the

treatment group



The dependent t-tests for the girls of control group A (Table 4.17) resulted in t(9)=-

3.46, p<.05, and a large effect size of .44. The boys’ dependent t-test (Table 4.18) showed

t(15)=-6.15, p<.05, and a large effect size of .5. Significant differences in the pre and post

test were observed in both the boys and the girls of control group A.


girls in control group A

pre postMean 42.34 61.77Variance 306.278 479.182Observations 10 10Pearson Correlation 0.613Hypothesized Mean Difference 0df 9t Stat -3.456P(T<=t) one-tail 0.004t Critical one-tail 1.833P(T<=t) two-tail 0.007t Critical two-tail 2.262Table 4.18 Dependent t-Tests between the pretest and posttest of the boys in control

group A


There was no significant difference in the dependent t-tests for the girls of control

group B (Table 4.19). This test yielded the following results: t(13)=-1.16 and p>.05. The


boys, however, did show a significant difference on the dependent t-test (Table 4.20).

The results for this test were t(14)=-4.51, p<.05, and a large effect size of .61.


girls in control group B

pre postMean 56.286 64.236Variance 572.943 383.782Observations 14 14Pearson Correlation 0.318Hypothesized Mean Difference 0df 13t Stat -1.159P(T<=t) one-tail 0.134t Critical one-tail 1.771P(T<=t) two-tail 0.267t Critical two-tail 2.16Table 4.20 Dependent t-Tests between the pretest and posttest of the boys in control

group


After running the ANOVAs and dependent t-tests, I used descriptive statistics to

further analyze and process the data. The percentage of students who scored below 70

percent and failed the pretest and above 70 percent the second time, thus passing the

posttest, was 56% for the on level students in the treatment group, 38% for the special

education students in the treatment group, 74% for the on-level students in control group

A, 0% for the special education students in control group A, and 58% for the students in


control group B. 84% of the on-level students and 45% of the special education students

in the treatment group scored above 70 percent on the posttest; 79% of the on-level

students and 0% of the special education students in control group A scored above 70

percent on the posttest; 62% of the students in control group B scored above 70 percent

on the posttest. The number of students exceeding the standard on the posttest in the

treatment group and in control group A was 11% and 5% respectively. None of the

special education students in either group exceeded the standard on the posttest.

Likewise, all the students in control group B failed to exceed the standard. Ninety- two

percent of the boys in control group B scored above 70 percent on the post test while

only 69% of the boys passed in the treatment group and and 63% of the boys in control

group A passed the test. Only 7% of the girls in control group B passed the posttest,

while 10% of the girls of control group A passed the test. 33% of the treatment group

girls, on the other hand, passed the posttest. At 82%, the percentage of boys that went

from below 70 percent on the pretest to above 70 percent on the posttest was greatest in

control group B. Control group A was the second highest at 54%, and the treatment group

boys were last at 45%. With a 50% pass rate, the percentage of girls who went from fail

to pass was highest in the treatment group. Control Group A was second with a 44% pass

rate, and control group B was last with a 38% pass rate. The number of African American

students who went from fail to pass was also highest in the treatment group. Sixty-seven

percent of these students failed the pretest and passed the post test. This is compared to

control group A in which 38% of African American student went from fail to pass, and

control group B in which 50% of the African American students went from fail to pass.


Quantitative data for focus question three, which was concerned with how well

students and teachers adapt to this new type of instruction, were collected from student

surveys, which were analyzed with a chi square. Qualitative data were gathered from the

student surveys, reflective journal, and the instructional plan. All qualitative data were

coded for themes.

The quantitative data for focus question three came from Likert Scale surveys.

The surveys were only given to the students in collaborative classes, so the students of

control group B were not invited to participate. I used a Chi Square to analyze the student

surveys. Table 4.20 displays the results of this section of the survey. Cronbach’s Alpha

for the treatment group’s responses was .55; Cronbach’s Alpha for the control group’s

responses was .65.

Table 4.21: Chi Square

n = 42 χ² (Treatment)

χ² (Control Group A)

Item 1: Having two math teachers has helped me better understand math

17.19** 15.73**

Item 2: I enjoy the activities that I am able to do in my class(es) with two teachers more than the activities that I do in my classes that only have one teacher.

17.67*** 6.64

Item 3: My math class is better because it has two teachers.

11.14* 15.73**

Item 4: My grades in math this year are better than they were last year.

4.75 0.73

Item 5: My behavior is better because I have two teachers in my math class.

12.9** 7.5

Both the treatment and control group trended toward significance in agreement in

their response to item one. On item one, the chi square analysis yielded χ² (4) = 17.19,

p<.01 for the treatment group and χ²(4) = 15.73, p<.01 for the control group. The majority


of the students in both the treatment and control groups indicated that learning in a two

teacher classroom has helped them to better understand math. Both groups also trended

toward agreement in response to item three claiming that their math class was better

because it had two teachers. With two asterisks, the treatment group showed a stronger

trend than that of the control group, which had only one asterisk. The analysis for the

treatment and control group showed χ²(4) = 11.14, p<.05 and χ²(4) = 15.73, P< .01

respectfully.

Unlike the control group, the treatment group also showed trends on items two

and five. Item two, which stated that students enjoyed the activities more in collaborative

classrooms than they did in single teacher classrooms, showed a three asterisk trend

toward agreement. The chi square analysis for item two yielded χ² (4) = 17.67, p<.001.

