TEACHER STUDENT RELATIONSHIP QUALITY AND
STUDENT WRITING PERFORMANCE
B. A., University of Colorado, 1991
M.A.. University of Colorado, 1994
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Educational Leadership and Innovation
2001 by Juree Griffin
All rights reserved.
This thesis for the Doctor of Philosophy
has been approved
Griffin, Juree (Ph.D., Educational Leadership and Innovation)
Teacher Student Relationship Quality and Student Writing Performance
Thesis directed by Professor Alan Davis
A two-level correlational design was used to determine the association between the
quality of teacher student relationships and writing performance while controlling
statistically for differences in student-level and classroom-level variables. Teacher
ratings of close relationship quality were significantly correlated with student
performance in writing. Close teacher student relationship quality also explained a
statistically significant amount of variance in student performance in writing.
This abstract accurately represents the content of the candidate's thesis. I recommend
Purpose of the Study..................................I
Limitations of Early Research.........................3
Guiding Framework.................................... 4
Connection of Framework to Current Study..............7
Research Questions................................... 11
Structure of Dissertation........................... 14
2. REVIEW OF THE LITERATURE............................ 15
Risk and Resilience...............................21
Teacher Student Relationships and Student Outcomes...22
3. METHODOLGY.......................................... 48
Subjects and Sampling Procedures.................... 50
Writing Prompts....................................... 54
Writing Sample Administration Guidelines.............. 54
CSAP Scoring Procedures............................... 55
Inter-Rater Agreement................................. 56
Student Teacher Relationship Scale.................... 57
Measures of Other Variables........................... 60
Data Collection Procedures............................... 61
Data Analysis Procedures..................................62
Approach to Analysis..................................... 66
Teacher Student Relationship Quality..................... 67
Teacher and Student................................... 72
Relationship Quality and Writing Performance............. 74
Relationship Quality, Writing Performance,
and Other Variables.......................
5. DISCUSSION................................................. 82
Summary and Explanation of Results......................... 82
Correlation Analysis.................................... 84
Regression Analysis..................................... 86
Integration and Implications of Finding with Previous
Convergent Findings..................................... 88
Divergent Findings...................................... 89
Contribution of Findings to Literature.................. 90
Implications of Findings................................... 91
Applied Implications.................................... 92
Design and Internal Validity............................ 93
External Validity and Generalizability.................. 93
Analysis and Statistical Power.......................... 94
Lessons Learned and Future Directions...................... 95
A. My Teacher......................................... 97
B. Student Teacher Relationship Scale.................98
C. Teacher Survey.......................................100
D. Writing Prompts..................................... 101
E. Consent Letters..................................... 102
F. CSAP Scoring Rubric................................ 106
Figure I. Factors that Influence Writing............................... 9
Figure 2. Research Model............................................... 49
Figure 3. My Teacher Distribution...................................... 68
Figure 4. Distribution of Conflicted Relationship Quality Ratings...... 70
Figure 5. Distribution of Close Relationship Quality Ratings........... 70
Figure 6. Distribution of Dependent Relationship Quality Ratings....... 71
Table l. Levels of Ecological Influence on Development................ 5
Table 2. Teacher Demographics......................................... 52
Table 3. Inter-Rater Scoring Consistency.............................. 56
Table 4. Writing Activities............................................61
Table 5. Mean Scores for Relationship Quality..........................68
Table 6. Teacher Student Relationship Quality Ratings Raw........... 72
Table 7. Teacher Student Relationship Quality Ratings Centered...... 73
Table 8. Correlation Analysis of Relationship Quality and Writing.....75
Table 9. Regression Analysis of Student Writing Raw Scores.......... 76
Table 10. Regression Analysis of Student Writing Centered Scores.... 77
Table 11. Writing Instruction, Time, Relationship Quality.............79
Table 12. Writing Sample Components and Relationship Quality Raw... 80
Table 13. Writing Sample Components and Relationship
Purpose of the Study
The purpose of this study is to explore the link between the quality of teacher
student relationships and student performance in writing. Relationship experiences
shape children's learning and can disproportionately contribute to their academic
success and later adjustment as an adult (Pederson et al., 1978). Children grow and
learn from a variety of people, circumstances, settings, and the interactions between
them. Teachers, however, have special influence to help children adjust and excel
through their unique encouragement, relational style, and instruction. The goal of this
dissertation is to contribute unique evidence documenting the link between the quality
of the relationships that students have with their teachers and student academic
The teacher student relationship is defined as the ongoing, stable patterns of
interrelationships between students and their teachers. At a theoretical level, it
reflects behavioral patterns, individual style, relationship histories of both student and
teacher, and ongoing forces and influences on their interactions (Pianta, 1999; Pianta
& Walsh, 1996).
Early research on interactions between teachers and students focused
primarily on teacher attitudes, expectation, behaviors and student traits (Brophy &
Good, 1974). While this information has had positive impact on teaching practices,
there continues to be a gap in practitioners' understanding of the power of teacher
student relationships to affect or transform a student's school experience and to
promote healthy academic and social development (Pianta & Walsh, 1996).
More recent work has focused specifically on teacher student relationships
and showed that teacher beliefs (and subsequent behaviors) about the quality of their
relationship with students can significantly influence children's achievement
orientation and childrens behavior (Alexander & Entwistle, 1988; Pianta, 1992).
However, many questions about the impact of teacher student relationships on
achievement remain unanswered. One such question could be asked about the
agreement between teacher and student perceptions of relationship quality, and
An efficient and reliable measure of the quality of teacher student
relationships has only recently been developed (Pianta, 1992). Early studies of
student teacher relationships used time intensive observation methodologies to study
attachment relationships between children and their primary caregivers (Pianta,
1997). The development of a self-report pencil and paper measure has allowed
research to be completed much more quickly, accelerating the growth of professional
knowledge about teacher student relationship quality.
Limitations of Early Research
Considerable work needs to be done to further professional understanding of
relationships as contexts for student development (Pianta & Walsh, 1996; Pianta,
1999). Few studies have examined the connections between relationship quality and
achievement (Birch & Ladd, 1997). Research must continue to statistically
substantiate the link between the quality of teacher student relationships and student
performance before the potential of this topic as an intervention context and
commitment to staff development around this topic will be recognized as a legitimate
issue of concern in the educational arena. While this connection may be recognized
at an intuitive level, it continues to be under-represented in the literature base.
Further study of the link between student teacher relationships and student
achievement is important for several reasons. First, developing a deeper
understanding of the patterns of teacher student relationships, the correlation of
teacher and student beliefs about their relationships, and related student outcomes will
contribute important information to the literature base currently available on this
topic. Second, studying the quality of the teacher student relationship and its link to
student outcomes can expand approaches to intervention with students at risk for
academic failure. Instead of viewing failure as child-centered, evidence from this
study may expand options for addressing academic needs to include the teacher
student relationship (Pianta & Walsh, 1996). Third, results from research can
inform university preparation programs about the importance of teaching new
educators about the link between relationship quality and school outcomes. Finally,
further study of student-teacher relationship quality and its link to student
performance can expand school-based interventions for all students.
Chapter 2 describes several guiding frameworks through which to understand
teacher student relationship quality. These include the perspectives of attachment
theory, relationship framework, risk and resilience, and the ecological model of
human development. One framework, the Ecological Model of Human Development
by Bronfenbrenner (1979), is applied to the current study.
Bronfenbrenner's Ecological Model of Human Development (1979) provided
the guiding framework for the current study because it reflects the dynamics of
interactions that define student teacher relationships. It is inaccurate to view the
teacher student relationship as one where the teacher teaches and the student receives
instruction. Instead, the teacher student relationship must be examined using a
framework that reflects its complexity. This section concisely reviews the principle
components of Bronfenbrenner's model.
Bronfenbrenner argues that the developmental ecology of childhood is an
interdependent and nested system (Fraser, 1997). This system is comprised of
interactive environments (Bronfenbrenner, 1978; Richman& Bowen, 1997).
Influences on development occur at four levels in the Ecological model. These are
shown in Table 1.
Table 1. Levels of Ecological Influence on Development
Levels of Ecological Influence on Development
Ecological System Principle Influence on Development
Microsystem A pattern of activities, roles, and interpersonal relations experienced by the developing person in a given setting with particular physical and material characteristics.
Mesosystem The interrelations among two or more settings in which the developing person actively participates (e.g. a child's relationship among home, school, neighborhood peer group).
Exosystem One or more settings that do not involve the developing person as an active participant, but in which events occur that affect, or are affected by, what happens in the setting containing the developing person.
Macrosystem Broad social structures and institutions that influence and/or support belief systems that are part of a particular culture or society.
Bronfenbrenner's model proposes that cognitive and social development is
based on the interaction of environmental interconnections and their impact on
experiences that directly affect development. This relationship between adult-child
interactions and the interdependence of the environment and these interactions is
what defines Bronfenbrenners Ecological Model of Human Development. In order
to support development, microsystems (schools, classrooms, school activities) require
joint participation, communication, and knowledge about the characters in the setting
(teachers and students). The "health" or supportiveness of these three elements is key
to the effect of a relationship as a context for development.
Applied to the student-teacher relationship, the model supports the idea that
the relationship between student and teacher, student and classroom, and teacher and
classroom interact multi-directionally to account for the contribution of experiences
and perceptions of both on the relationship itself. While Bronfenbrenner's model
does not judge the effectiveness of development based on particular quality
descriptors, it does uphold that relationships and patterns of interactions with people
and environments have a profound impact on developmental health.
Positive development can be traced back to the health of the relationships and
to the supportiveness of the relational contexts in which a child is a part. These may
include repeated patterns of interactions with parents, relatives, teachers, siblings,
friends, and other caregivers. Bronfenbrenner (1979) characterizes optimal
development by the degree to which the developmental context accommodates a
child's needs, the setting, and the way in which this relationship is mediated by other
systems (i.e. family, social group, culture, economy). The teacher student
relationship and its quality is an example of a developmental context.
Relationship quality is comprised, in part, of the extent to which the
relationship provides the child with resources that enhance or support the child's
developmental progress (Wachs, 1992). Teachers mediate a child's experience of
learning materials and situations. The quality with which they interact with the child
also contributes substantially to the value of their relationship. This implies that the
patterns and style of interactions between children and their teachers makes or forms
Connection of Guiding Framework to Current Study
Bronfenbrenner (1979) argues that relational dyads have an affective
component. Related to the current study, Bronfenbrenner would suggest that the
feelings of people in ongoing dyads (teachers and students in this case) about the
relationship become distinct and stable, making the affective quality of the
relationship salient to the effectiveness of the relationship itself to support
development. If a teacher and student experience positive interactions with one
another, then the relationship quality can directly support a child's learning. When
interactions between teachers and children are stressful, the feelings between them
may be negative or ambivalent, contributing to uncomfortable, but stable patterns of
interactions. Patterns of interactions that are strained cannot, according to
Bronfenbrenner, optimally support student development. Relationship quality
between teachers and children have the capacity to disrupt or to enhance learning.
Numerous factors influence academic performance. This study focuses on
writing in particular. The link between teacher student relationship quality and
writing performance is one such influence. Other factors might include building level
practices, individual characteristics of children, and family attitudes and value of
education (Pianta & Walsh, 1996). Figure 1 offers a visual depiction of the types of
factors that impact writing performance.
Figure 1. Factors that Influence Writing Performance
The writing process operates at a relationship level. Writing for elementary school
students is in part based on repeated reciprocal interactions between the writer and the
teacher. Calkins (1994) asserts that children become invested in writing at an
emotional level, making it a personal and interpersonal experience. She believes that
students feel success as writers through repeated genuine interactions with teachers
that express love, relationship, trust, alliance, and encouragement. The quality of
these relationship patterns communicates feelings of safety or apprehension to
students and consequently impacts the effort they invest in writing. Calkins believes
that children cannot write well if "...they are afraid to put themselves on the page" (p.
A child's emotional experience of writing, from choosing a topic, to drafting,
to conferring with the teacher, and sharing it with the class, strongly contributes to
their feelings of mastery as a writer. Writing exposes thoughts and feelings. If the
relationship a child shares with their teacher does not encourage self-disclosure or it
does not support the emotional component of learning, then they may view writing as
risky and frightening. Teachers have a significant role in how young children
experience writing (Atwell & Newkirk, 1987). Viewed through the combination of
Calkin's perspective and Bronfenbrenner's model of development, the importance of
teacher student relationship quality to writing performance is unmistakable.
While educators may have intuitive knowledge of the impact of their
relationships with the children in their classrooms, this knowledge is underutilized in
planning support for academic and emotional development of students. Instead,
teachers typically use strategies to control behavior so that students are receptive to
instruction. Many teachers are not accustomed to or comfortable with strategies to
support their students emotional needs. Teachers must develop new beliefs and
behaviors around meeting the emotional and learning needs of students. To learn,
some children need considerable emotional support that is not provided by the
behavior management techniques used by many teachers. Children need to feel safe,
to feel comfortable with risk taking in their learning, and to feel accepted. Teacher
student relationships should ideally reflect features of development-enhancing
contexts (relationships based on positive, trusting, reciprocal interactions) in order for
students to feel connected to the learning environment and to be receptive to
instruction and relational supports in the classroom.
The studies described in the review of literature illustrate the connection
between teacher student relationship quality as a context for social and cognitive
development and important student outcomes such as adjustment to school and school
oriented behaviors. The literature review shows that empirical evidence regarding the
link between relationship quality between teachers and student and academic
performance is incomplete.
My interest in studying student-teacher relationships and their link with
student performance is driven by a belief that quality student-teacher relationships are
as important to educational outcomes as instructional delivery and actual learning
materials. Within my sphere of influence, I would like to encourage teachers,
principals, and administrators to apply knowledge that student teacher relationships
are integral to student success.
The association between the quality of teacher student relationship and the
performance of elementary students in writing is the primary research question
explored through this study. These specific hypotheses guide the current study:
Hypothesis I: Student and teacher ratings of teacher student relationship quality
will be significantly correlated.
Hypothesis II: Teacher ratings of teacher student relationship quality will be
correlated with student performance in writing. These correlations are predicted:
Close relationship quality ratings will be positively correlated with writin
Conflicted relationship quality ratings will be negatively correlated with
Dependent relationship quality ratings will be positively correlated with
Positive relationship quality from the student perspective will be
positively correlated with writing performance.
A two-level correlational design was used to determine the association
between the quality of teacher student relationships and writing performance while
controlling statistically for differences in various student-level and classroom-level
variables. The current study design is structured around one dependent variable and
five independent variables. One independent variable, the teachers' rating of student-
teacher relationship, is comprised of three sub-scales. The other independent
variables include student perception of teacher student relationship quality, early
writing performance, average time devoted to writing instruction, and type of
classroom activities that support writing development. The dependent variable in
this study is a measure of writing performance collected at the close of the study.
Data were collected using self-report questionnaires, collection of district level data,
and structured writing to pre-selected prompts based on the Colorado Student
Subjects for this study included 4th grade teachers and their students in a small
urban school district. The focus of data analysis occurred at both student and teacher
level information. The sample pool includes 4th grades teachers and their students
across six elementary schools in Adams County District #14 schools. Sampling was
based on voluntary participation by both teacher and students. The location of the
study was chosen for convenience as the researcher is employed there and has
established relationships with administrators and building principals. Building
principals and fourth grade teachers were recruited uniformly using a standard
description of the purpose and goals of the research. A complete description of the
procedures employed in the study are provided in Chapter 3.
Structure of Dissertation
The current study will be organized in five sections or chapters. These
Chapter 1 Introduction to the research problem and methods
Chapter 2 Review of relevant literature related to research problem
Chapter 3 Methodology. Instrumentation, Data Analysis
Chapter 4 Results
Chapter 5 Discussion
REVIEW OF LITERATURE
Social and cognitive development are functions of patterns of interactions
between an individual and the multiple settings and environments that influence them.
To understand the contribution of the school environment to student performance, and
more specifically the link between teacheT student relationships and academic
performance, a variety of concepts and past research must be reviewed. Using the
ecological model as a conceptual framework, this review of previous studies outlines
important concepts, findings, and measurement issues in the sequence they appear in
the literature, ending with a brief discussion of unanswered questions revealed
through past research.
