Content uploaded by Lia Elaine Sandilos
Author content
All content in this area was uploaded by Lia Elaine Sandilos on Jan 11, 2018
Content may be subject to copyright.
Warmth and Demand: The Relation Between Students’Perceptions of the
Classroom Environment and Achievement Growth
Lia E. Sandilos, Sara E. Rimm-Kaufman, and Julia J. Cohen
University of Virginia
Theory suggests that African American students benefit from warm and demanding teachers. This study
examines the relation between students’perceptions of 634 teachers’warm demander characteristics and
achievement growth in fourth and fifth grades (M
student age
=9–11.5 years). Analyses explored whether rela-
tions were moderated by the proportion of African American students in the classroom or the ethnic match or
mismatch between African American students and their teachers. Results indicated that students’perceptions
of teachers’demand (challenge and control) related to student achievement growth. Findings showed a stron-
ger relation between challenge and academic growth in classrooms with more African American students, but
no significant findings were identified for ethnic match or mismatch.
In U.S. schools, teachers are charged with the task
of establishing learning environments that foster
measurable academic growth in diverse learners
(Zaslow, Martinez-Beck, Tout, & Halle, 2011). As a
result, education researchers strive to develop a
deeper understanding of the instructional practices
and interactions that are most beneficial to students.
In particular, the persistent disparities in achieve-
ment that exist between minority and nonminority
students has prompted research examining cultur-
ally responsive pedagogical practices that support
ethnic minority children (Villegas & Lucas, 2002).
Classroom processes and teacher–child interactions
hold promise for reducing these disparities (Pianta
& Walsh, 1996). Yet, prior research has indicated
that the quality of instruction and teacher–child
interactions is lower in low-income elementary
schools, which have disproportionately higher num-
bers of ethnic minority students (National Center
for Education and Evaluation, 2011; Pianta, Belsky,
Houts, & Morrison, 2007). Given these patterns, it
is essential that research continues to explore the
social interactions and instructional patterns that
can support positive developmental outcomes for
ethnic minority students (Cabrera, Beeghly, &
Eisenberg, 2012).
Warm demander pedagogy is one example of a
culturally responsive teaching framework that
describes effective interactions and instructional
practices with African American students. The
warm demander theory was developed from small-
scale qualitative work with samples of both African
American students and teachers, and posits that
teachers who are high in both warmth and demand
toward their students produce the best outcomes
for African American students (Ford & Sassi, 2014;
Ware, 2006). In theory, teachers who embody a
high-warmth and high-demand teaching style pro-
mote note only a culture of academic achievement
by managing their classrooms very efficiently but
also conveying to students that they care for their
well-being and hold high expectations for academic
achievement (Ware, 2006). However, since the
emergence of this theory from qualitative examina-
tions of effective teachers, little empirical work has
been done to test the relation between warm
demander teachers and student achievement out-
comes.
The present study addresses a problem at the
intersection of two lines of research. One line of
research measures and examines effective classroom
instruction, by considering interactions between
teachers and students (Pianta & Hamre, 2009). The
second line of research examines warm demander
The data used in this study were collected as part of the Mea-
sures of Effective Teaching (MET) Project funded by the Bill &
Melinda Gates Foundation (2009–2011). The research reported
here was supported by the Institute of Education Sciences, U.S.
Department of Education, through Grant R305B130013 to the
University of Virginia. The opinions expressed are those of the
authors and do not represent views of the U.S. Department of
Education.
Correspondence concerning this article should be addressed to
Lia E. Sandilos, Curry School of Education, University of Vir-
ginia, 268 Ruffner Hall 405 Emmet Street, Charlottesville, VA
22904. Electronic mail may be sent to lia.sandilos@gmail.com.
©2016 The Authors
Child Development ©2016 Society for Research in Child Development, Inc.
All rights reserved. 0009-3920/2016/xxxx-xxxx
DOI: 10.1111/cdev.12685
Child Development, xxxx 2016, Volume 00, Number 0, Pages 1–17
teaching practices that benefit ethnic minority stu-
dents, specifically African American students
(Ware, 2006). Bridging these two lines of research
has potential to create a deeper understanding of
the classroom processes that are most beneficial
and culturally responsive to students. To that end,
this study explored the extent to which the warm
demander teaching style relates to fourth- and fifth-
grade student achievement growth in reading and
math and considered the extent to which these
practices are particularly important for African
American students. Furthermore, this study
explored the influence of teacher ethnicity on the
relation between warm demander practices and
African American students’achievement growth.