Item five, which was concerned with the students’ perception of their own behavior in

collaborative classrooms, trended toward the middle and strong agreement. The chi

square analysis for item five showed χ²(4) = 12.9, p<.01 with two asterisks.

Five free response survey questions generated qualitative data, all of which were

coded for themes. A single theme emerged from both groups’ responses to the first

question on the survey. When asked about the best part of having two teachers in their

classroom, the majority of the students in both groups commented on the extra help

afforded by the second teacher. “the reason why I like having two teachers in one class,”

writes one control group student, “is because [they] can help more than one student.”

According to another control group student, two teachers provide “double the teaching.”

Another student from this group said “I can get more one on one [instruction].” Several

students from the treatment group expressed a similar sentiment. “If I have a question,”


asserts one student, “it’s easier to have it answered [because] there are more people

around to answer it.” In response to question one, two other themes emerged only from

the treatment group. The first centered upon students’ appreciation of the skills of both

teachers. One treatment group student wrote “I may not get it from one teacher, but when

I ask the other I get it more.” “Because if one teacher does not explain it as well, then you

can go to the other one and see if they explain it better,” adds another student. The final

theme coming from the treatment group’s response to question one concerned students’

appreciation of the grouping strategies used in the treatment. One student remarked “I

think I like that we can split up and learn different things at a time and still get it,” and

another student liked that “[the teachers] can teach more lessons at once.”

The second free response survey question was “what parts did you not like about

having two teachers?” Although the control group had a few more dissenters than the

treatment group, an overwhelming majority of students in both groups either left this

question blank or indicated that there was not anything that they disliked about having

two teachers in their classroom.

The third free response survey question asked students why they thought that two

teachers were in their math classroom. Only one of the treatment group’s respondents

asserted that the other teacher was in the classroom to help “certain students.” Most

respondents within the treatment group indicated that they thought that two teachers were

in the room to either help all students or to help each other teach. Within the control

group, however, several respondents suggested that the other teacher was in the room to

help specific students only. According to one control group student, the other teacher was

in the room “because some of the people need more help than others.” While many


responses were very general expressions of this sentiment, one of the control group’s

students overtly wrote that there were two teachers “because our class needs extra help

because some people are not as smart as others.” Student responses to the final two

survey questions did not yield viable qualitative data.

At the study’s conclusion, I used a personal interview to gather qualitative data

from the special education teacher who participated in the study. When asked about her

opinion of the effectiveness of the treatment, she stated that she believed the collaborative

structures being studied seemed to work well overall. Even so, she indicated that she

actually preferred using the one teach, one assist structure that was implemented with

control group A. “There is a lot of redirection necessary,” she said, “so it almost works

out better that one person is teaching.” She elaborated, “I like to allow the content teacher

to teach the material while I put out fires and deal with things that are going on in the

classroom.” She explained that she believed doing so frees the content teacher to “teach

the material that [he or she] knows so well.” After describing the collaborative structure

that she prefers to use, the special education teacher listed the pros and cons of each of

the other structures. About parallel teaching, she said “I like this, but I just can’t find it

effective.” “The classroom,” she explained, “is just too small, and the noise level makes it

difficult for me to get my point across.” She did say that parallel teaching seemed to work

best on the days that one of the groups worked outside of the classroom. Station teaching

seemed to be the special education teacher’s least favorite structure. Concerning student

led stations, the teacher asserted “I often feel out of control.” Continuing she said, “[the

slower students] often wait for the brighter students to get the answers and then copy off

their paper--I really don’t know who is doing what.” She said that she liked that they


were learning on their own, but she did not like the lack of individual accountability. Out

of the three strategies being tested, alternative teaching seemed to be the special

education teacher’s favorite. “The students really like that,” she exclaimed, “ they liked

being pulled out to review their tests when they made poor grades.” They didn’t just wad

up their paper and throw it away,” she explained, “because they knew that everyone in

the room was in the same boat that they were in.” She asserted that the students were

more focused and more willing to engage in discussion when pulled to the side in a small

group.

The final qualitative data for focus question three were collected through my

reflective journal, which was coded for themes. The following themes emerged from the

reflective journal: increased level of student engagement within the treatment group,

improved energy level and outlook on teaching when using the structures with the

treatment group, the importance of communication within collaborative classroom, and

the level of noise when using the parallel, station, and alternative teaching structures.

Within the reflective journal, I found several comments centered upon student

engagement. Students in the treatment group showed more focus during instruction and

independent work time than students of the two other groups. Furthermore, because many

of the structures allowed for more peer cooperation, the control group students were

considerably less reliant upon the teachers for help. The reflective journal also indicated

that the students in control group A, which was co-taught, were more focused and

engaged that the students in control group B, which had no co-teacher. I also found

several references to my energy level and outlook upon teaching in the reflective journal.

I noted that I felt better while teaching the smaller groups of students within the treatment


group. It was much easier for me to focus on the learning needs of the students who were

in my smaller group than when I had to focus upon the large groups of students who were

in the two control groups. At the end of the first week of the study, I wrote “I just feel

better about working with [the treatment group].” I elaborated, “I feel like I am producing

greater results with much less effort when compared with the other two groups.” Another

recurring theme within the reflective journal concerned the effect of poor communication

between the teaching partners upon the outcome of each strategy. “We didn’t seem to be

on the same page today,” I wrote on one particularly unsuccessful day of instruction.