To understand the connection between the teacher student relationship and
student outcomes, a reminder of the social foundations of the educational process is
important. Schooling can be described as a social process. Rogoff (1990) asserts that
the development of academic skills is a process dependent on social context. She
describes it this way:
Instruction occurs in an interactive context, at first often naturalistic, but
increasingly formalized and constrained. Although these interactions have
intentional goals of increasing the child's skills in a particular area, they
nonetheless occur in the context of the relationships in which those
interactions are embedded and they are influenced by the qualities of those
relationships...The same qualities of relationships-as-regulators that operate
with respect to emotional development operate with respect to instruction in
more academic skill-oriented areas of development (p. 20).
Since the educational process reflects countless formal and informal interactions
between teachers and students, the critical contribution of relationships and
relationship quality should be recognized. Studies have established that relationships
with teachers during the early grades have a disproportionately strong influence on
achievement and social competencies in later childhood (Alexander & Entwistle,
1988). The student teacher relationship can act as an effective intervention context
for children. It can provide a child with new relationship experiences of sufficient
strength, intensity, and consistency that enables them to use adult support more
competently to further their development (Pianta, 1999).
Three other bodies of literature or models for understanding relationship
systems are summarized briefly in the next section as they are frequently referenced
in the background research relevant to the current study. These include attachment,
relationship framework, and risk and resilience. These frameworks are not
considered as alternatives to Bronfenbrenner's model, but as congruent or
complementary with his model. Most of the studies reviewed reference some
combination of these three perspectives to frame their understanding of teacher
student relationships. A brief overview of the major contributions of these fields to
the current investigation is offered to further establish the tenet that relationships
between students and teachers are important to development. Ultimately, however,
they can be understood in the context of Bronfenbrenners model.
What is understood about the nature and impact of the student teacher
relationship is based on several frameworks and bodies of research. The framework
that logically structures the nature of the student teacher relationship is the Ecological
Model of Human Development. Bronfenbrenners (1979) model asserts that human
development is based on multi-system level experiences, interactions, and patterns of
Theoretical Frameworks for Studying Student-Teacher Relationships
Human development does not occur in isolation. Instead, Bronfenbrenner
(1979) argues that development is a function of the interaction between the individual
and various developmental contexts. Social and cognitive growth is strongly
influenced by family, school, teachers, peers, neighborhood, and community settings.
According to Bronfenbrenner, behavior and development is transactional and subject
to the dynamics of social exchange (Fraser, 1997). While Bronfenbrenner postulated
that the quality of adult-child relationships are important to learning and
development, he did not focus his work on examining the detailed nature of these
Early research by Bowlby (1980) and Ainsworth (1989) established a
foundation for the accepted concepts and definitions for understanding attachment
and its impact on development. Basic principles of attachment theory describe
social-emotional development and its subsequent impact on all other facets of
development. Attachment develops from regular interactions with a primary
caretaker. These interactions help children develop an internal working model of the
relationship between the self and the world. Ainsworth extended Bowlby's research
to investigate qualitative differences in attachment relationships. She documented
four types of attachment behavior that represent attachment relationships of differing
quality. These include secure attachments, anxious-avoidant attachment, anxious-
ambivalent attachment, and anxious-disorganized attachment. Both Bowlby and
Ainsworth claimed that the quality of the child's attachment relationship has a direct
and powerful impact on the child's experience of all subsequent relationships as well
as on their cognitive development. Children who have not experienced consistency
in their primary attachment relationship require a restoring, healing relationship with
another significant adult in order to progress towards healthy development.
Much of what is known about the impact of adult-child relationships on later
developmental outcomes is based on studies conducted with mothers and their young
children (Ainsworth, 1989; Bowlby, 1980). One very important implication of this
research is that children who do not form positive and consistent attachment
relationships with caregivers are at great risk for poor transition to school, poor peer
relationships, and later academic failure. Werners (1982) pioneer research in risk
and resilience extended the importance of positive relationships beyond early
childhood. She discovered that children could overcome a rocky beginning in life
and go on to become well adjusted adults when provided with a sustained, supportive
relationship with another adult in their lives. This early work contributed to a
recognition of the importance of teachers interactions with their students as well as
the power of positive relationships in shaping childrens outcomes.
A relationship quality framework has also been applied to the study of the
impact of teacher student relationships. Sroufe (1996) described relationships in
terms of their health or support of development. While a considerable amount of
attachment research has documented the importance of positive relationships between
parents and their children, less is known about what impact teacher-child
relationships can have on a childs academic performance and adjustment in school.
Positive relationships between a child at risk for academic failure and poor social
outcomes and a caring adult can alter the childs vulnerability and enhance learning
(Garmezy, 1985; Rutter, 1987). Repeated negative relationship experiences with
teachers can disproportionately reinforce subsequent negative school experiences,
impeding healthy development.
Negative school experiences have been associated with undesirable outcomes
such as academic failure, poor peer relationships, and behavior problems. Clearly,
early school experiences are important. Alexander and Entwisle (1988) conducted a
longitudinal study documenting the developmental progress of a group of students.
One very important pattern they uncovered was that the trajectory for childrens
success or failure in school is relatively stable by the time they reach third grade.
One implication of this result is that teachers and practitioners need to connect with
students before they pass this critical period, particularly those students at high-risk
for academic failure. While a childs school day is comprised of a variety of
interactions, routines, and tasks, his experience of school is largely shaped by
relational experiences. A simple, but extreme example illustrates this point. If a
child feels unsuccessful with peers, adults, and with academic tasks, his overall
experience of school will likely be negative, leading to unpleasant, but confirming
experiences with peers, teachers, and academics. Childrens relationships with their
teachers can be a strong source of feedback and experiences that shape their efficacy
beliefs. These beliefs, in turn, shape a childs attitude about learning as well as their
actual learning behaviors.
Learning of any kind is an emotion-laden experience. Attempting a new skill
often elicits excitement, apprehension, and anxiety all at the same time. Learning
about reading, math, and writing all excite emotional responses in young learners.
The relationship between children and their teacher mediates the emotional intensity
of learning. Recognizing the health of the relationship and its ability to support a
child's learning is one way to apply the relationship system model to student teacher
relationships. The relationship system perspective, though more theoretically based,
offers another salient viewpoint for understanding the contribution of relationship
quality to student performance (Sroufe, 1988).
Risk and Resilience Framework
A growing body of research has focused on the applications of risk and
resilience constructs in studies of adult-child relationships (Anthony & Cohler, 1987;
Bernard, 1991; Cowen et al., 1990; Doll & Lyon, 1998; Felsman & Vaillant, 1987;
Garmezy, 1985; Garmezy & Masten, 1994; Garmezy & Rutter, 1983; Haggerty et al..
1994; Luthar & Zigler, 1991; Masten, 1994; Masten & Coatsworth, 1998; Rutter,
1987; Seifer et al., 1992; Werner & Smith, 1992). Risk and resilience research is
characterized by the view that developmental outcomes emerge out of experiences
and traits at the individual and environmental level, and that the nature of these
outcomes differs depending upon the intensity, chronicity, and unique interaction
between risk and protective experiences. Research has approached studying the
nature of school based adult-child relationships from somewhat of an anthropological
perspective in that the goal of the inquiry was to document the features of both
effective and negative relationships. Its goal has been to define critical characteristics
of effective relationships and the contexts in which they are supported (Pederson et
The quality of the teacher student relationship can be affected by traits of
either the teacher or student or by environmental features of the classroom. These
characteristics interact as protective factors, risk factors, or some combination of
both. Individual factors may include the degree to which the teacher and child bond
or connect, the teachers sensitivity, the teachers style of emotional expression
(regardless of the childs style), and the teachers responsiveness to the childs
emotional needs (Pianta, Nimetz, & Bennett, 1997). Other environmental and
contextual factors that impact the supportiveness of a relationship include the number
of adults and the number of changes a child is exposed to during the school day. The
more adults a child interacts with, the more fragmented their relationships. Also, the
more transitions a child must negotiate, changing environments in the school day, the
more fragmented their interactions will be with various adults in the school setting.
The impact of these issues on academic progress need to be understood at a greater
depth so that the research can be used to actively modify teachers interactions with
Teacher Student Relationships and Student Outcomes
This section reviews and highlights important findings from key studies that
explore student teacher relationships from the conceptual viewpoint (noticing that
there are qualitative differences in student teacher relationships), methodological
viewpoint (exploring means of assessing relationship quality), and from a
developmental impact viewpoint (how student teacher relationship quality relates to
Early studies showed that student teacher relationships (STR) could be
categorized and mapped much like the relationships between children and their
mothers. Then, research began to focus on the quality of student teacher
relationships and their impact on school adjustment in young children.
Methodologically, these studies remained closely aligned with tools used to measure
attachment relationships. A shift in teacher student relationship studies led to the
development of new measures to assess relationship quality and to explore the link
between this relationship and its impact on student adjustment and performance.
Each study reviewed ends with a summation or key point that reflects the contribution
of the study to the broader understanding of the impact of student teacher
relationships on student outcomes.
The work of Robert Pianta has become an important source of knowledge
regarding how student teacher relationships are conceptualized and understood and
serves as the starting point for the research review. Pianta (1997) constructs a
compelling argument about the importance of relationships for development and the
specific influence of the student teacher relationship on student adjustment. The
teacher student relationship involves feelings of closeness, sharing information and
experiences, histories of behavior patterns and interactions, and expectations for
future responses based on these patterns (Pianta, 1992; Hamilton & Howes, 1992).
Based on these relational ingredients, Pianta believes that relationships act as contexts
for development (classrooms, relationships, emotional regulation). Several studies
have contributed evidence that shows that the quality of patterns of child-adult
interactions exercise important influence on the adaptation of the child within a given
setting (Howes, Matheson, & Hamilton, 1994; Howes, Hamilton, & Matheson, 1994).
Student teacher relationships serve as a regulatory function with respect to emotional
and academic skill development. These qualities, in part, determine the settings' level
of support for development. The review of research that follows summarizes early
studies of teacher student interactions, the shift in the literature to teacher student
relationship constructs, and measurement of teacher student relationship quality.
Brophy and Good (1974) were early contributors to current understanding
about student teacher interactions. In their exhaustive analysis of their own research
as well as their review of other important studies of student teacher interactions,
Brophy and Good strove to document student differences that affect teacher
expectations and attitudes and to determine how these affect student-teacher
interactions and student performance. Stressing the importance of examining
student-teacher interactions at the individual relationship level, the authors suggested
that aggregate analysis of teacher student interaction data may mask rather than reveal
important relationship factors that contribute to understanding specific instructional
Brophy and Good's analysis revealed and showed repeated evidence of
student factors that are associated with positive teacher expectations and attitudes
towards their students. Some of these included socioeconomic status, race
(Caucasian), gender (female), achievement orientation, and having an appealing
personality. Positive attributes in these areas tended to yield higher expectations
from teachers as well as more positive interaction patterns with teachers. They
confirmed that seating arrangement, struggles with handwriting, and non-native
English speaking status were more often associated with negative patterns of teacher
As part of their original goal, Brophy and Good examined teacher
expectations and attitudes, the stability of these, and their impact on student-teacher
interaction behaviors. Brophy and Good concluded that expectations can be self-
defeating. If a teacher believes a student to be a low-achiever, and if this is viewed as
a permanent characteristic of the child, then the teacher may be relatively
unsuccessful in teaching the student. Their conclusions have significant implications
for the meaning of student teacher relationship quality. The impact of this pattern of
perceptions on relationship quality can be very damaging. If teachers internalize a
belief that a child cannot learn, the belief will observably affect their behavior. It
may be as subtle as not calling on the student as often as others, not pushing the
student to produce quality work, or overlooking certain behaviors in an effort to avoid
engaging the student or battling with them. These patterns of relational interactions
become mutually reinforcing, almost cementing the pattern as permanent.
Brophy and Good also viewed attitude as a meaningful element in the
interactions between students and teachers. For example, when a teacher decides
that a student is a low achiever, the teacher may feel angry or frustrated. Such
attitudes towards students, once formed, can begin to act as self-fulfilling prophecies.
The teacher may begin rigidly and inflexibly to treat the student differently based on
her attitude towards the student. Though Brophy and Good's (1974) exhaustive
analysis of research on student-teacher interactions, expectations, and attitudes was
not focused on an intervention level, their work contributes important information to
understanding factors that impact teacher student relationships.
Pederson, Faucher, and Eaton (1978) uncovered important features of the
impact of positive teacher student relationships. Using a historical case study
approach, Pederson et al. (1978) investigated the impact of 1st grade teachers on later
adult outcomes of their students. Results from the first stage of data analysis included
the discovery that teacher bias as shown through seating arrangement, degree of
interaction with particular children, and quality of teaching for preferred students
showed differences in later student outcomes. The authors concluded that teacher
behavior, based on preferences, may lead to very different experiences of school and
school success for non-preferred students.
A cross-tabulation of 1st grade teacher and adult status of former students
showed that one particular teacher had a disproportionately higher number of students
ranking in high' adult status than her teacher peers (64% vs. 31%, 10%). A
regression analysis showed that Ms. A, the exceptional teacher, added 10% to the
explained variance in adult status after other significant contributors such as family
background were controlled.
The authors concluded that their data supported their hypothesis that a good
first grade teacher could provide children such a major head start that the effects, in
terms of both academic self-concept and achievement, would continue to be felt in
later life. They also concluded that teacher-student relationships may have powerful
influence on reshaping the course of school adjustment and subsequent adult
Reports from participants in the study reflected that the special 1st grade
teacher made them feel worthwhile, supported their independence, motivated them to
achieve, and provided them with support to interpret and cope with environmental
demands. This teacher's students differed from their same-age peers on dropout rates,
academic achievement, behavioral competence, and adjustment in the adult world.
The developmental link between experiencing a quality relationship with an adult and
feeling encouraged to explore the world in a competent manner is a defining feature
of the parent-child attachment relationship. Based on the work of Pederson et al., the
student teacher relationship can be defined this way as well.
In their important early work on cognitive development, student-teacher
interactions, and the transition to full time school attendance. Alexander & Entwistle
(1988) documented the relationship between teacher behaviors and student
performance. Viewed through a life-course framework, Alexander and Entwistle
suggest that experiences during certain periods in life have disproportionate effects on
subsequent development. The authors emphasized two critical life-course events in
their work: transition between 1st grade and the implications for children who repeat
first grade as an indication of success/less successful completion of the life event.
Social development strongly influences how children manage developmental
hurdles such as attending school. Alexander and Entwistle suggest that the academic
achievement of children in their first years of school is related to the social
developmental processes of school of which the student teacher relationship is a part.
They believe that the first years of school have a strong influence on a child's
experience of school during subsequent years. One example of this is the cumulative
record begun for each student at the point of their enrollment in school. Documents,
attendance histories, and narratives from teachers strongly influence the opinions
formed about the student by others, often permanently.
Alexander and Entwistle (1988) draw two important conclusions about early
school experiences. First, a child's successful negotiation of social and cognitive
demands during their first year of school largely determines the trajectory of their
future educational experiences -- positive or negative. Additionally, teacher
attributions, attitudes, interactions, and expectations are associated with children's
adjustment to school. These patterns are considered to be fairly flexible during this
early period, but become rapidly fixed during a child's second year of schooling.
Second, children who experience low success (e.g. retention) during their early years
of schooling are more likely to continue to encounter lower levels of success with
school in the years beyond. This downward spiral, the authors suggest, is sustained
by changes in parental and teacher beliefs and expectations about a child's
performance. Few school based experiences after 3rd grade have enough weight or
influence to alter these established achievement pathways for children. Furthermore,
efforts to alter the trajectory of student achievement and adjustment after this point
require significantly more time and intensity than for students who experience
positive school adjustment and patterns of achievement prior to 3rd grade.
The practical implication of this work is clear. Early school experiences, both
academic and relational, are extremely important to how a child experiences school
for the rest of their educational careers. The importance of this work to the
understanding of student-teacher relationship quality centers around a teacher's
capacity to mediate patterns of school experiences or to interrupt the negative
emotional experiences some children have with school and to change the course of
their school experiences through stable, consistent, supportive relationships.