Particularly at the beginning of the study, I wrote several times about having an excessive

amount of down time while the special education teacher and I made sure that we both

knew what was going to happen. At the time, I recorded that I felt like we should have

discussed the logistics and content before each day of instruction. There were also many

days in which classroom noise was an issue. In general, the noise was not negative.

Instead, the noise came from both teachers talking at the same time or from students

working in groups. Still, this was noted as a problem when one group was teacher

directed and the other was student centered.

CHAPTER FIVE: ANALYSIS AND DISCUSSION OF RESULTS


Analysis of Results

This action research study was designed to test the effect of using parallel, station,

and alternative collaborative teaching structures in a sixth grade mathematics classroom.

The three focus questions that guided the research process were concerned with

discovering how to implement the structures, student learning outcomes, and teachers’

and students’ appreciation of and ability to adapt to working in a classroom in which the

three structures were being used.

Focus question one was concerned with the process of implementing the three

structures in a collaborative classroom. To research this, I first created a unit plan and

interviewed another math teacher and my assistant principal about the plan’s deficits and

merits. Throughout the study, I also recorded reflective information about classroom

implementation in a daily journal. Finally, I interviewed the special education teacher

who participated in delivering the treatment to gather her thoughts on the

implementation. Because nearly all data were qualitative, I coded most of the data for

themes. The minimal amount of quantitative data were analyzed solely by using

descriptive statistics.

Throughout the entire research process, a primary consideration when gauging the

success of the implementation of the collaborative structures was the on-task behavior of

what is, for me, the most difficult group of students to keep on task: students with

disabilities. After carefully monitoring this group of students, I was able to determine that

the students with disabilities in the treatment group were on-task about 85% of the time,

while the students with disabilities in control group A were only on-task about 51% of

the time. The shortcomings of this method were related to reliability and bias. To


alleviate these issues as much as possible, I alternated with my collaborative teaching

partner when collecting the data. The checklist, combined with my own broader

observations, left me convinced that the implementation of the structures being tested led

to a significant increase in student engagement within the students with disabilities

subgroup.

The major sources of qualitative data for focus question one were the peer teacher

and assistant principal interviews regarding the unit plan. Two themes clearly emerged

from both interviews concerning its shortcomings. One theme was the absence of clear

real-world application. The two interviewees explained that all students need to know

how the content applies to their life. This, they said, is especially true for special

education students. Real-world application, therefore, is an important component of any

classroom instruction, particularly in collaborative classrooms, regardless of the

structures and strategies being used. A second theme that I discovered from the

interviews was the importance of higher order thinking. Noticing the performance tasks

listed in the plan, the assistant principal said that I should look within them for evidence

of higher order thinking. She said that higher order thinking was key to mathematical

success for all students. Most teachers, she asserted, use higher order thinking, but do not

properly document it.

Several other themes emerged from the interviews while we were discussing the

positives in the instructional plan. The first theme discovered was the importance of

consistently using the instructional skills of both teachers throughout instruction. When

two teachers work in a classroom together, both teachers should feel equally important

and valued. Perhaps more important, the students should perceive teachers as being equal


in the classroom. Any instructional plan in a collaborative classroom should lead to the

above results. The two interviewees also discussed the importance of properly balancing

teacher led and student directed instruction. Special education students--as well as most

other groups of students--need a certain degree of structured, explicit, and direct

instruction, but they also need time to work, explore, and interact with the content

themselves and in small groups. A well designed instructional plan makes allowances for

both teacher centered and student centric instructional activities. A final theme that I

discovered during the assistant principal interview was the importance of assessment data

when planning for instruction. According to her, instruction should be framed around pre-

assessment data. Pre-assessment data should also be a primary consideration when

grouping students into the proper parallel and station groups. Formative assessment data

should also be relied upon throughout instruction.

I also relied upon the reflective journal and collaborative teacher interview as

secondary data sources for focus question one. One consideration from the reflective

journal was transition times. This applied to implementation because it seemed to have

resulted from inadequate planning. Before attempting to implement parallel, station, and

alternative teaching structures, teachers may need to devote additional time to

collaborative planning to group students and ensure that all teachers are on the same

page. This would marginalize or alleviate many problems including those involved with

lengthy transition times. A final consideration for teachers planning to successfully

implement parallel, station, and alternative teaching structures that was consistently noted

in the reflective journal is space. Due to a large class size and the small size of the

physical classroom, it was often difficult to arrange groups of students and maintain


appropriate volume levels so as not to disrupt the other teacher. While teachers facing

similar obstacles should not be deterred from using these structures, they must carefully

consider the physical environment of the classroom when planning instruction.

The special education teacher involved in the study reiterated the issues involved

with an inadequate amount of planning time. She also discussed the importance of

considering each group of students when choosing instructional techniques. According to

the interview, some groups of students, like the treatment group, respond very well to less

traditional instruction, but many groups of students, like the first control group, need

more traditional instruction. The primary characteristics that she said led to the successful

implementation within the control group were the group’s overall behavior and the

students’ willingness to help each other and work together. A class that does not display

those particular characteristics may not respond to parallel, station, and alternative

teaching as well as the treatment group responded in this study.