Pianta and Steinberg (1992) documented fundamental conceptual and
methodological advances in the study of the student-teacher relationship. Pianta and
Steinberg used an early version of the Student Teacher Relationship Scale (STRS)
developed by Pianta (1992) to assess teacher feelings, beliefs, and perceptions about
student's behavior towards them, as well as tapping their own internal working
models of their relationships with individual students. Items were designed to
measure dimensions of warmth and security, anger and dependence, and anxiety and
insecurity. A detailed review of their study is necessary to establish rationale for
using Pianta's approach to relationship quality measurement.
The relationship model reflected in the STRS can guide teacher behavior with
students in meaningful ways (Bretherton, 1985). Pianta and Steinberg (1992)
suggest that the student teacher relationship can act as a protective factor against risk
(i.e. referral to special education or retention). Operating from this perspective, the
primary goal of their study was to document patterns of student-teacher relationships
and their relatedness to student behavior at school and later school adjustment. The
sample included 26 teachers and 436 students total. The authors administered the
STRS in the spring of the school year to allow for stable relationship patterns between
children and their teachers.
Preliminary results from the Student Teacher Relationship Scale revealed that
emotional and affective experiences of the student teacher relationships were
reflected in several descriptive categories. The authors conducted a factor analysis of
a 31-item version of the STRS. Initial analysis yielded a total scale alpha level of .90.
Four distinct descriptive groupings were clear. Coefficient alpha estimates of internal
consistency reliability were as follows: conflicted/angry, .93, warm/close, .83, open
communication, .64, and troubled/closed, .53. This factor analysis was replicated by
The data showed a clear relationship between student-teacher relationship
quality and students' classroom behavior. The most significant of these involved
student-teacher relationships rated as conflicted/angry. Conduct problems in the
classroom correlated with conflicted student-teacher relationship ratings. More
specifically, conflicted STRS ratings correlated with behavior problems (r = .51),
conduct problems (r = .67), learning problems (r = .47), and low levels of social
competence (r = -.51) as rated by the Teacher-Child Rating Scale (Hightower et al.,
The authors found that this pattern, once established, often carried over to new
contexts as reflected in student-teacher relationship ratings by 1st grade teachers.
Student-teacher relationships rated as conflicted correlated with behavior problems (r
= .53), and conduct problems (r = .61). Decisions to retain/refer were not based
solely on a child's ability or classroom behavior, but also on the teacher's beliefs
about their relationship with that child. Overall, the author's concluded that teachers'
internal working models of their relationships with students contribute to an
understanding of children's school adjustment. The data confirmed that student-
teacher relationships are built on repeated patterns of behaviors and interactions.
Establishing and maintaining supportive relationships with teachers is
important to student adjustment and contributes greatly to how a student experiences
school. Lynch and Cicchetti (1992) explored this hypothesis in detail with children
considered high-risk due to past documented abuse and neglect. They predicted that
maltreated children would use their teachers as secondary sources of positive
relationship experiences when their own primary attachment figures (family) did not
provide this. Teachers, the authors suggested, are significant non-parental adults to
which maltreated children have frequent, consistent exposure. Repeated supportive
experiences with teachers may influence negative relationship models, thereby
allowing children to become more engaged in school.
Lynch and Cicchetti framed their understanding of the student teacher
relationship through a concept they described as "relatedness." Using this
perspective allowed them to assess children's perceptions about the quality of their
relationships with significant others. Wellborn and Connell (1987) developed a self-
report measure that assesses children's perceptions of the quality of their relationships
with a variety of people. The instrument was designed to record student's
evaluations of the degree to which they believe their psychological needs are being
met by significant relationship figures in their lives. The sample included 115
maltreated (based on participating in services from the local DSS for abuse and
neglect) and 100 demographically matched, non-maltreated students ranging between
ages 7-13. All data were collected in the context of a summer program provided for
children with the goal of treatment, prevention, and research around the effects of
Lynch and Cicchetti (1992) assigned student scores to categories
characterizing the patterns of their relatedness to teachers. These categories included:
optimal, deprived, disengaged, confused, and average. Maltreated children reported
more proximity seeking with teachers than non-maltreated children. The results
showed an overall effect of maltreatment status on children's patterns of relatedness
with teachers. Maltreated children were less likely to report optimal levels of
relatedness to their teachers than their non-maltreated peers.
The authors concluded that maltreated children's lower levels of reported
relatedness to teachers could be based on their skewed model of relationships.
Because the study did not concurrently investigate the perspective of the classroom
teacher, the unique opportunity to leant from both views was lost. Empirical study
of relationship quality from teachers and students in tandem would reveal more about
how the two perceptions are related. Examining both teacher and student perception
of relationship quality could also document which perception of relationship quality,
the teacher's or students', relates more to student outcomes.
While this sample did not confirm that maltreated children feel high levels of
relatedness to their teachers, the results do suggest that teachers have a great
opportunity to help children change their model of how relationships work by
attending the relationships they have with children. These relationships can become
corrective emotional experiences for children who have a history of negative school
experiences as well as histories of negative relationships with caregivers (Masten,
Best, & Garmezy, 1990).
Pianta's (1994) goal was to develop an instrument that would produce
descriptive labels for relationship qualities between teachers and students and to
document the performance of students compared with these groupings. Pianta's work
is based on attachment and relationship models that assert that children form
relationships with teachers that have a unique influence on their development.
The Student Teacher Relationship Scale (STRS) (Pianta, 1992) and the
Teacher Child Rating Scale (TCRS) (Hightower et al., 1986) were distributed to 26
teachers to collected information about student-teacher relationships tor 436
kindergarten age children. Using a 31 item Likert response format, the STRS was
completed and collected in May of the school year targeted by the study. This timing
allowed students and teachers to establish stable patterns of relationships.
Factor analysis grouped scale items into several distinct clusters describing
quality indicators of the student teacher relationship including: conflict/anger,
warmth/closeness, open communication, dependency, and troubled. The coefficient
alpha for total scale was .90. Alphas for the cluster groupings were as follows:
Conflicted/Angry, .93; Warm/Closeness, .84; Open Communication, .83;
Dependency, .64; and Troubled Feelings, .52.
Further analysis showed that these five factors accounted for 63% of the
variance in teacher ratings on the STRS. The conflicted/angry factor accounted for
30.4% of variance, warrath/closeness accounted for 14.1% of variance, open
communication 7% of variance, dependency 3% of variance, and troubled feelings
- 1% of variance. Pianta learned that 5-50% of teacher ratings characterized their
relationships with students as negative.
To explore this further, analysis was conducted to examine cluster group
membership across all classrooms. The five main cluster groups were regrouped into
two groups describing generally positive relationship patterns or generally negative
relationship patterns. A total of 74.5% of the sample were classified as generally
positive and 25.5% were described as negative. STRS scores for eight classrooms
indicated that at least 30% of the student teacher relationships were rated as negative.
These 8 teachers (31% of teacher sample) explained 51% of the student teacher
relationships described as difficult or negative. Results from Pianta's study showed
that student teacher relationships rated as conflicted/angry reflected negative affect
between teachers and students. Student teacher relationships described as warm/close
reflected positive affect between student and their teachers. The dependent cluster
described over-involved or over-reliant relationships with teachers.
Pianta concluded that the STRS could contribute valuable information about
the health of the relationship between a student and their teacher as well as guiding
plans for intervention to support the student's needs. He suggests that intervention at
the student teacher relationship level is often overlooked as a potential avenue for
intervention. Past research has documented, however, that the power of the
relationship is often the key to school success for high-risk students (Werner &
Expanding upon the early success of the Student Teacher Relationship Scale
(Pianta, 1992), Pianta, Steinberg, and Rollins (1995) went on to explore the tool's
application to studying student adjustment across multiple school years.
Specifically, the study explored the relationship between student teacher relationships
and student adjustment over a two-year period.
The authors proposed that the quality of the student teacher relationship is
related to later change in adjustment by comparing 1st grade and Kindergarten teacher
reports with false positive predictions of retention/referral during those grades.
Using the Student Teacher Relationship Scale and the Teacher Child Rating Scale
(Hightower et al., 1986), 70 Kindergarten-2nd grade teachers were surveyed about
their relationships with 413 students.
Overall results showed that children who experienced positive student teacher
relationship quality during Kindergarten were rated as better adjusted by their 1st
grade teachers (based on TCRS Teacher Child Rating Scale (Hightower et al.. 1986)
than peers who did not experience positive student teacher relationships during
Kindergarten. Further, children who experienced warm, close, and communicative
student teacher relationships with Kindergarten teachers were rated as better adjusted
and had more positive student teacher relationships in 2nd grade versus their peers
rated as having angry or dependent relationships with teachers during Kindergarten.
Pianta et al. also explored the link between student teacher relationship quality
and teacher decisions about referral for special education evaluation and retention in
Kindergarten. They found that students categorized with false positive predictions for
retention and referral (predicted referrals/retention that did not occur) experienced
more positive student teacher relationships than their peers who were true positives
for referral for evaluation and retention in Kindergarten. While these findings are
compelling, one possible limitation is that the findings are based on correlation and
causal interpretation is not possible. Another issue that weakens the potential power
of these results is the absence of the student perspective of the student teacher
relationship. Lynch & Cicchetti (1992) suggest that if the student's perspective were
explored different dimensions of student teacher relationship properties might be
A common thread throughout the studies reviewed to this point is that the
quality of the student teacher relationship makes a difference in a child's ability to
access the teacher as a resource and impacts their experience of school. Birch and
Ladd (1997) maintain this theme in their recent study of the connection between
quality of student teacher relationships and children's school adjustment. The
distinguishing factor between Birch and Ladd (1997) and much older studies of
teacher-student dynamics and student outcomes (Alexander & Entwistle, 1988;
Brophy & Good, 1974) is that their current study focused on the interaction between
relationship and several dimensions of student adjustment. Their view is that
relationships and relational dynamics are the basis for understanding student
adjustment and performance in school.
Using a sample of 206 kindergarten age children across 16 classroom
teachers, the authors paired the exploration of quality descriptors of the student
teacher relationship with measurement of student adjustment to school. School
adjustment was defined specifically as student performance, school affect and
attitude, and student involvement or engagement with the school environment.
The authors used several statistically reliable measures including the Student
Teacher Relationship Scale (Pianta, 1992); the Metropolitan Readiness Test
(Harcourt Brace Jovanovich, 1986) for an assessment of language, visual, and
quantitative readiness; the Loneliness and Social Dissatisfaction Questionnaire for
Young Children (Cassidy & Asher, 1992) to gauge children's affective experience of
school; the School Liking and School Avoidance Scale Ladd & Price, (1987) to
determine children's attitudes towards school; and the Teacher Rating Scale of School
Adjustment (Birch & Ladd, 1997) developed by the authors to document various
aspects of school adjustment. Classroom teachers completed the STRS and TRSSA
during the fall semester of the school year. Children were interviewed during fall of
their Kindergarten school year by trained research assistants.
Birch and Ladd performed a series of hierarchical regression analyses to
determine the relationship between student teacher relationship quality and school
adjustment. Each factor was regressed separately to clearly determine its contribution
to the variance in student outcomes. Their preliminary findings led the authors to
conclude that the relationship quality framework is useful in terms of investigating
the link between student teacher relationship quality and school outcomes for
The regression analysis (with relationship variables entered in varying order
into the analysis and controlled for gender and relational environment information)
showed that student teacher relationship quality ratings of closeness and dependence
accounted for a significant portion of the variance in student achievement as
measured by the MRT in Visual Skill (Closeness 7%, Dependence 7%) and Language
Skill (Dependence 3%, Closeness 5%). Children with close relationship ratings also
had higher MRT visual and language stanine scores than did children with less close
Analysis also demonstrated an important relationship between school affect,
attitude, and student teacher relationship quality. Teacher ratings of dependency
contributed a significant amount of the variance to ratings of children's loneliness in
school (3%). Both conflict and dependency explained a significant amount of
variance in reported school liking (conflict 3%, dependence 3%). Children in
classrooms with higher means for conflicted patterns of student teacher relationships
reported liking school less than their peers with lower ratings of conflicted student
teacher relationships. Also, students whose student teacher relationships were rated
as dependent reported liking school more than their peers with lower levels of
relationship dependency. Closeness ratings accounted for a significant portion of the
variance in school liking (17%).
School involvement and engagement was also explored within the context of
student teacher relationship quality. Student teacher relationships described as
conflicted or dependent were associated with lower levels of school engagement and
poorer attitudes towards school overall. Dependency in the student teacher
relationship was significantly related to children's reported school avoidance,
accounting for 4% of variance. Higher ratings of dependency led to higher reports of
school avoidance by children. Analysis of this association at the individual student
teacher relationship level revealed that dependency again explained a significant
portion of the variance in school avoidance (7%). Conflicted student teacher
relationships explained an additional 5% of the variance in school avoidance. Higher
ratings of dependency and conflict and the individual student teacher relationship
level was strongly related to higher levels of school avoidance.
Other areas of student adjustment that were assessed included levels of self-
direction in the classroom as well as level of perceived cooperation. Conflicted
student teacher relationship ratings at the classroom level accounted for 6% of the
variance in student levels of self-directedness. Higher levels of conflict in the
classroom was linked with lower levels of perceived self-direction in students.
Individual student teacher relationships rated as dependent accounted for 22% of the
variance in perceptions of their self-directed behaviors. Closeness ratings explained
an additional 25% at the individual student teacher relationship level.
Higher ratings of conflict and dependency at the classroom level were related
to lower levels of perceived cooperation (conflict 4% of variance, dependence 3% of
variance). Individual student teacher relationships rated as conflicted accounted for a
significant 52% of the variance in teacher ratings of cooperation.
The significance of dependent relationships in this sample may be interpreted
through the relationship framework. Birch and Ladd suggest that students may be
relying on teachers as a source of emotional support. Their level of investment with
the teacher then may impede or disrupt the remaining interest the child has in
engaging in the broader school environment. Conflict also appeared to be an
important factor in children's adjustment to school. Student teacher relationships
rated as highly conflicted were also related to poorer attitudes about school, lower
levels of engagement, and lower levels of compliance. Finally, closeness in student
teacher relationships is important to recognize and study. This relationship quality
was strongly related to student performance, attitude, and engagement with school.
The results of this study are commanding. Without discounting the
contributions of early work on understanding the link between teacher behaviors,
beliefs, and perceptions and student performance (Alexander & Entwistle, 1988;
Brophy & Good, 1974), the significance of this work is clear. The perceptions
teachers hold about their students appear to be significantly associated with children's
early academic performance, children's feelings of loneliness, school avoidance, and
school engagement. This study established important results around the idea that
student teacher relationship quality is related to academic performance in young
elementary students. Similar to previous studies, the authors focused on very young
children. Further study including older elementary populations is an important next
step for expanding practical understanding of how relationship quality impact student
Using attachment and self-system theories as frameworks, Lynch and
Cicchetti (1997) believe that investigation of children's perceptions about their
relationships is important to understanding how to support their success in school.
Other investigations have documented the importance of feelings of relatedness to
student adjustment in school (Lynch & Cicchetti, 1992).
Lynch & Cicchetti (1997) describe the shift in student-reported quality of
relationships that occurs from elementary to middle school by asking them about their
feelings of relatedness to others. Surveying 1,226 students in grades 2-8 in an upper,
middle-class school district, children were asked to answer questions about their
relationships with their mom, teacher, best friend, and siblings. The survey was
administered at the end of the year to maximize the child's time to develop and
sustain relationships with teachers and friends. The authors used the 17-question
Relatedness Questionnaire adapted from Wellborn and Connell (1987). Two sub-
scales, emotional quality and psychological proximity seeking, were extracted from
the measure for evaluation. Previous use of this measure indicates that it has
sufficient internal consistency (Lynch, 1992). The emotional quality scale yielded an
alpha range of .67 to .83 and the proximity seeking scale yielded a range of .83 to .93.
The authors hypothesized that children who reported optimal levels of
relatedness would report lower levels of proximity seeking. Responses reflecting
adequate patterns of relatedness were predicted to be balanced between quality and
proximity seeking. Students reporting deprived patterns of relatedness would show
low levels of emotional quality and high levels of proximity seeking. Students
reporting disengaged patterns of relatedness would report low levels of both
emotional quality and proximity seeking. Finally, confused relatedness patterns
would show high reports of quality and extremely high proximity seeking.