Focus question two was concerned with student learning outcomes. All of focus

question two’s data were drawn from pretest and posttest grades, so all the data were

quantitative. After running ANOVAs on the pretests and posttests, I found that the

treatment had not led to a significant difference among the groups. I also ran dependent t-

tests between each group’s pretest and posttest. At r = .87, Pearson’s correlation was

highest for the treatment group; control groups A and B had medium correlations at r

= .67 and r = .65. The correlations were all high enough to mark all of the pretest and

posttest data as reliable. While the dependent t-tests for all three groups demonstrated that

each group had made significant gains, the group with the largest effect size was control

group B. This group had a large effect size of .52. At .48, The second largest effect size


came from the other control group. With an effect size of .34, the treatment group had the

only medium effect size. When taken as a whole, the treatment seems to have led to

fewer gains than the control. Furthermore, the control group in which collaborative

teaching was not taking place made the most academic gains overall. Although these data

are significant, they are not the most important indicators of the effect of the treatment

upon student outcomes.

Because of several very high pretests within the treatment group, the most

significant evidence of the treatment’s effect came from running the data with outlying

pretest and their corresponding posttest grades excluded. Even with these data excluded,

the ANOVAs still did not demonstrate a significant difference among the groups. Again,

I ran dependent t-test for each group. Pearson’s correlation for this set of tests was nearly

identical to the reliability value of the previous set of tests. There was, once again, a

significant difference between the pretests and posttests of each group. While the

treatment group’s effect size was large on this t-test, it was still the smallest at only .38.

The effect size of the two control groups also went up and were now equal at .55. The

treatment group had still demonstrated the smallest gains. Both control groups, however,

seemed to have demonstrated remarkably similar learning outcomes. These data seem to

suggest that parallel, station, and alternative teaching have little effect on overall student

learning outcomes. If anything, the implementation of these structures seems to have led

to decreased growth in treatment group students.

The data was also disaggregated by sex to determine if the treatment had been led

to greater student gains in either the boys or the girls. Like before, the pretest and posttest

ANOVAs for the boys showed no significant difference among the three groups. After


running the two ANOVAs, I used dependent t-tests to examine the pretest and posttest

data of each group. The boys in all three groups had demonstrated academic gains. With

a very high Pearson’s correlation of r = .93, the treatment group’s data was the most

reliable. Control group A’s data was also very reliable with a correlation of r = .72. At r =

.13, the data for control group B were not as reliable. Following the trend set by the

previous two sets of tests, the effect size was smallest for the treatment group, whose

effect size was .28. The largest effect size was found in control group B, which had an

effect size of .61. This was followed by control group A, whose effect size was an even

0.5. Although the reliability of control group B’s data may be questionable, the data

would seem to indicate that learning in a collaborative environment may not be ideal for

boys. When in a collaborative classroom, the boys in this study seemed to perform better

when the class was more structured and directed by a single teacher. Using the parallel,

station, and alternative collaborative teaching structures did not seem to work for the

majority of the boys involved in this study.

Like the boys, the ANOVAs for the girls‘ pretest and posttests showed no

significant difference among the groups. The dependent t-tests, however, did not follow

the pattern shown by nearly every other test. With reliability statistics of r = .67 for the

treatment group, r = .61 for control group A, and r = .32 for control group B, most of the

data were reliable. The girls in both the treatment group and control group A showed a

significant growth between their pretests and posttests. The girls of control group B,

however, did not show significant growth. The girls were also the only group in which

the effect size was larger in the treatment group than in the other groups. The treatment

group’s effect size was .47; control group A’s effect size was slightly smaller at .44. The


two classrooms with collaborative teaching produced better results for the girls involved

in this study. Although the data may be too close to make a bold determination, the

treatment seems to have produced greater growth in the female students involved in the

study than the single teacher led control group.

Although the most useful data were derived from the statistical tests described

above, a few observations drawn from the use of descriptive statistics are worth noting in

this paper. Although the treatment group did not demonstrate more growth than the two

control groups, the students without disabilities in this group had the highest percentage

of students who passed the post test. At 84%, this group’s pass rate was 5% higher than

control group A’s and 22% higher than control group B’s pass rate. The treatment group

also had a pass rate of 45% within the students with disabilities subgroup while control

group A had a 0% pass rate within this subgroup. Black students and female students also

had a much higher pass rate and fail to pass rate within the treatment group. One group

that displayed especially interesting results was male students. On the pretest, the

treatment group had the highest percentage of boys with passing grades. This group had a

pass rate of 42%, which was 23% higher than control group A and 15% higher than

control group B. On the post, however, control group B had the highest pass rate among

male students with 92% of the male students passing. This was 23% higher than the

treatment group, which had previously led in this category. For the most part, these data

confirm the other data presented in this paper. The male students in this study performed

better in the non-collaborative environment; the female students performed better in the

collaborative classrooms, and they performed best in the collaborative classroom in

which the parallel, station, and alternative teaching structures were being implemented.


These data also show that the treatment group students performed best overall--even if

they did not make more significant gains than the other two groups.

Focus question three, which was concerned with students and teachers perception

of and ability to adapt to this type of collaborative instruction, generated both quantitative

and qualitative data. The quantitative data were derived from students’ responses to five

Likert Scale survey items. After receiving the treatment or control activities, the survey

was administered to both of the collaborative classes that were involved in the study. The

weakness of this approach was the difficulty in determining if the treatment caused the

differences in the treatment group students’ perceptions, or if the treatment group had a

different outlook upon collaborative teaching before the study. To increase the reliability

and validity of the qualitative and quantitative data produced by the surveys, I

administered the treatment immediately following a school holiday and implemented the

structures for over four weeks. In doing so, I hoped to increase the likelihood that

students would mostly reflect upon the treatment when responding. A Chi Square was

used to analyze the quantitative data. Cronbach’s Alpha for both groups was high, so the

this data was reliable. Both classes agreed at the .01 level that having two teachers has

helped them better understand math. Both classes also expressed that they believed their

class was better because it had two teachers. These two responses suggest that the

students in both the treatment and control groups have an overall positive outlook upon

collaborative teaching. The treatment group, however, showed trends in two other areas

that were key to this study. They agreed at the .001 level that they enjoyed the activities

that they were able to do in collaborative classes more than activities in non-collaborative

classes. This is very significant because, prior to the study, both the treatment and control


groups had been taught using identical methods. Therefore, it is very unlikely that a

variable other than the treatment could have caused such a strong response in this group

while remaining insignificant in the control group. Students clearly enjoyed being in a

classroom in which parallel, station, and alternative teaching were being used as

instructional tools. The last statement that was significant only for the treatment group

was that they felt like their behavior was better because they had two teachers. The

majority of students in this group either strongly agreed or remained neutral. Because the

control group did not show a trend in their response to this item, the treatment may have

led to improved student behavior.

Qualitative data concerning student outlook was gathered from five free response

survey questions. The data generated from students’ responses were coded for themes.

Both the treatment group and control group expressed appreciation for the help given by

both teachers. Many members of the treatment group, however, stated that they enjoyed

being able to have two perspectives in their classrooms. While the control group students

generally seemed to regard the additional teacher as a source of individual help similar to

that provided by a teacher’s assistant, the treatment group clearly appreciated the teacher

as a teacher. Several of the treatment group’s students even commented upon how much

they enjoyed splitting up and working in groups during instruction. Another important

theme that emerged from the free response survey questions concerned students

understanding of the purpose of collaborative teaching. The majority of students in the

control group seemed to hold the belief that the teacher was in the classroom because

certain students needed extra help. The students in the treatment group, however, seemed

to believe that both teachers were in the room to help all students. One possible result of


the treatment, therefore, was to remove the stigma involved with students working with a

collaborative teacher.

Qualitative data referring to focus question three were also gathered from an

interview with the special education teacher and the reflective journal. Overall, the

special education teacher did not reflect kindly upon the treatment. She felt that her

purpose in the classrooms was to help manage behavior and work with students

individually as needed. As such, she said that she tended to favor the one teach, one assist

approach used in the control group. She indicated that she likes classrooms that are more

structured and teacher directed. The teacher also stated that she doesn’t have much

tolerance for the noise level when both teachers were actively teaching within the same

classroom. While my personal reflections included issues with the level of noise, the

themes found in the reflective journal differed significantly from the special education

teacher’s reflections. I felt like the class was considerably more engaged throughout the

course of most lessons. I also recorded feeling the benefit of having another teacher’s

voice and perspective in the classroom. Not only did I perceive that the structures being

tested benefited the students, I felt like they helped me as well. Being able to equally

share the hard work of managing and teaching a classroom full of students left me feeling

much more effective as a teacher. I also noted that I did not feel as tired after working

with the treatment group as I had in the past.

Most of the research agreed that frequent and effective planning is a necessary

part of any effective collaborative teaching program (Seleo & Garderen, 2010; Murawski

& Dieker, 2004; and Hammeken, 2010). Gurgur and Uzuner (2010) claimed that a lack of

common planning often leads to poor teaching practices. Likewise, both the special


education teacher and I cited inadequate planning time as a major hinderance to the

efficacy of our collaboration. According to the research, some of the problems involving

classroom noise, space, and transition times could have been eliminated had we had more

time to plan together.

During the instructional plan interview, both the peer teacher and assistant

principal confirmed most of Dingle et al. (2004) and Patterson et al.’s (2009) work,

which explained the importance of using collaborative teaching structures to differentiate

instruction and meet the instructional needs of all students. Their opinions differed from

the literature by emphasizing the implementation of a healthy mixture of student centered

and teacher centered instruction while the literature urged teachers to remain mostly

student centered.

Patterson et al. (2009) asserted that students benefit more from collaborative

teaching when teachers use a variety of co-teaching models and strategies. As a whole,

the students in this study failed to confirm Patterson et al’s assertion because there was

no significant difference among the groups at the study’s conclusion. This is in sharp

contrast to several studies cited by Thousand et al. (2006), in which students showed

large academic gains when presented with a variety of collaborative structures. Two

major differences between this study and those cited by Thousand et al. are the scope and

duration of each study. Most of the described studies involved several classrooms or

several schools and lasted much longer than this study.