Analysis showed that statistically significant shifts in reports of
optimal/adequate relationships occurred between elementary and middle school age
students. For grades 2-5, 57% of students reported optimal/adequate relationship.
Relationships described as disengaged accounted for 23% of responses in grades 2-5.
For grades 6-8, 27% reported optimal/adequate relationships with teachers. Sixty
percent of students reported feeling disengaged from teachers in grades 6-8.
Alarmingly, 45.6% of the total sample reported a disengaged relatedness pattern with
their teachers. This pattern supports the importance of teacher-student relationships
and their impact on student attitude and investment in school. The authors offered
some cautions or explanation of their results that are noteworthy.
The authors suggest the significant differences between elementary and
middle school reports reflects a developmental shift in the adult-oriented nature of
elementary school settings to the more peer-oriented culture found in middle schools.
Lynch and Cicchetti (1997) also concluded that the results may have been a function
of administration differences during the study. Middle school age students were
asked to respond to questions about their teachers "on average" whereas elementary
age students were asked to respond with their primary teacher in mind. While this
study contributes important information towards understanding student perceptions of
relationship quality with teachers, the scope of the research was one-sided. The
teacher point of view is necessary for a fuller understanding of the link between
relationship quality, school adjustment, and student performance.
Several of the studies reviewed in Chapter 2 incorporated some measure of
school adjustment or estimates of academic performance into their design. Few, if
any, of these studies measured actual academic products as variables dependent upon
relationship quality between teachers and students. This may be due to the difficulty
of uniform measurement across multiple classrooms. No two teachers deliver
instruction nor would they uniformly apply standards for assessment. These factors
are driven as much by the teachers' level of training and their personality and teaching
style. The manifestation of these factors in a classroom are further tailored by student
characteristics and the dynamics of interactions between the teacher and their
students. As was introduced in Chapter 1, the current study intends to evaluate
student writing as a measure of academic performance. Careful consideration was
given to measurement and evaluation of writing based on the assumption that
assessment of writing varies from classroom to classroom. A standard measure and
evaluation procedure adopted by the state in which the current study occurred was
selected to equalize these differences between classes.
Childrens school experiences powerfully shape their achievement and social
developmental pathways. Teachers are especially influential in shaping a child's view
of himself as a learner and as a friend. This relationship can be particularly critical
for learning when children do not experience stability and support in the other
environments in which they live.
Children who experience stress-ridden relationships with adults can be
especially sensitive and fragile in their approach to school. Like parents, teachers
cannot be neutral in their interactions with their students. The patterns of interactions
between children and teachers constructs part of the foundation that supports social,
cognitive, and academic development. However, when students and teachers share a
strained relationship, the effectiveness of the support and instruction offered by the
teacher can be significantly impeded. One logical outcome of long-term stress in a
student-teacher relationship is diminished academic performance. Strained student-
teacher relationships may contribute to reinforcing behaviors that diminish the
effectiveness of a student's learning time or the quality of student learning behaviors.
When students and teachers become locked in a struggle of personality, power and
control, or mismatched expectations, the energy of both individuals often becomes
channeled into the effort of controlling or coping with the conflicted relationship.
Emotionally strained from regular struggle, teachers and students end up with less
energy to devote to instruction, guidance, engaged learning, and academic risk taking.
Students who enjoy a positive, nurturing relationship with their classroom teacher are
generally more emotionally receptive to learning and can benefit from instruction and
guidance from the teacher more than their peers who do not. While the studies
reviewed in Chapter 2 provide a foundation for understanding teacher student
relationship quality as a construct and address issues regarding its measurement,
available research that offers evidence of the link between relationship quality and
learning behaviors and academic performance continues to be limited.
The importance of studying the link between the quality of teacher student
relationships and learning is evident. One strong influence on the academic
performance of students is the quality of the relationship shared between students and
teachers. Interest in enhancing the quality of instruction in core academic areas
consumes a large proportion of professional development efforts. Statistical
documentation of the link between the student teacher relationship and academic
performance will contribute to professional practices and professional development in
several ways. An understanding of the student teacher relationship and its influence
on students can help individual teachers work through struggles they may experience
with particular students, it may guide teachers through some reflective study of their
teaching and classroom structures, and it can positively impact the experiences of
many children through professional development devoted to incorporating knowledge
and strategies that enhance the quality of the student teacher relationship (Pianta,
Empirical work describing student teacher relationships is invaluable to
building an understanding of what impacts student success as well as its potential as a
context for intervention. While progress has been made towards this end, few studies
have attempted to link the perspective of both student and teacher at the relationship
level, nor have they compared these with actual measures of student performance at
the upper elementary school grades. Exploring this combination of factors that
influence student experiences of schooling will contribute new information about
relationship dynamics that continues to be underrepresented in its recognition as a
context for supporting student success.
The current study examines the relationship between teacher student
relationship quality, at the student and teacher level, and student performance in
writing as a measure of academic achievement.
Two primary research questions frame this studys design and methods:
1. What is the relationship between teacher ratings of teacher student
relationship quality as measured by the Student Teacher Relationship Scale
and students ratings of the teacher student relationship quality as measured
by the My Teacher Scale?
2. To what degree are teacher and student ratings of teacher student relationship
quality related to student writing as an index of academic performance?
A two-level correlational design was used to determine the association
between the quality of teacher student relationships and writing performance while
controlling statistically for differences in student-level and classroom-level variables.
One dependent variable (writing performance) and five independent variables were
measured. Data were collected at two levels: the student level and the classroom
level. Student level data included: (1) the quality of student-teacher relationships as
perceived by the classroom teacher, (2) the quality of the student-teacher relationship
as perceived by the student, and (3) the student performance on a writing sample
collected at the onset of the study. Two independent variables were collected at the
classroom level (on the assumption that they are approximately the same for all
students in a class). Classroom level data included: (1) time devoted to writing
instruction and practice in each participating classroom, and (2) activities chosen by
teachers to reinforce and practice writing. Instructional activities used to support
writing development are also generally applied uniformly within a classroom and
were considered to be an influence on writing performance. Figure 2 shows the
variables for comparison:
This model led to the development of two hypotheses. First, students'
perceptions about their relationship with their teacher would be positively related to
relationship quality ratings by teachers. Teacher student relationship quality was
expected to be congruent for teachers and students. That is, teacher ratings reflecting
close teacher student relationships should be confirmed by a positive rating of
relationship quality by the student. Second, student performance in writing would be
positively related to the quality of relationships between students and teachers. After
controlling for time and type of instruction and differences in student performance on
the 'pre-test' writing sample, no significant differences were predicted between classes
that explain student performance on the final writing sample other than relationship
quality. Ratings of the closeness of relationships were expected to be positively
correlated with writing performance. Ratings of the degree of conflict in
relationships were expected to correlate negatively with writing performance.
Ratings of the degree of dependence in relationships were expected to correlate
positively with writing performance. Finally, student ratings of their relationships
with their teacher are expected to correlate positively with student performance in
Subjects and Sampling Procedures
The current study was conducted in a low income urban school district near a
major metropolitan area in a Western state. Demographic statistics for the district
show that the community ethnic composition is 47% Hispanic, 3% African American,
2% Native American, 1% Asian, and 47% Caucasian. The demographic composition
of the four schools included in the current study as well as the demographics of the
sample is reflective of the ethnic breakdown of the community. Families with school
age children have a median income of $18,000. During the 1999-2000 school year,
an average of 63% of students in the district qualified for the ffee/reduced lunch
program. Over one third of families are headed by single parents, including many
teen parents or seniors parenting their grandchildren. All of the district's elementary
schools are designated Title 1 schools based on the low income of families.
Teachers. Fourth grade teachers from four elementary schools were recruited
for participation in the study with a goal of securing a commitment from 6-8 teachers.
A total of seven teachers committed to participate in the current study. All teachers
were female and reported between 1 and 8 years of experience as a classroom teacher.
Teachers reported devoting between 60 and 120 minutes to writing instruction and
practice per day. Table 2 describes the traits of teachers included in the study.
Table 2. Teacher Demographics
Teacher Years of Experience Minutes Per Day for Writing
1 8 120
2 l 90
-> 2 90
4 1 90
5 4 60
6 4 60
7 2 60
Classrooms with a majority of monolingual Spanish speakers were not recruited for
participation in the current study. The association between the student-teacher
relationship, differences in instructional practices in bilingual classrooms, and the
process of dual language development and its impact on writing may be very different
from instruction and learning experiences of English speaking students.
Students. All students from each participating classroom qualified as a
participant in the study with two exceptions. Spanish speaking student and students
for whom 4th grade level standards were significantly modified through special
education accommodations were not recruited for inclusion in the sample. These
students were excluded because there are no alternative measures of accommodated
writing for these students that yield equivalent assessments of their writing
The current study included complete data gathered from 97 students across
seven different classrooms, including 46 boys and 51 girls. Data from nineteen
additional students with positive parent permission were included in classroom level
analysis of relationship quality. Data on their writing performance was not available
because they were absent during the data collection period or had moved from the
school. Positive parent permission was required for student participation in the study
(See Appendix E for Consent Letters). A total of 158 students were enrolled across
the seven participating classrooms. Twelve students were excluded from the study
due to their Spanish speaking status. Another two students were not considered for
the study based on the modified status of their educational program. Sixty-seven
percent of the remaining 144 students returned positive parent permission. Levels of
participation in the current study varied across classes, ranging from a high of 24
students to a low of six students.
Performance in writing, the dependent variable, was measured using the
procedures outlined by the Colorado Student Assessment Program (McGraw-Hill,
1993). Student writing samples were generated from prompts used in previous
official administrations of the CSAP. CSAP instructions to students included the
direction to plan, draft, revise, and edit their writing. Students produced two writing
samples during the course of this study as a measure of their performance in writing.
Samples were evaluated and scored using the Colorado Student Assessment Program
Writing Prompts. Two writing prompts were pre-selected for this study by the
researcher. All available prompts from released (available for public review) CSAP
items were reviewed and evaluated based on the appeal of the prompt to the
researcher. The researcher also considered the type of writing required by each
prompt (i.e. creative, persuasive, explanatory). After informally surveying literacy
specialist teachers in her building, the researcher selected prompts targeting creative
writing and persuasive writing. Complete administration procedures and the exact
prompts are provided in Appendix D.
Writing Sample Administration Guidelines. When samples were collected for
the current study, classroom teachers were given specific guidelines about the
administration of the writing sample. Teachers were encouraged to review
organization and planning skills with their students. They were also encouraged to
explain any words in the prompts that their students did not understand in an effort to
give the students the best possible conditions under which to write. As was
described earlier, students were given the direction to draft, edit, and revise their
CSAP Scoring Procedures. Writing samples were evaluated using the scoring
rubric developed for the state of Colorado to assess the development and performance
in writing of its 4th grade students. The Colorado Student Assessment Program
(CSAP) was first administered during the 1996-1997 school year and has been given
to students in 3 rd, 4th, and 7th grades since that time. The internal consistency
reliability of the CSAP writing assessment was estimated at .91 using Cronbach's
coefficient alpha (McGraw-Hill, 1993). The rubric is designed to assess student
writing at four levels: content and organization (0-3 points), style and fluency (0-3
points), language usage (0-2 points), and evidence of planning (0-1 point). Essays
earn a score ranging between 0 and 9 points. Scores are then categorized based on
predetermined performance levels indicating the quality of the writing compared with
state writing standards. These levels include: advanced, proficient, partially
proficient, and unsatisfactory. For the purposes of the current study, however, these
quality descriptors were not used. See Appendix F for the details of the rubric.
Writing samples were evaluated by three independent raters, and writing
sample scores were the average of the three raters scores. Raters practiced scoring
using the standard rubric before beginning to score writing samples used for the
study. Each rater was assigned to score 35 anchor writing samples released from
CSAP. The researcher assembled sample packets to reflect a full range of possible
scores. These samples were compiled so that scores assigned by CSAP raters were
unknown to the scoring team in this study. The rating team (including the researcher)
familiarized themselves with the scoring rubric and then scored practice samples
independently. Because the possible range of scores for writing samples was narrow,
differences of more than one point between the three raters were discussed to explore
how points were assigned.
Inter-Rater Agreement of Writing Sample Evaluation. A correlation analysis
was done to determine the level of agreement in writing sample scoring for the
current sample. Inter-rater agreement was analyzed for both the 'pre-test' writing
sample and the final writing sample. These results are described in Table 3 below.
Table 3. Inter-Rater Scoring Consistency for Evaluation of Writing Samples
Inter-Rater Consistency Rater 2 Rater 3
Rater 1 Sample 1 .907** .868**
Rater 1 Sample 2 .873** .839**
Rater 2 -Sample I .855**
Rater 2 Sample 2 .782**
** g < .01 two-tailed.
When evaluations of writing samples differed by more than 2 points between any of
the three raters, the samples were rescored (n=l 1).
Several steps were taken to prepare student writing samples collected through
the current research for scoring. To minimize possible bias, student identity,
classroom identity, and the essays designation as pre- or post- was removed and
replaced by a scoring template pre-marked with a research identification number.
Student Teacher Relationship Scale. Developed by Pianta (1992), the Student
Teacher Relationship Scale (STRS) is a survey completed by classroom teachers to
describe their relationships with their students. Based on work that combines adult-
child attachment theory with the importance of elementary school experiences in
establishing the pathways of student progress, the Student Teacher Relationship Scale
(STRS) has been used to clarify the link between relationship processes and student
adjustment. Further, it has been used as a tool to guide consultation with teachers
(Pianta, 1999). Responses are used to describe a student-teacher relationship along
three subscales close relationships, dependent relationships, or conflicted
relationships. Results from several studies show the total measure as well as the
subscales to have good internal consistency. Alpha ratings of internal consistency for
the STRS as a total scale were reported at .89. The reliability of subscales describing
relationship quality are as follows: closeness .90, conflicted .93, and dependent .69
(Birch and Ladd, 1997; Pianta and Steinberg, 1992; Pianta, 1996; Saft, 1994).
Construct validity of the three subscales was supported by factor analysis (Pianta,
1994). Teacher responses are recorded using a Likert format. Items on the STRS are
worded such that scale meaning is uniform. High scores on a particular subscale are
interpreted as a characterizing the relational quality of that teacher student
relationship. Mean scores of whole class STRS were used to describe the quality of
the relationships between students and teacher at a systemic level using the same
subscales (Birch & Ladd, 1997). See Appendix B for a copy of the measure.
A principal component factor analysis of the STRS yielded three distinct
factors. These factors were congruent with items contributing to the three
relationship categories described in Pianta's standardization work around the STRS.
Estimates of internal reliability coefficient alpha based on responses from the current
study were as follows: closeness .87; conflicted .92; dependent .78 (n=97).
Mv Teacher. My Teacher is a brief 8-item measure assessing elementary age
students' perceptions about the quality of their relationships with their classroom
teachers. Developed as one part of a package of tools to be used with the ClassMaps
program (Doll, Zucker, Brehm, 1998), it has been used to aid consultants and
graduate level school psychology students in their analysis of the health of a
classroom and of particular teacher-student relationships. The tool was originally
adapted from a measure developed by Ainsworth, Bell, and Stayton (1971) to assess
the degree of maternal sensitivity and responsiveness in mother-child interactions.
Their work reflects the importance of attachment systems in developing strong
relationships that, in turn, support cognitive and social development. Healthy
attachment relationships have also been linked with the development of positive peer
relationships, emerging literacy, and the capacity to problem solve in the school
context (Pianta & Walsh, 1996).
In a recent extension of the ClassMaps model to the elementary level, My
Teacher was modified for readability and understanding by elementary age students.
Results from standardization probes used with elementary populations indicate that
the scale possesses sufficient construct validity and reliability (Alpha = .84) for
further use. In the current study this tool was used to gather information efficiently
about student perceptions regarding their relationships with their teachers.
My Teacher was described to students as a way for the researcher to learn
about how things are going for them with their teachers (Doll, Zucker, Brahm.