The female subgroup, however, did seem to benefit from the treatment. A similar

phenomenon was not found in the literature that was presented in chapter two. Deak

(1998), however, described ideal conditions for female students. Certain elements, such


as smaller groups and opportunities for more peer interaction, of the treatment bore

remarkable similarity to the characteristics that Deak described. The girls may not have

performed as well in the two control groups because their instruction was more

traditional and teacher directed, which was the antithesis of Deak’s list of effective

practices. Geist and King (2008) also described the differences between boys and girls.

Boys, they say, generally like to work alone. Even in a group, they often work by

themselves and report to a group leader. They generally like to compete and want to look

better than the rest of the group. Girls, on the other hand, like to see themselves as part of

a group. They like to communicate as they are learning and want the whole group to

succeed. Geist and King’s findings explain, at least in part, the results of this study. The

boys clearly achieved more in the more traditional individualistic classroom

environments, but the girls seemed to learn more in the social, group based environments.

A large portion of the research referenced the importance of using collaborative

teaching to remove the stigma involved with special education and invite students with

disabilities into the larger group. Cook and Friend (1995) suggested that students with

disabilities prefer to be placed into mainstream classrooms, but also cautioned that

special education teachers should be careful not to stigmatize these students by working

solely with them. The data generated from the student surveys suggested that the

collaborative structures used in the treatment group worked to eliminate this stigma.

When asked about the purpose of the other teacher, nearly every student in this group

wrote that the other teacher was in the room to help all students. When asked the same

question, however, students within the collaborative control group acknowledged that

they believed certain students were not as smart and needed extra help. A strong stigma


clearly existed in the collaborative class that relied solely upon the one teach, one assist

model.

Discussion

Upon completion of this four week study of the effects of using parallel, station,

and alternative teaching structures in a sixth grade classroom, the data seems to indicate

the following: effective implementation of collaborative teaching should begin with

frequent collaborative planning to work out details regarding teaching strategies,

transition times, and space; instruction in a collaborative classroom should begin with

students’ needs as determined by assessment data and contain real world application and

higher order thinking; parallel, station, and alternative teaching do not lead to

significantly higher overall student growth when compared to other collaborative

structures and non-collaborative classrooms but may lead to an improvement in girls’

learning outcomes; students enjoy classes more and have a greater appreciation for both

teachers when given the treatment; and teachers may have differing opinions regarding

parallel, station, and alternative teaching.

Most of the data that were derived regarding the implementation of the structures

led to conclusions similar to those found in the literature. These data clearly indicate that

planning is of upmost importance throughout the implementation process. Teachers who

wish to establish a strong collaborative program of any type should invest a significant

amount of time in working out the details before beginning instruction and an even larger

amount of time planning instruction each week. Teachers cannot hope to establish an

effective collaborative program without this investment of time. Throughout the planning

process, teachers should consistently use assessment data to make grouping decisions and


other instructional decisions. Basing decisions upon students’ needs does not mean that

teachers of students with disabilities should dumb down instruction. All students,

regardless of their ability level, should be able to encounter real-world application as well

as higher order thinking. These principles are relevant to all collaborative programs, but

will prove to be especially important components of collaborative programs in which

teachers wish to use a variety of collaborative teaching structures.

The student outcome data did not show the improvements that I had anticipated.

In retrospect, I have realized that four weeks of using the collaborative structures was not

enough time to truly test the effectiveness of the strategies. Developing and teaching the

structures took too much time on the front end. Had we had several months or even an

entire year to work with the structures, I believe that the student outcomes may have been

very different. I feel like the girls responded well to the treatment because of their

preference for working cooperatively as part of a team. These data are not strong enough

to make broader conclusions without further testing. The assertion could be made,

however, that the treatment did nothing to harm or inhibit student learning outcomes. The

data also suggests that female students benefit more than male students from using a

variety of collaborative structures.

Neither the students’ nor the teachers’ feelings about using the collaborative

structures surprised me. The students clearly enjoyed working and learning in smaller

groups and with both teachers. Another possible benefit of using a combination of

collaborative teaching structures is that they seemed to have removed the stigma

associated involved with working with a special education teacher. When both teachers

are seen as equals in the classroom, they can both be considered viable instructional


options for all students. The special education teacher probably responded negatively

toward the treatment because she holds a very traditional teaching philosophy in which

one teacher lectures most of the class while the students listen and take notes. The

treatment did not change this, but I never expected it to change her outlook. Additional

planning time before instruction each day probably would have made the treatment more

pleasant for the other teacher involved. My teaching philosophy is more progressive, so I

went into the study believing that it would work for me. Still, I would suggest that

collaborative teachers carefully work out the details before attempting to implement

parallel, station, and alternative teaching in order to make the process more enjoyable for

all parties involved. They should make sure that they can agree on how they are going to

use the classroom space, their tolerance for noise, and how the strategies are going to be

used.

To achieve structural corroboration, I derived data from multiple sources

including several interviews, observational and reflective data, surveys, and student test

scores. With the exception of focus question two, every focus question was answered

with data that came from at least three different sources. To increase the amount of data

available to address focus question two, I drew test scores from a second non-

collaborative control group.

To maintain fairness, I presented dissenting points of view in chapter two and

when reporting the data. Within the literature, however, opposing points of view were

often limited or of little significance. Still, I tried to make a tight case and present all

points of view. After acknowledging the faults and building upon the strengths of this

research design, I feel like I have established rightness of fit.


Implications

The quantitative data produced by this study suggest that using a mixture of

collaborative teaching structures including the parallel, station, and alternative models

does not necessarily lead to improved student testing outcomes. It also suggests,

however, that testing outcomes may be improved for female students. I do not feel like

these data are conclusive enough to generalize the findings to larger populations. Other

quantitative data, such as the student survey data, build a much stronger case. The

students clearly enjoyed learning more and felt like they were better behaved in the

treatment group. This finding most certainly seems reliable enough to warrant the

assertion of broader claims. Most students will enjoy working more in a classroom whose

teachers employ a wider variety of collaborative teaching structures.

Parts of the qualitative data confirm the quantitative survey data. Students

genuinely appreciate collaborative classrooms that equally employ the skills of both

teachers. When a classroom is structured in this way, the students have a better outlook,

not only upon collaborative teaching, but on learning in general. Teacher perception

seems to be split, so it is hard to make this a part of any argument for or against using a

variety of collaborative teaching structures. This split most likely mirrors reality. In most

schools, about half of the teachers may enjoy and appreciate the collaborative structures

and about half will not respond well to them. Teachers would do well to study the

research on collaborative teaching and find structures that will work for them in their

classrooms. The purpose of collaborative teaching is, after all, to benefit the students.

Regardless of teachers’ opinions of how to structure their classroom for collaborative

teaching, the data from this study clearly indicate that collaborative planning and grade


level appropriate, differentiated instruction is a necessary component of any special

education program.

While conducted in a middle school math classroom, this study might be regarded

as important by all special education and grade level teachers who work with students of

disabilities regardless of content and grade level. This study could be replicated to

determine the treatment’s effectiveness in other subjects or grades by following the

procedures outlined in chapter three.

This study transformed the participants in the treatment group by changing their

attitudes about collaborative instruction. After working in groups with both classroom

teachers, they developed a new appreciate for the special education teacher. They will

also be more accustomed to learning in this type of classroom should their future

collaborative teachers choose to use any of the parallel, station, and alternative teaching

models. The special education teacher who participated in the study was able to work in a

classroom in which she was able to teach and work with students more often than she

may have done in the past. Although she did not favor this type of instruction, these new

interactions may have changed the way that she views collaborative teaching. It is hoped

that she will remember the parts of the treatment that she preferred, like alternative

teaching, and use them in future classrooms.

This study transformed me in a number of ways. I feel like I have come away

with a better understanding of how to make collaborative teaching work well in my

classroom. If I were to repeat the study today, I feel like I would have better results. It

also gave me more confidence to try new strategies in my classroom. Because this was

my first attempt at action research, I had never before systematically researched and


tested an idea. Before this, every attempt at trying something new felt like a journey into

unknown territory. I now feel like I have the tools to try new ideas and evaluate the

results, which will make me a much better teacher.

Impact on Student Learning

The quantitative data showed little impact on overall student learning. Still, the

female students did demonstrate a little more growth in the treatment group. What may

be more important to reiterate here is that the treatment also did not seem to have a

negative impact of student learning. Because the qualitative data all indicate that the

students enjoyed learning more in the treatment group, the fact that the strategies did not

negatively effect learning outcomes is noteworthy. If teaching students to appreciate

learning is one of the school’s primary goals, then strategies and structures like the ones

tested in this study are worth implementing, so long as they do not impede students’

learning.

Recommendations for Future Research

Future researchers should create a research design that involves multiple

classrooms for a minimum of an entire school year. These classrooms could assess

student progress using common benchmark exams rather than trying to make

determination based upon a single unit of study. The previous year’s benchmark scores

could provide a control group should researchers desire to implement the structures in

every class. In doing the study, researchers should frequently monitor the girls’

performance in order to determine if the treatment’s perceived benefit to girls in this

study was genuine or simply an anomaly. The boys involved in future studies should also

be carefully monitored to ensure that their academic growth is not being impeded by the


treatment. Future researchers in areas with higher minority populations may also want to

study the effect of the collaborative structures upon different minority groups.


References

Cook, L., & Friend, M. (1995). Co-teaching: Guidlines for creating effective

practices. Focus on Exceptional Children, 28(3), 1-16. Retrieved

for Eric at EBSCOhost.

Deak, J. (1998). How girls thrive: An essential guide for educators

(and parents). Washington, D.C.: National Association of Independent

Schools.

Denzin, N., & Lincoln, Y. (1998). The fifth moment. In N. Denzin & Y.

Lincoln (Eds.), The landscape of qualitative research:

Theories and issues (pp. 407-430). Thousand Oaks, CA: Sage

Publications.

Dingle, M., Falvey, M.A., Givner, C.C., & Haager, D. (2004). Essential special

and general education teacher competencies for preparing teachers for

inclusive settings. Issues in Teacher Education, 13(1), 35-50.

Retrieved from Eric at EBSCOhost.

Eisner, E. (1991). The enlightened eye. New York: MacMillan.