1998). It was administered using the prompt, Circle the one that is true for you the
answer that makes the most sense. Response choices included yes (scored as 3). no
(scored as 1), and sometimes (scored as 2). Scores range between 8 to 24 using this
system of assigning points. These scoring conventions were slightly different, but
congruent with scoring procedures used in the standardization of My Teacher. Scores
of zero were used to denote missing data rather than reflecting a score on this
measure. Results from My Teacher can be used to evaluate the relationship quality
between one student and their teacher as well as lending itself to whole-class
aggregate scoring. The resulting classroom profile can provide a broad measure of
student perceptions of the relationship they share with their teacher. See Appendix A
for a copy of the measure.
A principal component analysis of My Teacher was done to confirm that all
items were measuring a common construct. The items from the My Teacher scale
clearly clustered into one principal component with factor loadings ranging from .357
to .761. All but one item loaded above .400 in the principal component. Internal
consistency for My Teacher was estimated at .71 (n=97), indicating an acceptable
level of internal consistency.
Measurement of Other Variables. Using a brief, informal measure developed
by the researcher, teachers were surveyed about their years of experience in the
classroom, the total time per day spent on writing instruction and support activities,
and the types of writing activities they used to reinforce instruction. Working from
the argument outlined in Chapter 1 that writing is a relationship dependent process,
the researcher wanted to compare descriptive data regarding daily time spent on
writing and relationship quality at the classroom level. The content of writing
instruction provided by teachers may also contribute important descriptive detail to
understanding writing performance as well as relationship quality in this sample.
Activities that supported writing instruction were grouped into two categories: whole
text writing activities and mechanics based writing activities. These activities and
categorizations were chosen based on a review of texts devoted to instructional
techniques for writing. The content of this survey can be found in Appendix C.
Examples of each type of writing activity are described in Table 4.
Tabic 4. Writing Activities
Whole Text Writing
Planning (webbing, brainstorming)
Writing to a prompt
Responding to text (reactions)
Organization of writing
Mechanics Based Writing
Editing own work
Editing teacher samples
Other spelling activities
Data Collection Procedures
Data collection occurred at two points during the 2000-2001 school year. The
first data collection occurred during late October, 2000 and the second occurred
during January, 2001. Each measure was administered using a standard set of
instructions. Writing sample administration was conducted in accordance with the
guidelines developed for use with the standardized administration of the CSAP
(Colorado Student Achievement Profile) writing test. These guidelines and the rubric
itself are included in Appendices D and F. The 'pre-test' student writing sample was
collected within a two-week period of time during the first data collection period.
The second data collection period occurred during January, 2001. The final writing
sample was administered during this time, also administered and completed within a
two week time period. Classroom teachers chose to administer the writing prompts to
their full classes regardless of permission status. They viewed the writing prompt as
a good opportunity to practice for the upcoming CSAP.
Teachers and students completed their respective surveys about relationship
quality during the final data collection period. Participating teachers were asked to
complete the Student Teacher Relationship Scale (Pianta, 1992) for their students.
Students completed My Teacher as a whole class with the researcher. Directions and
individual questionnaire items were read aloud to students to address variation in
reading proficiency within the class. Classroom teachers were asked to leave the
room for 15 minutes to allow the researcher to explain the measure and to complete
the questions with students. Students were assured that their individual answers
would not be shared with their teachers. Classroom teachers completed their informal
survey regarding time and type of instruction devoted to writing and provided
information about students' previous reading achievement records.
Data Analysis Procedures
All survey responses were scored and entered into a data management file
using SPSS software, Version 10.0. Data were coded according to predetermined
instructions related to each measure. The current study is exploratory in nature. In
order to avoid overlooking important results, alpha was set at. 10. Data analysis
occurred at two levels student and classroom.
Three types of statistical analyses were applied to the data set to the answer
the research questions posed through the current study. Research question one was
answered by applying a Pearson Correlation analysis. Research question two was
approached in two stages. In the first stage, multiple regression using student level
predictors were applied across classrooms to account for variance in post-test writing
performance. This analysis was carried out using raw score relationship scales and
also relationship scores centered around the respective classroom means. The
classroom means of the post-test writing were then compared using Analysis of
Covariance (ANCOVA) to determine whether unexplained differences remained
among classrooms to be explained by differences in the quantity or type of writing
Chapter 4 describes the results of the current study in detail.
Analyses were completed at two levels (student and classroom) to determine
how data from the current study supports the research hypotheses. Data from the
current study were analyzed using SPSS Version 10.0. Results are presented
following the order of the two hypotheses. To review, the main research question and
guiding hypotheses are as follows:
Hypothesis I: Student and teacher ratings of teacher student relationship quality will
be significantly correlated.
Hypothesis II: Teacher ratings of teacher student relationship quality will be
correlated with student performance in writing. These correlations are predicted:
Close relationship quality ratings will be positively correlated with writing
Conflicted relationship quality ratings will be negatively correlated with
Dependent relationship quality ratings will be positively correlated with
Positive relationship quality from the student perspective will be
positively correlated with writing performance.
A two-level correlational design was used to identify the association between
the quality of teacher student relationships and writing performance while controlling
statistically for differences in student-level and classroom-level variables. Analysis
was conducted at two levels. First, student level data was analyzed. Performance in
writing at the student level is hypothesized to be stronger for children who experience
close relationship quality with their teachers than for children who experience
conflicted or dependent relationships with their teachers. Performance on the final
writing sample should reflect a child's performance on the 'pre-test' writing sample
and the quality of the teacher student relationship. After controlling for performance
differences in the pre-sample, results should reflect a correlation between student
performance on the final writing sample and the quality of the teacher student
The second level of analysis focuses on the classroom level. Because multiple
classroom teachers participated in the current study, import differences between
relationship quality and student writing needed to be explored. Writing performance
is hypothesized to vary across classes due to the time and method devoted writing
instruction as well as the quality of teacher student relationships in the class.
Results from student level data analyses are presented first, testing the
hypothesis that teacher and student ratings of relationship quality correlate. The
connection between relationship quality and performance in writing is presented next.
Data analysis across classrooms tested the hypothesis that there are significant
differences among classrooms in writing performance after controlling for 'pre-
writing' differences. Chapter 4 concludes with a presentation of results from a
regression analysis to determine the amount of variance in the dependent variable
explained by relationship quality and other variables.
Approach to Analysis
Analysis of classroom level data of relationship quality and writing
performance was conducted using two approaches, one using raw scores, and the
other using scores centered around each classroom's mean for each relationship
quality type. The use of raw scores as a measure of relationship assumes that the
meaning of a score remains constant from classroom to classroom. However,
inferences about students and teachers applying similar standards when evaluating
their relationship quality with one another may misjudge the meaning they give to a
particular score. Individual evaluations of relationship quality depend on that
person's relationship history, their beliefs about relationship parameters in a given
context, and their perception of the relationship in comparison with other
relationships in that environment (Pianta & Walsh, 1996). Using a raw score
approach to this analysis also suggests that the score describing the individual
relationship is the strongest predictor between relationship quality and writing
performance conceivably overlooking the effects of the individual relationship
compared with other teacher student relationships in that classroom.
The other method involved calculating a centered scored based on the
difference between the classroom average and reported scores from individual
students and teachers. If the assumption is correct that relationship quality is not
absolute, but context dependent, then the numeric reflection of a student's view of the
relationship or the teachers view of relationship quality is expressed most
meaningfully in terms of how it compares with other relationships in that setting.
Initially, no compelling justification arose to exclude one avenue of analysis over the
other, so both were executed. However, results of analysis using centered scores
showed more powerful statistical relationships between the independent and
dependent variable than was shown using raw scores. In the sections that follow, raw
scores are used first to compare relationship ratings across classrooms. Both raw
scores and centered scores are then used in correlational analysis to examine the
relationship with writing performance. A maximum alpha level of. 10 was used for
all statistical tests.
Teacher Student Relationship Quality
Student. Teacher student relationship quality was measured using two self-
report questionnaires, one version for students and another for teachers. My Teacher,
used as a measure of student perception of their relationship with their primary
classroom teacher, resulted in generally positive perceptions of relationship quality.
Scores could range between 8 and 24 points. Higher points reflect a more positive
perception of the teacher student relationship by the student. Total mean scores for
My Teacher and STRS are reported in Table 5.
Table 5. Mean Raw Scores for Relationship Quality by Teacher
1 2 3 4 5 6 /
Conflict 19.4091 21.6087 22.5714 29.9583 20.8571 21.2500 20.8571
SD 4.8370 9.4568 9.5369 12.0560 4.7409 9.9121 4.5981
Closeness 47.2727 39.6957 39.0000 45.7500 36.4286 38.5000 46.8571
SD 6.1504 9.7814 6.5574 5.8328 6.1334 2.6458 5.0143
Dependent 7.3182 10.8261 12.0000 13.2500 11.1429 10.2500 12.7143
SD 3.5642 4.4380 4.5826 3.9259 3.1848 5.9652 5.0897
My Teacher 20.82 20.70 22.00 21.58 20.43 22.25 22.14
SD 2.50 3.30 2.36 1.79 2.37 2.22 1.68
Distribution of student reports of relationship quality are reflected in the box graph in
1 2 3 4 5 6 7
Figure 3. Aiy Teacher distribution by classroom
Figure 3 reflects that individual student ratings were positive overall. Students in
Teachers One and Two's classroom showed some differences in the way they view
their relationship quality with their teachers. Student responses in Teacher two's
class showed the greatest variation in their perceived quality of teacher student
relationship. An analysis of variance showed no significant differences in mean
ratings of teacher student relationship from class to class using this measure (F6.86 =
1.039. g = .419).
Teacher. Using the Student-Teacher Relationship Scale (Pianta, 1992; 1996),
classroom teachers were asked to evaluate their relationship quality with their
students. Item analysis yielded three distinct factors congruent with Pianta's original
standardization analysis of the measure. They fell into what he describes as the
closeness, conflict, and dependent categories of relationship quality. Total mean
scores for each of these scales were calculated for each teacher as shown in Table 5.
The distribution of mean total scores for relationship quality is clear in the
graphs in Figures 4, 5, and 6.
1 2 3 4 5 6 7
Figure 4. Distribution of Conflicted Relationship Quality Ratings by Classroom
O 20 i _ _ _ ___
1 2 3 4 5 6 7
Figure 5. Distribution of Close Relationship Quality Ratings by Classroom
1 2 3 4 5 6 7
Figure 6. Distribution of Dependent Relationship Quality Ratings by Classroom
Teacher Four reported more conflict in her relationships than the remaining
teachers. She also rated more of her relationships as dependent than the other
teachers. Teacher One's reports of relationship quality with her students showed the
highest mean for close relationships and the lowest mean for relationships
characterized by conflict. She also rated fewer of her relationships with her students
as dependent than the other teachers. It is noteworthy to mention that this teacher
reported the most years of experience (8 years) in the group. An analysis of variance
showed significant differences between mean scores of relationship quality reported
across classrooms (Conflict F6 87 = 3.288, g = .006; Closeness F6 87 = 4.780, p = .000;
Dependent F6.87 = 4.323, p=.00l).
Teacher and Student. A correlation analysis between relationship quality
measures for teachers and students was conducted using two approaches, one using
raw scores and the other using centered scores for each scale. Table 6 shows
correlations between relationship scales using raw score data.
Table 6. Teacher Student Relationship Quality Ratings Raw Score
Scale My Teacher Conflicted Close
Close .230** -.254**
Dependent .111 .573*** .009
* "correlation significant at the .05 level (2-tailed)
*** correlation significant at the .01 level (2-tailed)
Relationship quality rated as close by teachers showed a small but significant
correlation with positive ratings by students (r = .230, p = .027). The link between
relationship quality ratings by students and conflicted and dependent relationship
ratings by teachers were not strong, nor were they statistically significant. Close
relationship quality ratings by teachers correlated negatively with conflicted
relationship quality (r = -.254, p=.013). Teacher ratings of dependent relationship
quality correlated with conflicted relationship quality at a highly significant level (r =
.573, p =.000). Features that characterize conflict or dependency in a teacher student
relationship may overlap somewhat in teacher perceptions. Finally, close relationship
quality did not correlate with dependent relationship quality ratings.
The same correlation analysis was conducted using centered scores for each
classroom. Patterns of statistical relationship using centered scores were similar to
those using raw score data with one important distinction. Correlation relationships
using centered scores showed stronger statistical relationships than those resulting
from raw scores. Table 7 shows these results.
Table 7. Teacher Student Relationship Quality Ratings Centered Score
Scale My Teacher Conflicted Close
Close .255** -.381***
Dependent .054 .522*** .074
correlation significant at the .05 level (2-tailed)
*** correlation significant at the .01 level (2-tailed)
This analysis also reflected a significant relationship between close relationship
quality ratings from teachers and positive relationship ratings from students (r = .255,
E = .014). Conflicted relationship quality correlated negatively with student ratings of
relationship quality at a nearly significant level. Student ratings of relationship
quality were less positive for relationships characterized by conflict. Centered score
correlation analysis showed a stronger statistical relationship between teacher ratings
of close and conflicted relationship quality (r = -.381, p = .000). Relationships rated
as dependent were statistically significantly related to ratings of conflict in
relationship quality (r = .522, p = .000).
Relationship Quality and Writing Performance
Hypothesis two states that teacher ratings of teacher student relationship
quality will be positively correlated with student performance in writing. To explore
how the data from the current study confirm this hypothesis, further analysis of the
link between relationship quality and writing performance was conducted. Table 8
shows results from a partial correlation analysis between relationship quality and
writing performance using raw score means and centered score means. Because
student performance on the pre-test writing sample correlated significantly with
performance on the final writing sample (r = .546, p = .000), the analysis was
executed controlling and not controlling for the link of pre-test writing scores with
final writing scores.
Table 8. Correlation Analysis of Relationship Quality and Writing Performance
(Controlled and Uncontrolled for Pretest Writing Performance)
Relationship Quality Measure Final Writing Sample Performance (Pretest Writing Uncontrolled) Final Writing Sample Performance (Pretest Writing Controlled)
My Teacher Raw Scores -.003 .005
My Teacher Centered Scores .009 .017
Dependent Raw Scores -.067 -.042
Dependent Centered Scores -.107 -.085
Conflicted Raw Scores -.147 -.036
Conflicted Centered Scores -.152 -.081
Close Raw Scores .195* .193*
Close Centered Scores .284** .207*
* significant at. 10 (2-tailed)
significant at .05 (2-tailed)
With the exception of a small, but significant, correlation between close relationship
quality and writing performance, relationship quality overall was not statistically
related to student performance on writing. Teacher ratings reflecting close
relationship quality were significantly correlated with writing performance using both
raw and centered score means for each class.
Additional analysis was completed to establish how the combined factors of
relationship quality and the pre-test writing sample explained variance in student
performance in writing. Controlling for differences in performance on the 'pre-
writing' measure and comparing variance explained using both raw score relationship
quality data and centered score relationship quality data, a regression analysis yielded
the following results:
Table 9. Regression Analysis of Student Performance in Writing Raw Score Data
Change in r2
Score on Writing Sample One
3 rd Grade Reading CSAP Score
Close Relationship Quality Raw
Conflicted Relationship Quality Raw
Dependent Relationship Quality Raw
* Beta significant at. 10 (2-tailed)
** Beta significant at .05 (2-tailed)
Beta significant at .01 (2-tailed)
Table 10. Regression Analysis of Student Performance in Writing Centered Score
Variable Change in r2
Score on Writing Sample One .299***
3rd Grade Reading CSAP Score .136***
Close Relationship Quality Centered .081**
Conflicted Relationship Quality Centered .023
Dependent Relationship Quality Centered .012
* Beta significant at .10 (2-tailed)
** Beta significant at .05 (2-tailed)
*** Beta significant at .01 (2-tailed)
This analysis shows that student scores on writing sample one explains approximately
one-third of the variance in student performance on writing sample two. The conflict
(2/10 of 1%) and dependency (1/10 of 1%) relationship quality scales explain very
little of the variance in student performance on writing sample two. The closeness
scale, when using centered score data, accounted for approximately 4.3% of the
variance in student performance on writing sample two (p = 10).
Relationship Quality. Writing Performance, and Other Classroom Factors
Types of writing instruction employed by classroom teachers were surveyed.
Analysis revealed a strong negative correlation between the use of whole text
activities to support writing instruction and writing activities focused on mechanics (r
= -.716, p = .071). Teachers in this sample delivered writing instruction that tended
to follow one or the other broad type of activities to support writing development.
The type of writing activity used by teachers was correlated with relationship quality
ratings by teachers. Whole text writing activities (See Table 2 in Chapter 3 for
examples) and all categories of relationship quality were negatively correlated for the
seven classrooms in this study. This correlation was strongest between close and
conflicted relationship quality and whole text writing approaches. Teachers
describing their relationships with students as close, conflicted, or dependent tended
to employ fewer whole text writing activities to support writing instruction.
Mechanics based writing activities showed moderate positive correlations with close
and conflicted relationship quality.
Correlations between time devoted to writing instruction and relationship
quality were not statistically significant despite the reported range of time spent on
writing per day. Table 11 details the results of this analysis.
Table 11. Writing Instruction, Time, and Relationship Quality
Relationship Quality Time Whole Text Writing Mechanics Based Writing
Close .455 -.659 .479
Conflicted .073 -.664 .564
Dependent -.522 -.220 .184
My Teacher -.340 .120 .095
A correlation between the adjusted mean scores for writing sample two and teacher
reports of writing activities used in their classes was conducted to extend this analysis
one step further. Results revealed very little statistical relationship (whole text, r = -
.256, g = .580; mechanics, r = -.060, g = .899) between performance on the final
writing sample and the methods of instruction employed by teachers in this sample.
The statistical link between time devoted to writing instruction and performance on
writing sample two was also weak (r = .493, g = .261). Mean student performance
on the final writing sample did not differ statistically by classroom (F61S = .598, g =
.731). Because very little statistical variance in writing performance between
classrooms remains, no hierarchical analysis was done to include time and type of
Additional analysis was conducted to correlate writing sample components
and relationship quality to further explore the relationship between particular
components of writing and their sensitivity to student-teacher relationship quality.
Table 12 shows correlation relationships between writing sample components and
relationship quality using raw score classroom means and controlling statistically for
student performance on the pre-test writing sample. A conflicted relationship quality
correlated negatively with components of the writing sample. High ratings of
conflicted relationship quality correlated with lower performance in each of the
component areas evaluated in the final writing sample. Close relationship quality
correlated significantly with student scores in content and organization.
Relationships rated as dependent in quality correlated negatively with student
performance on writing sample two.
Table 12. Writing Sample Two Components and Relationship Quality (Raw Score
Writing Components Conflicted Close Dependent
Content/Organization -.008 .296*** -.054
Style/Fluency -.138 .146 -.052
Language Usage .129 -.065 .039
*** significant at .01 (2-tailed)
Table 13. Writing Sample Two Components and Relationship Quality (Centered
Writing Components Conflicted Close Dependent
Content/Organization -.045 .211** -.075
Style/Fluency -.146 .212* -.069
Language Usage significant at .10 (2-tailed) ** significant at .05 (2-tailed) .07 -.056 -.048
Table 13 reflects the analysis of writing sample components and relationship
quality using centered scores for each classroom controlling statistically for pre-test
scores. Higher levels of conflict in the teacher student relationship was again
correlated with lower scores in the subcomponents of the writing sample. This same
pattern was also reflected in the correlation between dependent relationship quality
and the sub-pieces of the writing sample. Relationship quality rated as close
correlated significantly with two areas of the writing sample: content and
organization and style and fluency. This result shows that teacher perceptions of
close relationship quality is statistically related to student performance on writing.
A full discussion of the interpretation and implications of the results is
presented in Chapter 5.
The current study investigated the link between student teacher relationship
quality and student performance in writing. A two-level correlational design was
used to determine the association between the quality of teacher student relationships
and writing performance while controlling statistically for differences in student-level
and classroom-level variables.
Discussion of the results are grouped based on research hypotheses proposed
in the current study followed by a explanation of the implications of the findings at
the theoretical, research, and applied level. Limitations of the study are considered.
Chapter 5 concludes with a discussion of suggested directions for future research.
Summary and Explanation of Results
Results from the current study offer partial confirmation of Hypothesis I.
Both raw and centered score approaches to analysis of teacher student relationship
quality showed significant correlations between close relationship quality ratings by
teachers and positive ratings by students. Relationship quality ratings by students
correlated minimally with conflicted and dependent relationship quality ratings from
teachers. Close teacher ratings and high scores on My Teacher were congruent as
expected. Teachers who view a relationship as positive likely engage in behaviors
with those students that confirm their perceptions. The positive relationship
experience is mutually reinforcing. The statistical relationship between relationship
quality ratings by students and conflicted relationship ratings by teachers was
expected to be negative. While the analysis showed a very small negative correlation,
it was statistically significant. One explanation could be that 4th grade students
generally have a positive regard for their teachers and tended to show patterns of
responses reflecting this. That is, when surveyed about their perceptions about the
relationship they share with their teacher, responses of 4th grade students in this
sample tended to be very positive overall. Another possible explanation for this result
could involve the relationship construct measured by My Teacher. A comparison of
item wording on My Teacher and the STRS led to the conclusion that items from My
Teacher may not assess perceptions of relational quality in the same manner achieved
through the STRS. The limited number of items on My Teacher may also have
narrowed the capacity of the measure to estimate a range of relationship quality
indicators from the student perspective.
A similar explanation can be offered for the link between My Teacher and
dependent relationship quality ratings from teachers. Students whose teachers
described their relationships as dependent could be expected to rate their relationship
with their teacher positively. If observed, they would probably appear to be very
engaged with the teacher at a behavioral level. However, a teacher student
relationship characterized by dependency could impact student performance in one of
two ways. A student emotionally dependent on a teacher may use the attention they
get to improve their class work. Although the teacher may feel drained by the
students' lack of confidence and independence, the student continues to make
progress. An alternative scenario could involve a student who continually engages
the teacher's attention and support, but does not apply that support to their learning,
possibly diminishing the student's feeling of support from the teacher.
A second research hypothesis was considered in the current study. Its scope
went beyond the link between teacher student relationship quality to explore the
association between relationship and writing performance.
Hypothesis II: Teacher ratings of teacher student relationship quality will be
correlated with student performance in writing. These correlations are predicted:
Close relationship quality ratings will be positively correlated with writing
Conflicted relationship quality ratings will be negatively correlated with
Dependent relationship quality ratings will be positively correlated with
Positive relationship quality from the student perspective will be
positively correlated with writing performance.
Correlation Analysis. Hypothesis II was partially confirmed through the
results of the current study. The most important result related to this hypothesis was
the significant correlation between close relationship quality and writing
performance. Teachers who perceive their relationships with students to be close and
positive likely engage in behaviors that maintain this perception. Likewise, "good"
students may be more likely to connect with their teachers, contributing to the
establishment of positive patterns of interactions from the beginning of their
relationship. This could include praise, attention, encouragement, and demands for
high work quality. Students who receive this confirmation and support will likely try
harder and persist even if a task is difficult.
While analysis revealed very little statistical relationship between conflicted
quality and writing performance in this sample, the negative direction of the
correlation is in agreement with the prediction made in Hypothesis II. Predictions
made regarding the link between dependent relationship quality and writing
performance were not confirmed in the current study. The correlation between these
factors was not statistically significant, nor was it congruent with the predicted
direction of the relationship.
Low writing performance was expected for conflicted teacher student
relationship ratings. Teacher student relationships characterized by conflict and strife
can exhaust the attention, energy, and attitude that supports learning. Students who
experience conflict with their teacher may also exert less effort to produce quality
work as a way to control the relationship or are unable to produce quality work
because most of their attention is devoted to managing their feelings and responding
to relationship dynamics. The opposite effect of conflict between teachers and
student should also be considered as a possible explanation for the low correlation
between conflict and writing performance found in the current study. Some students
may be able to compartmentalize their feelings about how relate with their teachers
and achieve in spite of it. Because learning is largely a social and relational process,
however, evidence of this explanation probably would not be likely in actual
Regression Analysis. Regression analysis applied to Hypothesis II showed
that conflicted, dependent, and student perceptions of relationship quality explained
virtually none of the variance in student writing performance. Close relationship
quality ratings by teachers explained a small, but statistically significant portion of
variance in writing performance (4.3%). This result indicates that positive
relationship quality is related to student performance. Several possible explanations
exist for the lack of statistical association between student assessment of relationship
quality and writing performance.
Examination of item wording and content in My Teacher suggests that
questions focused on behavioral features of the teacher student relationship and not
on indicators of quality. Further, the lack of variance in My Teacher responses
overall may have further weakened its power as a predictor of writing performance.
The total sample size and the modest numbers of available relationship quality-
writing performance examples within each descriptive category could also have
diminished the power of the analysis. The small link between relationship quality and
writing performance could be attributed to a number of factors. One explanation is
measurement. Using a survey format for assessing relationship quality may not have
been optimal for measuring the association between teacher student relationship
dynamics and student writing. A one-time assessment of relationship quality may
lack the power to portray the ongoing relational dynamics between a student and
teacher. The size of the student sample analyzed for this study may have impacted
the sensitivity of the measures to reveal a statistical relationship. Time may have
also impacted the link between relationship quality and writing. Pre and post-test
writing samples were collected within five months of each other. Final writing
samples collected later in the school year may have reflected greater student
development in writing. Finally, the researcher recognizes that writing performance
was certainly affected by a wide range of factors that were not measured in this study,
only one of which is relationship quality. Pianta and Walsh (1996) suggest that a
child's view of school and subsequent performance in school is affected by a number
of factors outside the setting of the classroom. Some of these include: procedures
and structures specific to the school a child attends regarding instruction, other adults
available at school to support learning, the socioeconomic status of their family and
neighborhood, and their parents' experience and attitudes towards school. Some of
these factors could have impacted the results of the current study.
Ratings of conflicted and dependent relationship quality by teachers also
explained very little of the variance in student writing performance. Dynamics of
relationships characterized as dependent or conflicted appear to be more complicated
than anticipated in their impact on student performance. Teacher student
relationships described as dependent may actually be supportive from the students'
point of view, encouraging them to perform well in the classroom. The opposite
effect could also occur. A student could consistently seek reassurance and guidance
from their teacher driven by an emotional need, but put less effort into engaging in
Conflicted relationships could also result in two distinct achievement profiles.
Friction between teacher and student logically affects student performance negatively.
However, the opposite effect could occur as well. Teachers and students may
experience conflict, but with no apparent effect on student performance. This may
clarify why the variance explained by conflicted and dependent relationship quality
was so low. Both types of relationship dynamics may have been at work, in effect
canceling out the influence of one another.
Integration and Implications of Findings with Previous Research
Convergent Findings. Groupings of teacher responses on the Student Teacher
Relationship Scale from the current study were concurrent with factors established by
Pianta (1992), adding to the substantial evidence that the STRS is a reliable measure
of relationship quality. Correlations between relationship quality descriptors from the
current study also aligned with early standardization research using the STRS. The
subscales measure distinct features of relationship quality.
Close STRS ratings were significantly associated with student performance in
writing. This is similar to the findings of Birch and Ladd (1997) to the extent that
both studies examined some aspect of student achievement and its link to relationship
quality. These findings add to the evidence that positive relationship quality between
teachers and students have some relationship with student outcomes.
Divergent Findings. Student and teacher ratings of relationship quality
explained little of the variance in student writing. Close relationship quality as rated
by teachers explained 4.3% of the variance in growth in student writing. In their
study of the association between STR quality and a test of school readiness. Birch and
Ladd (1997) found that close and dependent relationship quality explained variance in
visual and language skills at a significant level. Several differences between the
current study and Birch and Ladd (1997) could explain the disparity in results. The
current study analyzed data from fewer subjects than were included in Birch and
Ladd's study. With the exception of using the STRS as a measure of teacher
perception of relationship quality, the current study used different measures as well as
an older population of students. The results found by Birch and Ladd indicated that
variance can be explained by relationship quality. While close relationship ratings
from the current study explained variance in writing performance as a significant
level, results did not confirm a relationship between conflicted and dependent
relationship quality and writing. More discrete information is needed to understand
how the dynamics of dependent or conflicted relationship quality affects individual
students. This could be achieved through a more qualitative approach using
observation and interviews.
Contributions of findings to literature. Results from the current study
contribute unique information to the literature about teacher student relationships.
The current study compared measures of student perception of relationship
quality with teacher perception of relationship quality. Positive relationship quality
ratings from students correlated significantly with close relationship quality as rated
by teachers. No statistically significant relationship resulted between student ratings
of relationship quality and teacher ratings of conflicted or dependent relationship
quality for this sample. While the results correlating student and teacher ratings of
relationship quality were not significant across all types of relationship quality, the
results are important nonetheless.
No previous studies have examined the perceptions of relationship quality
from both student and teacher perspectives. These findings offer evidence in support
of the importance of positive, close relationship quality between teachers and students
to student performance. The lack of statistical relationship between student
perceptions of relationship quality and conflicted and dependent ratings by teachers
offers little to expand practitioner understanding of the impact of these relationship
characteristics on student learning except to establish that more research is needed.
Results from this study showed very little statistical relationship between
conflicted and dependent relationship quality and writing performance. This adds
50@70/67A@ @BC/5 DC $)"5$=33= )"5$=33; )"5@) @/" 7) ?99=
@/" #" DC ) IG @) '@ ):(
C$D,/ @ $7)6"B#/ "/@) '5 %.)#%% &"%#%# ".).)) #&"%#"% %# 5&"! "#)%# "J # @)
50/ = @5 /" 7" 8 C#%( ; 5%(5" F 6 == #" =8 @ =; ? KB07 => %( =F =< =3 ( ?= ?? 5 ;> 8 0@7CA ;< ;< L#/ >9
>= >8 >; B#/ >; B#C >; 5/#/ >> .# >1 >F >< K 19 @5/ 1= @"/ 1? ; 7 1; 11 6" 1F 1F 13 F? 6"B#/ F; M
6"$B#/$ K F< > @5 "! 5" <; #" <1 ##%/) << 5)## << @)## <3 5#7 39 # 3= 3= 3? 3? 7 38 @#K" 38 !K"C*" 38 "/% 3; 3; 77@ 3> K
//@5 3F 3F 3< 5 )" =99 @ B#/ =9= 57 =9? 5/# =91 5 =9< )
C # # B# 3 #? ;3 #8 "@ 1< #; @56"# F9 #> @56"# F9 #1 @@6"# F= NO!
'7 7)#@) > ? @# >? 8 .#5" >1 ; B#) 1= > 6" 1< 1 6"#.% F? F 6"#.5F8 < 5"6"B# F> 3 #"B#.% F1 =9 #"B#.5FF == B#$$6" F3 =? B#56".% <9 =8 B#5 6".5 <=
50/= @5 /" "!(%&" %# J#" L,/ $=3F<5#% )"$$#$% $%)$)L! #$"$ # &)#(%&" )% ##$ % )$ )$)"$ $##,/$=333N/ PB$=331
"% "$!$),'"P C$=3F;B)#$ #J#% J! "),/PB$=331%(" %,&)-&" %#"J) 2),!P%$=3<
&("$##% (%#&" 7" 5%(# !),/PB$=331N/$ =333%)!%&" ),'P7$=33F" (%&" )! )%#*# B"#* ))$. "(% )) $)# #$ $% ") $"#&"( !)%( )%#.$) ""!# Q 8 M
,/PB$=331$ )"##% (%&" "$ ".&"( !.) C#%( 5?)##%(#% &" ) "$%($($# ) %($#0@) "',=3F3-$" 'J#0@),=3F3-) ##%("" )% %) $!# %(!" ")% 'J '#)#" ",$=33F" )),'$=3F
))# %= = 7)#@) )$$ !")##)#% " #%#% )#)", # J #$$##!" #)))# )$%) $"$%# #)# 'R " 'J#)) ) !") %. %
) %'2#0@) )$",$$)-& L$$(%## ,44)(" !) .$ %$$ ."! B'J L#))&" $% ))) /))( )!% %$)$$#$ $#) ',=3F3-* )"#%)! J$#$%"%" ", "$#$$"&"!)!