Falvey, M.A. (2004). Toward realizing the influence of “Toward realization of the

least restrictive educational environments for severely handicapped

students.” Research and Practice for Persons with Severe

Disabilities, 29(1), 9-10. Retrieved from

https://tash.org/publications/rpsd/rpsd.html


Falvey, M. A., & Givner, C. C. (2005). What is an inclusive school? In R.A. Villa

& J.S. Thousand (Eds.), Creating an inclusive school (2nd ed.),

(pp. 1 - 11). Alexandria, VA: ASCD

Friend, M. (2008). Co-Teach! A manual for creating and sustaining effective

classroom partnerships in inclusive schools. In M. Friend & D. Bursuck

(Eds.), Including students with special needs: A practical

guide for classroom teachers (pp.71-105). Upper Saddle River, NJ:

Pearson.

Geist, E. A. and King, M. (2008). Different, not better: Gender differences in

mathematics learning and achievement. Journal of Instructional

Psychology, 35(1), 43-52. Retrieved from Eric at EBSCOhost.

Gilles, C., Wilson, J, & Elias, M. (2010). Sustaining teachers’ growth and renewal

through action research, induction programs, and collaboration. Teacher

Education Quarterly, 37(1), 91-108. Retrieved from Eric at

EBSCOhost.

Gurgur, H., & Uzuner, Y. (2010). A phenomenological analysis of the views on

co-teaching applications in the inclusion classroom. Education

Sciences: Theory and Practice, 10(1), 311-331. Retrieved from

Eric at EBSCOhost.

Hammeken, P.A. (2000). Inclusion: 450 strategies for success.

Thousand Oaks, CA: Corwin Press.


Hendricks, C. (2009). Improving schools through action research: A

comprehensive guide for educators. Upper Saddle River, NJ:

Pearson Education, Inc.

Kinchloe, J., & McLaren, P. (1998) Rethinking critical theory and

qualitative research. In N. Denzin & Y. Lincoln (Eds.), The

landscape of qualitative research: Theories and issues

(pp. 260 299). Thousand ␣ Oaks, CA: Sage Publications.

Kosheleva, O., Kreinovich V., Longpre, L., Tchoshanov, M., & Xiang, G. (2007).

Towards interval techniques for processing educational

data (Technical Report Paper 256). Retrieved from

http://digitalcommons.utep.edu/cs_techrep/256.

Lagrange College Education Department (2008). The conceptual

framework. Lagrange, GA: Lagrange College.

Little, M.E., & Dieker, L. (2009). Co-teaching: Two are better than one.

Principal

Leadership. 9(8), 42-46. Retrieved from

http://www.principals.org/KnowledgeCenter/Publications.aspx.

Mahony, M. (1997). Small victories in an inclusive classroom. Educational

Leadership, 54(7), 59-62

Meijer, R.R. & Oosterioo, S. J. (2008). Measurement: Interdisciplinary research

and perspectives. Psychology Press, 6(3), 198-204. Retrieved from

Eric at EBSCOHost.

http://www.principals.org/KnowledgeCenter/Publications.aspx


Murawski. W. W., & Dieker. L. A . (2004). Tips and strategies for co-teaching at

the secondary level. TEACHING Exceptional Children. 36(5). 52-58

Patterson, J.L., Connolly, M.C., & Ritter, S.A. (2009). Restructuring the inclusion

classroom to facilitate differentiated instruction. Middle School

Journal, 41(1), 46-52. Retrieved from

http://www.nmsa.org/Publications/MiddleSchoolJournal/Articles/Septemb

er2009/tabid/2011/Default.aspx

Popham, W. J. (2008). Classroom assessment: What teachers need to

know. Boston, MA: Pearson Education.

Powell, K.C., & Kalina, C.J. (2009). Cognitive and social constructivism:

Developing tools for an effective classroom. Education, 130(2), 241-

250. Retrieved from Eric and EBSCOhost.

Sileo, J.M., & Garderen, D. (2010). Creating optimal opportunities to learn

mathematics: Blending co-teaching structures with research-based

practices. Teaching Exceptional Children, 42(3), 14-21. Retrieved

from Eric at EBSCOhost.

Sturges, J. (1978). How to make the most out of course evaluations

forms. Paper presented at the Educational Innovations Exchange,

Council on Social Work Education Annual Program Meeting, New

Orleans, LA. Abstract retrieved from Eric at EBSCOhost.

Thousand, J.S., Villa, R.A., & Nevin, A.I. (2006). The many faces of

collaborative planning and teaching. Theory Into Practice, 45(3), 239-

248. Retrieved from http://dx.doi.org/10.1207/s15430421tip4503_6

http://www.nmsa.org/Publications/MiddleSchoolJournal/Articles/September2009/tabid/2011/Default.aspx

http://www.nmsa.org/Publications/MiddleSchoolJournal/Articles/September2009/tabid/2011/Default.aspx


Treahy, D.L, & Gurganus, S.P. (2010). Models for special needs students.

Teaching Children Mathematics 16(8), 484-490. Retrieved from

http://www.nctm.org/eresources/article_summary.asp?URI=TCM2010-

04-484a&from=B

Villa, R.A., & Thousand, J.S. (2005). The rationales for creating and maintaining

inclusive schools. In R.A. Villa & J.S. Thousand (Eds.) Creating an

inclusive school (2nd ed.), (pp. 41-56). Alexandria, VA: ASCD.

http://www.nctm.org/eresources/article_summary.asp?

effective collaborative teaching strategieshome.lagrange.edu/educate/advanced programs/m.ed....

Documents