&"$$!% )%J )#,B$=33?J! # &"%%"% ") "%( &" 5C#%(5" ',=3F3-#")) "$'%## ###",$(#)&" ) ) !)% $&""J# B %$#% "#))$#$ /$# '$") &" %)"#
" %# (%&" %# ##) $)$") ,/PB$=331#=)" %#
6" 6" B# B# / 5 !" ))# 5 ))# #= B#/
%#) B#" %% 5(,=33;-)%# )$(#! ) %##% !)$$$$# &" #" &"")%# 5() %%4 ")#4, =;8J!%#$#$#$ #%$#%$#" #"% B#!## %#. #$"")%%# ("## )#%"# !%#,%P%(($=3
#) $ ""#)) "%# 2 )% )## $ )" )#&"" 5$ %((##$ ").# !,)$#$#)) )% %&"!#) )L ) )%%)## (%&"% 6 ""#.(% )"&".
)"# B"$%(#$ $"(%# # %&" "%#"& !#" "#"E S0"E#&" %#" S0"E#&"% %%# E S5&"#%)"%%# S5&"#%#)"% %# S@&"#%)"% %# S/)&")% )"%%# =?
#" %.)#% %&"%#% #").).) ) "#) )) )$J#. $. )&"$" %#$)#)%#$" )%#) ) "%#" @%#.&$)$ %#.5 /# L";# ) ;# !"5"@T=; #% )"" "%)" %# '# =8
#%"# # "")58 @ "%#*) E 5= 5?)%) 58#" $@" 5; 5>@
50/? KB7 #)) %)#) )$ "(% $)")% # #%($)%) $#$&" $#%%&) # % $ # #,=339)(! %"E %
)!$$ #"* #) ##J($" !% .""& &..# %)% (.), ?9%$ &"#* ) %#"#)"# ),!P%$ =3<<))! )%%! #$"$" "),/$=333# "*"!"&" (#)" $ %($( %( )'J$# "% )% )# =1
))%L )# %) "$%)$ "!'2 B )%( %( #"# 0@) '2,=3F3).")!$$ %("#. 0) $' ,=3F3-#)%) ))! #)#%#" ""$$$$#$"# #'$))L "!#,$=33FB' &".# )$%(!# =F M
%( ""'%",=3<9-%,=3<3# ) '" .)& ) )#%" ( )%(# %% %!'%"J )#&) ")# &" $!.)$!. )$!.#* ''%" %&"J %J!&% #)) 5%)!" "$#% ##%") %(%. )%"# ,%$=3<3N'%"$=3<9)" %) =<
%#)#($ $ B2,=3-( !)"" ))("## #%L%)%$) %) "%( #2%% %)#2 %( &"%(" ,=331) B )% $(%%. )2L /)%( #2)"# ,C*"$=3<>N$=3
#)!)% $$) 5"$ "! !%,=3<<#"#)## )"")%L"2 )"""# % "U2$"#.( B2")" $$($!#"" $!! %$$%($) !%("#)$#$# !%$$ 52% #(!" $$2#% #) 7#"(.! #%( !$$!" 7# #$$%#!"# %" "#
# #*#" J#%""" ")$#"$ )%# &",$=3<<(%( #%#"( .,"P5$=3NC*"P$=33;NC*"P$=3<8N0##" =33;N7P:#$=33=N$=33;NP5%$=33
&"" ") )$($ )"#% $2)"$2"! ,#2"-$2)2 ,/$*$P'$=33F) !) #!#" %$# $ #$##)"$ #%%)# ## )""2% )%###(" !)%,# &)-$# )%,!##&"-$ ""
))%,%&" "%,#*(% $#&" L"# #"$"#% )%&"!( %L ")%%(" "# %(/(%# ##%* )#)% /,=33F##) L ))#$# !$)$! ,/$=33?N0P0%$=33?$
'#$/)! ),$$#) ))%&". !%#) #,0%$$P0$=33;N0%$0$P$=33;)#"% () &$$#J) ) )%%*" $ $&" '"C,=3F;-%"# !)"% %)%$ '"C) !%. #!# .))$## ###""() # )
'"CJ")%) %)! % $ ,5-$#,-$)$)## /)"#! %)% ##$##%%#$.) #(#%%#) ##$'"C! !$"$. ) '"C!. # )%.)$)% $")" # )# #&" &")"## $%)") "#$# &"%($)(#)) %
####% "#$# '"C)%# % !$% %)$"#" %$$#.# "##"!"" % #'"CJ,=3F;-!) ".$!$% ))$%( # /$$,=3F<-) ) #" $/ ,=3F<-)#=# #" )"%###$# %$&"# % )$$")"! "$
.=# %"# (#2#J,1;V) 8=V$=9V#"% $!$=9V !)#" (#% "# #)L$ .)$% ".")% #L& "=# %%$$) )$)%%) J.#$ )$)$L% )(%!#&"% ##!%# '%(/ $ %"% ?F
"%(#))$. $ !P% ,=3<<-%) K%#.%($!% ##!#) &) *%.) %(E%=#% #R) )#"%#) # !%## )" )% ")")#J !#&" !) # @$ $)#" "$" !%,=3<<-%%" $J##) #"#"L" !..)#) "$ "%
$$$!%J L "!# "$"!#J"# $%!%, # -#"" #("%)% "" %%$##$ "#!J %!8#)#%# )%" $ L")L ""%! )L)8# %( "!$ $!"%! %( #.&"J "!#) !)%# !#$$) /#,=33?#)". / 7
#"),)"/,=33?-#$$ J)%$%#%%(# %) %# %"$#$!" )%"" #/J&" #)% #%",'$=3<>/#,=33?##)#( #)$ "#"%. )L ?1;81 #"%% /") )!% ))# 8=.) "") 39 )##% 5 $
""%%ER#"$ 38$%R$ <8$ $ 1;$R$ >8 "%" ,=33;%%. &"J) #)) .R#" 5 %.# "$#%),W >=-$ ,W 1F-$#,W ;F-$%) ,W. >=-".5#,0#% $ =3<1$$)% !.#"=# .%), W >8-$,W 1=@R% "J")$J % )$JJ %(#% #JL ) $
##)% L#"%! 7"5,=33?-!"% #.(# %") !%%"#,") $##$#. %)&$! ) !%"#)$" %### 7"5# #"4 4# )%J&" %# B5,=3 ,#)@ #-=99#"$.##% #F.=8 %!#) $
%#$)$ "" 7"5,=33?-## *# #E $)$##$$)# !"(#%. %)J % %(") J%) (% '"")#) $&")%% &"%) %% !# &"%&"$ JJ$ B#) $##)# "#%%(" #")% 88
)!%)"#) !%#)%#),$ '$PC*"$=339/J,=33;-#%)% )&% %## /J%( %)&) ,-,/$=33?5#,5-,0#% $=3<1-%?1 .;81 (## #8=7($% ""#"" # % "#)# &"#ER#$ %R$$"$ % 39 ##%%E 5R#"$ 38NBR5$ <;N5$ <8N @"$ 1;N#$ >? $' X
"%)18V )# R#" 89 ;V)$%R=; =V)$ .FV)$".8V)$# .=V) />.>9V#* %#) !$"%!# )#%# %###")#"#) F; >V%#" )?> >V%#) # 89V%#) <,8=V-!>=V #) /J"% R#"#) % %R )% ).))).% /) %%## )J 0##) $%
))() ) /$%)$% ("#.(,BP $=3!#" ,/$=33?-$/$#$,=33>-%!J "#L" "$"!% L)%." &" #L"#=#Y# %)R## #5# ,0#% $=3<1-$F9Y#.?#%)" %;=8 )%%!) &"#Y#%L"= #,5.5#,0#% =3<1%!)# Y# $%!%$$) %Y#%L $4
)?#) )##"%#Y# / !(%&" ) Y# "#*%) ,R-! )%%) )Y# B# #$# %(% ) 7"P5,=33?-##J)% !# ) #)% &"(J" 7,=33F-"% &"JL ##%'7,=33F.",!P%$=3<
'"PC$=3F;-"% )L )% "# L #?91(##=1 $!&" %L L%"$ $))##%) )"# ,/$=33?-N ,0'D))$=3<1-##$)$ &)N7@6 A#5,5"P$=33?-##J)! N7(#)7P/$,=3
'7#" %&" L %#"" ) "# &"%()## (%&" #",%))"# "#)%&"# #)) "K(,5FV$@FV-7## (,@8V$5>V5%# #)##% # "%$ $&" #" #)#J ,8V'"!# )(#,8V$8V5 %# 83
(#%%# $%% (#%%) 5## )(#,=FV))##%!%! &" %%%)## %) @" %#"J)$ #;V) 0##"# )" ") ))"#!# )),FV5 )) 0# #") )%#"#)) L%). %)) 5 #) )). 0#) ;9
%(%%)). )??V ).) 5#! ?>V)) 0##")% %)),;V)$8V )) #>?V)# #" #%( '7##" "# ))% "# ###) 5 JL #"%$% )##$%) "$ #*" &" %#"$$##% "# B# "%(#(%)$ $,!P%$=3<
'"PC$=3F;-$#%( #"%J "$J#$)$ ## &""# )$)""# "#"! !##%&" #."%($7" 5,=33F-))#J #% )#)# L,7"P5$=33?7"P5,=33F-.&" ""(# # )"#=$??1#?.<$ .$%(%& %$$$# )"% "!*J) % =F.& ;?
6B5,=3$>FVR& ##?8V#?.> #1.<$?FVR&% !" ####1.< #"$;> 1V##% ;8
) !%" ###%" ). "#. 7"5,=33F-") #" #% (&4)#4%" #%(%" B "%# &"%$%. )%"#(% &"$L$ ))%5? L# %$ "$) &"% "" %) %""" )"J)#"# " M ;;
"% %5=$") %# 5% #))%# %#) )"%"% &*% 5 52!%") )%" "#J)% #%!" )%") 5%!.% ")# 7($ % %$ #)$) 0%)$% $)" #" ##. . ;>
"#) )2#&"#) B(##"$% $!$#") ##% "###$% #")$#$###$((# %L")$#% #"")# #% B )%5?)# &"##$ ))(%&" #) "#(%&" #) # &"% #&" #) (% %) ;1
)%" # )%(###""! %$"##)" #$)"! "#))#(%# #&",/$ =333%(#) ##%% !) B#%$% )() )$)"% "# !# !#%% "# !# ;F
50/8 0@ "!% &"$)$ %#) %"&"2#E = B%# &"" 2#&" ""[ ? %## &"%#![ %.)#% %&"%#% #".).)) ),%#-))% @%%)E) ) )E,=-&". )"$,?-&". )"$,8-%# ;<
" %)% ),"!" 5)E,=-)%# #$,?-)" %# ) %#)#""% %%# #?% )E &' 7)\ KE S.J ,S. ,() +] 57) KE S@) B# S"B# #? @ KE B# ;3
)%" $J %%)" &"#" &"% !# $## ")# &"" $%#% )"&"% #" J.J%#$#%% !%# &" #%!)" %%# ## %!#)"%%# ##%! )"%%# "$# %!)"% %# L#/ "%% LB @# %";FV0$ $ >9 M
?V)$=V$;FV5 # "%# )(%" % #)^=<$999 @#=333.?999"$ )#18V&R # )"#$#" ## J" #=% #"% "%##1.< )" %%=<"! )#%19=?9%# ?" >=
? @# A!/@" B# = <=?9 ? 39 ._ ? 39 ; =39 > ; 19 1 ;19 F ?19 5%L"#(% %. $#$ ##)%#")" #!#(# Q
#& "%%! (# %;#)%#"# % %!) %#"&)%# "#3F )$#;1">=# @ %)%) "&" @%#%) "%#) /)%&" ,!57=><% )# %)%!" (# %% "# !".) #=;;) 7) ")$###?; %! >8
/%#$)$%# "5/#,C%.0$ =338%#%#) 5/ 5/ $$)$%# %%# #"%# %)#5/# # B#/ %%#%."" ),))%-5/ %)%) "%#&" )$)$!"")"#" #$##) %#)%# 5! )!@ B#C B% "$%#)# %# %#)% #*#(% "%# %'
!"% #)%% % $%#)$$) %# 5/#/ B#%)## )5) %#;# 5/# ,5/-%#=331.=33F"#) 8$;$F# "5/%#% 3=#5J ,C%.0$=338# %#)E#*,9.8-$"",9.8 -$###,9.?-$)#,9.=" ##%93 #* )#&"%#% %# )E)$$" $" "$%)$ &"% B#%)"$%# %)#2 # ###%# >>
" %#8>%# 5/ (# %#"5/% (%#" #,#*)%# '#%#%%$ %%! %%# .#B#) %)#%## .#%"*J.J%# %# 8% 8 .#5")B# ## $%&'(( <1>OO "" F1
)%(%## # *$"$ "$"2#..%) "#.(% @)"/,=33?-$ ,-)"" % '%(. "%"! #%"#$ ,-"(% L $#% ,/$=333.# .$$ )%% )#" #" % <3 "# &"%E 39$ 38$ 13 ,'7$=33FN/#$=33?N/$=331N$=33;5)"%"",/$ =33;#7( %# 0# *+ =
*#&" %%&" %")# ,'P7$=33F!'" "" %#%# #/J*%( "%%E <
$##"$") !,/PB$=331!5")$ ) *+ %"#""# *%" )"",W <;"%#" ##% )*+ %%"] %###%a,@$:($' =33<%#$]5". %( a",8,=-$,?#%;# "## #)%#"$ #%#* )*+$ *%## )*+ )&" %%#%. #### #) "% >3
" )*+ % %# )*+ "%### 8>F F1= ) ;99 )*+ % F=,W3F-$# )" K #$) "$%)""! $"%#)$ "%#)" B(# #5=%#$ %)##" %#&") %# )"") #%#%&" )%#%#%#E% !%#)%#) ) #*%)%!) &%# )"!5 !"%#)"; $#
; B#) B!B# D%# /#,%#$#%# B# #!,#*%# /)%# )R/#%# /" 5)%# 'B# 0 # (# # !R/! #%%( # "R"" 5% #) CR @5/ @%#?999.?99=" #$?999 #D"$?99= %# B#%% #)%*5/ ,5)/-%# # 1=
@ J.J%#% %%.%(# #D"$?99= %# %#$% %%( 5%# # ")%%# #"#5/ ))" &"# /#%( ,/$=33?"%% @ )&%) #"% 5%() =>%! &% %)% %% 5 )"##")%#) J)#) @"/ )"%# #/%$K=9 9 @%# $"
"!" ))(#$% =9 @" %). ""%% " &% %""#/5" &%% %# #$##) %).%# "%#% ) .%#%#" 5),5K-%! #!"&""%# 5;"
50/; 7 "%%),%"" @ "%"*#/K=9 9 %#%" )%$& ##"%E 0"E#&"% #" 0"E#&"% %%# E S5&"#%)"%%# S5&"#%#)"% %# S@&"#%)"% %# S/)&")% )"%%# %.)#%"% &"%#%#
".).)) %%) $)%"* / %#)"*#%! &"%%! % / %#JJ.J%# &" # .$% %#&" )") "$% &"%#! B# "*)")%# %&" )"$# "#&" %&"%#! @""# #%##J. $*
%#J 5;% #")) !"&") ")&"%# %#%$#%$ #J &"" % # 0%)$ "#%)# &"%"L##"#) ))&" J"$#) !$% ),/PB$=331#% "###) #%&"%# )")(#) % ))# %)#) $$
&" $!$J)% 2)%&"! #"%%# "$#L!)") $%! 0%)$"# %%% )%%#% %$% # '% "! %%# !) =9% 6" &"%#%. &$) )*+ $ %" $#")&" #%; 0#) )*+ > 1F L
> %6"" ? > =3 ;93= ?= 19F=;?3 3><8 ?9 <>F= ?= ?>99 ?9 <>F= ; <8F9 3 ;>1< 3 >813 =? 9>19 ; F;93 3 3=?=; >3<= ;F ?F?F 83 13>F 83 9999 ;> F>99 81 ;?<1 8< >999 ;1 <>F= 1 =>9; 3 F<=; 1 >>F; > <8?< 1 =88; ? 1;>< > 9=;8 F 8= =9 1= =? 9999 =8 ?>99 == =;?3=9 ?>99 =? F=;8 8 >1;? ; ;8<9 ; >1 8 3?>3 8 =<;< > 31>? > 9<3F ?9 ?9 F9 ?? 99 ?= >< ?9 ;8 ?? ?> ?? =; ? >9 8 89 ? 81 = F3 ? 8F ? ??= 1< @&"!# #8 ?1 ?; ?9 ,$ =1 =; = #=? >=9 #8 )*+ =?8;1F 5 $%
#8)#%)) %J%%"")% &"% %J %#))&" ")%# ##,1 <1W = 983 #W ;=3#.,/$=33?N=331-$ %()&"% ""#%/J# *" "% $$#&" %%> 1<&" ##;$>$1 M 13
19 O> ` 8?9 W =9 --./=?8 >1F 5 #; @56"#"5 ;9 ?9 b 5 #> @56"#"5 F9
89 9GGGGGGGGGGGGGGGGGGGGGGGGGGGGGG=?8 F 5 #1 @@6"#"5 # J&"%% #% *" %% %" "!,<"-# ") %#%&" ,51$
"%&" %#%$# %# 1% %#% 1 6"#.% )*+!! .=?; .?89OO. ?>;OO .0 >F8OOO 993 O5# 9>),?.OOO# 9=),?.&""%# %)#",W ?89$W 9?F(% &"#" #"%#$%""# 5 &"#"#)"% &",W. ?>;$W 9=8# &"%&"#"#),W >F8$W 999*" F?
")% "$ &"%&"# "%# /#% #%% 5 #%## % F% F 6"#.5 )*+!! .=3= .?>>OO. 8<=OOO 9>; >??OOO 9F; OO# 9>),?.OOO# 9=),?."#% &"#)#,W ?>>$ W 9=;5&"#)"%# &""#) # &"%)*" 5 "%#%# F8
&",W. 8<=$W 999%"#"# &",W >??$W 9996"B#/ 0"%# &"%)"%%# %""$" (%&"%#% < %"%&" %##% .%##"% %#,W >;1$W 999-$"% !##(.%#% %# F;
< 5"6"B#/ ,5/B#/!"# $ %$ &%"# '( !"# $ %$ &%"# ( ) 998 99> 993 9=F ) 91F. 9;? =9F. 9<> ) =;F. 981 =>?. 9<= ) =3>O =38O ?<;OO ?9FO O# =9,?.c2# 9>,?.B!$#$% &"%#$&")%" %# ## &"%#"%%## % F>
"%% &".%#!) %# 5#J. %#J#)!#% &"&"$#"" %#E 3 #"/B#.%@ K5#? = B# ?33OOO 8C#5/ =81OOO 56".% 98 O'# =9,?.** '# 9>,?.OOO'# 9=,?.F1
M =9 #"/B#.5 @ K5#? = B# ? 8C#5/ 56".5 56".5 @6".5 O'# =9,?.OO'# 9>,?.OOO'# 9=,?."%%#!!" .)%#% ,?R=9=V-",=R=9=V-&"!)" )%#% $%#$!"; 8V )%#%,W =9?33OOO =81OOO 9<=OO 9?8 9=? FF
6" B#/$5 "%#""%)" ")##)%%! )%#%#), W. F=1$W 9F=)%# %")%#) "%#)""%%&" #" B!%#),?58 !-#&"%#)" )" %#% &"%!%# #%$$ "%%!%#)%# %#)%)% &" 5%)%# &"%"## %#" ==" F<
== B#$$6" +* .+*/ )+ 0. ;>> 1>3 ;F3 9F8 11; >1; >?? ??9 =<; )*+ 8;9 =?9 93> %L%#% %#)%!" ))",%!$W. ?>1$#W ><9N$W. 919$#W <33-% %#"" (%)%# %#%%%(,W ;38$#W ?1=%#"", # W >3<$#W F8=')")%#% $"%" %# "%%# &"!% F3
%#)".&" =?%%%# &"#%#" .%# &" #)"%%# 0## &"%% )%# 5&" #"%#* &"#)"% %#% =? B#%56",% .! ! !123 99< ?31OOO 9>; 14 =8< =;1 9>? 56 =?3. 91> 983 OOO# 9=,?.%#
=8 B#%56",5 .!! !123 9;> $"##(( 9F> 14 =;1 ?=?O 913 56 9F 9>1 9;< O# =9,?.OO# 9>,?.=8"%# &"##". 0#)%# %%%# %%&" .%# &" #"%%%#E #*"" % &""%# 5> <=
50/> @5 ")#(% &"%# %.)#% %&" %#%#".) .)) @#" "%"!# $$) 7" 5>%## "! "0" '%" &"%#%&"#" )#" &"#" "%&"# 5##"%# %"
! %)%)("##) % ) !"# % &"#"#"% !#) B"%)"#)$ %"# !;# #"))#% # $%)" "%$;# )")) ))"" %#"" "&") # "") %"#&" ) !(%" &"# % %)" )$"%")" ##%)) 0%)$ <8
*"" %%" """ #)%( #"" J($( # )))%"## J$"#$ "#J# "%" %"(%&"! %%# 0"E#&"% %%# E S5&"#%)"%%# S5&"#%#)"% %# S@&"#%)"% %# S/)&")% )"%%# 5" 0"%"# "% #%&"%# %)% <;
)("##) 7(%$4#4 "("%$# )## $$#$ #%(&" %)%("" )( B"))"% &"%#$#) #%0" / ##(%&"%# %" % %"#$%#% 7%%#%! # *" !$#"$# % !%"!&" %(%"&"%( )#### % !% <>
%%#" "*#%% ) '##"$ %)$)!"%(" #" #"0"% $$&"! )")%# 5 &"#"!$"# )%#,; 8V) &" )! !(% &"%# !%#"## &) &" $()" )")%(%%# *)&". %#!%)#") %" (%&" %# %$
#)"#&"") #% "%# .&"" (%"##"% *"*"") )") ") (%&"%# /. %#%%) %# "")# )%# "$#*%# %""%#%"$ "%&" /B,=331-## J)%&" # E ##$ )#$" #$J!% )" #&"" !)")%# @"
L "")J )%$##% "(# )"$### ( 5%) %#"#)" 0%)$% !$% "%")!"&" %% '""")%($ # ##%/) 5)## C# "%%" /,=33?-$#) &" 5%&" "#%"*# &" %%
5#%#"% %# #'7,=33F-! !)( &" #))&"% )% @)## #&" !)%# 5&" "!; 8V)#%%# "%&" 7,=33F-&"!) )##(#) )% "'7,=33F-!" ""*%L%' 7J" B!# &"$"% "'7 )!"&" B# "!)%## )$% &"%# %"&") <3
)#&)# ))% 5# & &"%&" /)&" ##"%&" "#%# &"# &" B## &"%#"&"$ ))!&" ) #) )$&"% (% &"#" !# #! "%)"% &"%# 39
&)%#(% &" #% & 5&"%% #""%# B)) #)"!% ),'"PC$=3F;-$ "&"!# (%&"%#" # "&%#) &""""%# # #" )%% %# #$&" ',=3F<-% %&" ) "$ %!)%(" 3=
# %) %*")" !#$#'J#$ %) $ "%,!P%$=3<
%&"% "$" #(%&" %#L% &" 7 @#K" )")# "#)% 5 #""") %) #"$" (#&"% !K"C*) / "#*"L ,/$=33?-%#% "" #*)"" B"" #)&"(% $"%## ") 38 Q
"/% %("% )% %# #"* #*%) #)&"%#%) % )$*" )#"" ")% "")" J ") )*+ ""% B %)!)" )*+ (J) &"% ) ")" %# "!%%" $) "")%) "$ 7'
%)(%##&"I #&" @)"$ #)%$#Z &""% K%#%" "$# %#%%# 77@ # #E$ $" B"% &"%)## $" )( "# &"%"#% )#J 6)$%!$ ###% 5"# #% B +,
)&"[0% %%[B%&) #)% %[(%# "#"# #)) "$##" ""% ) #%" )#% % &" $"% %! &"$$ "&" ")")) 31
. 1. 23 E 5 2= (##GGGGGGGGGGG" A ? A 8 A ; A > A 1 "(#%( A F (" A < (%" A 3F
!' EGGGGGGGGGGGGGGGGGGGGGGEGGGGGGGGGGGGGGGGGGGGG/#%%#" "% #%$ @""$ @" % =?8; > = $%% = ? 8; > ? %"###% = ?8 ; > 8 $%( =? 8; > ; %" = ? 8; > > )R% = ? 8; > 1 %R = ?8 ; > F B$R% = ? 8; > < #" = ?8 ; > 3 "R = ? 8; > =9 )" = ? 8 ; > == "#"% = ?8 ; > =? = ?8; > =8 R" = ? 8; > =; ("%R" = ? 8 ; > => "%%# =? 8; > =1 = ?8; > 3<
=F !L"%% =?8; =< #"# =?8; =3 B)#$R%"() =?8; ?9 @#%"#" =?8; ?= 2)"#")%"## =?8; ?? B$(%%2#" =?8; ?8 2#%#" =?8; ?; @"$2%%## =?8; ?> %%R%# =?8; ?1 (")% =?8; ?F "R#!% =?8; ?< "%() =?8; > > > > > > > > > > > > 33
. 1 432 J" D $%"### % "" = 0%"")"[ ? B"%#", %("-[5(""E D%# # /#,%$-#* %# 0 B# # ) !R! #%%( # !,-B77,)%R#"R"" )%# % #) #R (# 8 )#"%# =99
!@ B#/ =/"%(#%% B"%")"% ?A( ")" %"%"#$ B "[B%"") =9=
! 57 GGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGE 0 "DC 5) /"L% )%# #%. %(%# %%(## 7("$$ %(#")"5@) # ".% ##. #%"?99=$" @=;## )%(% !"% )# 2#%# %#&"J #"%%%# %)"% %#)% "5/%%# )",#)%-"(=9 %%"$#"2 ")$$ )" A2 "%(" B#$)"$ "%"" )"($%$ "$("#%,-)) B)""#"% ""#"$) "")#)! # )"%(%($" =9?
!%"## $$"$$2% )"$%($)% (")%%"%" ) /%"2" ")" "#)%%" ""$"%" ") ""$ %"#"$#$ ")&"#"2 #L$E$5.@) '#$F99$,898->>1.?>>9 %(""" %)")#. ""2#% / <>8.>383%"& ,#5
57 @;CE 0 "DC 5)" %()""% (%. ) "L)##% )%# #% %# %%(## )". ")" 5@) "% #. # %"?99=$"@=; ## )%(% "!"% )#2#% # ""$%&" #%##E=-## $?-#%%#5/ #$8-#)" %)" %"$;-#)"&"" %"" ""$%%%# %)"%" %#)% )"%%5/ )",#)%-"(=9$" %2( )"(=> %2"?99= "#) *"% %)$ )"(!"=9 ,=>.?9 -"%%")% "$%!$ B#$%"%(>9 $##%) =9;
"$)"" ""?99=$ #%)^?>#'(" '( B)""#"%" "$)"" )#)! #)" %(%($"!%" ## A$$ "##)""%" %%" $$ "$$%()"()"$ %($")") B ""*"$ "%%")22" )"!"$ () A%" "%( B#$ )"$)"%"" R#" )"($%$"$( "#%,-))"" 2 /%")" "# $"%"" ) ""$ %"#"$#$ )&"#"2# L$E$5.@)'#$F99$ ,898->>1.?>>9 %("""" %)")#. ""2#% / <>8.>383%"& =9>
M 5/# 8 ? = 9 5P%#E %#E%#E %#E #* S S S %S. & J J ) & & #*S "" S "" S )" # S S /)" S ) ) S /) S @ )" ) ) # S )# ) N S # #* # S '# $ # "# $ % N S N % % S /) % % S @ %% "P" %#E %#E %#E %#E ,) S %# S % S %S % )" S S S #"# %N) )" %N S )") S )" S ) )" S S #) % ) =91
## (# S)# #" ## ## R S& (# ## S S$" # S S S )" ) ) )" ## ## $ $ ## # # $ S"" # S $"" S 7### 7##(7##( 7##( "#E #N #N) % "$ %#E%#E SLR) SLR) # # S S S/E S/E & & ( ( S) S) S# S# ## ## ) ) ) /# # # =9F
%$ @ ,=3<3"" 7+8 ;;,;-$ F93.F=1 !$Y 7 $P%$@ ,=3<<)%" E/ )++ + !+ $>8,?$ ?=<"$ $P5$' -,=3,=-$1=.F3 '%"$D ,=3<97+$ %A(E''( '$ ,=3<>"E B, -$ =;+>*++$ )++? +!+8 ) >9$ =.?, ?935#E)"5#/ '$ ,=3F3*++>/: $ 5#$E0))"/ '"$D $PC$ 7 ,=3F;*++>! @$ %A(E0 5"$D $P$ ,=33?5+$ !+8 18$8>9.81> 5($7 ,=33;*++;$ 0E/$0 =9<
5))$ 7 $B"$/ $B($B 5 $P/($C ,=3395/LE))%""# ++ $?$=38.?=? @$' $P7"$ ,=33< -(E "#" ++ ;8 ?F$8;<.818 $D Y $PK$C ,=37$ B#$@ 5EB/ C*"$ $P$ ,=3<888+$ % A(EC%.0 C*"$ ,=3<>.E) D ), -$ +++8 ,;-$ ?=8.?88 !E/# C*"$ $P$ ,=33;5) $ "$ 7 0), !+7+$ 7E'(% 0##"$ D $$7 $C*"$ $P$ ,=33;898 +$ %A(E5#)" / 0$5 $P0%$5 ,=33?"#J % 5 /, -$%@5 @)EK >F +:;+ ;=.19ED".' 0%$5 $0$5 $P$5 5 ,=33;52% E@%. !+8 1>$?>8.?18 0%$5 $5 5 $P0$5 ,=33;$$Z "J% !+ 8 1>$?1;.?F8 =93
0'D)) ,=3<1) *>52$ @#E5E 0#%$ @ $B($B 5 $5%$ 7 $7"*%($' $$ $C$ D $P($5 ,=3<1.5#E L)"J) ++ ; $=>$838.;93 7$C $P/$D ,=3*+ 2+ 7!+A5$ ED".' 7"$ $P5$@ ,=33F5J%E !"L# <+ + $8>$<=.=99 $ ,=33;))E ()" 5 B#P B C, -$ 7>!+8 8.?80$DE7% $ P5%$D @ ,=33<) )))E7 7+ $>8$?9>.??9 /$ $$ $P$B ,=3F<%) .#J& B ;8 ;<,=-$=.8= /$ 5 ,=33?*+*+ +$ $ )"K#$5)$K
/$ 5 ,=33?%@5@)EK >F 0+ >*+++A$ ED". /$ 5 ,=33;/%(# <++ $8?.=>.8? /$ 5 ,=33F."# 8 % $==.?1 /$ 5 $*$ 7 $P'$ ,=33F.$. L(# !++ +-8 =?$?18.?<9 /$ 5 $P#$ ,=33?% (#J) 5 /, -$ 0+>*+++A 1=.<9ED".' /$ 5 $#$ $P$Y ,=33>%"E .2 L ++8 F$?3F.8=? /$ 5 ,=33?5# %. 5 / -$ 0+>*+ 2+7!+A5$ ED".' /$ 5 ,=33F."# 8 % $=$==.?1 /$ 5 ,=333+ +0;!+*+$ B#$@ 5 E/"# /$ 5 $P#$ ,=33?. L# 5 /, -$ 0+>*+ 2+7!+A5$ ED".' /$5 $PB$@ D ,=331B+9++>! +$ %A(E#
$D $P'%$C ,=33FE#.. )) -$ 9 ++>7 3>.==1B#$@ 5 E B/ #$' ,=3397++9>! /$ %A(E!)"/ $ ,=3F,8-$8=1.88= $ B ,=33;7+++ *+ +$ $)" K#$5) $ $$ $'%$5 $P'%$ ,=33?5" (%;=8"# <+7 7!+7+18 8=$<38.398 $7 ,=331>*+3 +$ %A(E5#)"/ $7 ,=3<7 ++$ %A(EC%0 B$ $P$ ,=33?2+>B+9+ :++$ $ A E5)"