Teachers as Warm Demanders
The warm demander theory emerges from a
broader literature on culturally responsive peda-
gogy, which explores effective teaching through a
deeper understanding of learning styles and interac-
tion patterns of students from diverse backgrounds
(racial, ethnic, economic, linguistic; Gay, 2002). The
term warm demander, coined by Vasquez (1989),
refers to a teaching style in which teachers are nur-
turing or caring toward their students but do not
lower academic standards or expectations and are
effective disciplinarians. Ethnographic work on
warm demanders focuses on effective instruction
with African American students and emphasizes
teachers’critical role in promoting educational
equity by holding all students to a very high stan-
dard academically while maintaining strong caring
relationships and effective classroom management
(Ford & Sassi, 2014). Thus, warmth is defined as
the teacher’s ability to exhibit unconditional posi-
tive regard, convey a sense of caring for students’
well-being, show authentic interest in children’s
lives, and demonstrate mutual respect (Bondy &
Ross, 2008). Demand refers to the teacher’s ability
to challenge students through academically rigor-
ous instruction and to insist that students exert
effort and perform to a high standard. Demand also
encompasses the teacher’s ability to manage or con-
trol classroom behavior to minimize distractions
and increase time engaged in learning (Bondy &
Ross, 2008; Ware, 2006).
Warm demander practices have been found to
be especially salient for African American students.
African American high school students reported
stronger trust in their teachers when they perceived
teachers as both caring and holding high expecta-
tions (Gregory & Weinstein, 2008). Additionally,
academically successful African American high
school graduates from single-parent, low-income
homes reported having supportive and warm men-
tors who challenged them to succeed (Williams &
Bryan, 2013). African American students have been
historically marginalized within the education sys-
tem, and teachers may be able to use warm deman-
der practices to counteract negative societal
messages and to convey high expectations for suc-
cess in school.
Moreover, several qualitative studies have sug-
gested that African American teachers might better
exhibit warm demander teaching style than Cau-
casian teachers because they draw upon a shared
cultural history with African American students
(e.g., Cholewa, Goodman, West-Olatunji, & Ama-
tea, 2014; Ford & Sassi, 2014; Irvine, 2002; Ware,
2006). A shared ethnic or cultural background can
contribute to positive student–teacher relationships
because students have the opportunity to view their
teacher as a role model who is racially similar to
themselves (Dee, 2005). Furthermore, some theorize
that African American students are able to perceive
cultural commonalities between themselves and
their teacher, contributing to a lower risk for nega-
tive expectancy effects and stereotype threat when
an ethnic match exists (Dee, 2005). Teachers’per-
ceptions of students may also be influenced by eth-
nic match or mismatch. For example, prior studies
of elementary and middle school samples indicated
that Caucasian teachers were more likely to report
a higher prevalence of problems among ethnic
minority students (e.g., inattention, lack of home-
work completion, difficulty following directions,
disorganized home environment) than teachers of
similar ethnic minority backgrounds (Dee, 2005;
Rimm-Kaufman, Pianta, & Cox, 2000). Given that
warm demander theory originated from observa-
tions of African American teachers and students,
ethnic match may be an important consideration in
understanding the practices of warm demanding
teachers.
Important qualitative work generated the warm
demander theory based on observations of African
American students and teachers. An essential next
step for the warm demander theory is to better
understand the influence of these practices on stu-
dent achievement outcomes in a large and diverse
sample of students. Validation of this pedagogical
style is important to improve teacher preparation
and professional development. Current views of
cultural competence emphasize that teachers hold
high expectations for students regardless of stu-
dents’background characteristics (Ladson-Billings,
2 Sandilos, Rimm-Kaufman, and Cohen
2009). Enhanced empirical understanding of the
warm demander style may provide important
nuance for future recommendations. Specifically,
establishing links between warm demander teach-
ing style and student academic growth may lead to
interventions to reduce racial academic disparities.
Unfortunately, there is no quantitative measure
of warm demandingness. However, the widely
used Tripod Survey (Ferguson, 2008) is a valid tool
that can be used reliably to operationalize key
facets of the theory. The Tripod uses student-report
data to measure teachers’level of academic chal-
lenge, classroom control, and care—all aspects of
the warm demander teaching style.
Considering the Student Perspective
Students’perception of their experiences in the
classroom provides a unique lens for measuring
classroom quality. Two students sitting in the same
classroom may have very different experiences. Stu-
dents’experiences vary because each student con-
stantly appraises his or her immediate context or
environment (e.g., classrooms, teachers, peers).
These subjective appraisals lead students to either
engage in or avoid the available learning opportuni-
ties (Skinner, Kindermann, Connell, & Wellborn,
2009). When measures are designed well and imple-
mented properly, surveying students about their per-
ceptions of the classroom environment can provide
critical information about how students perceive
their teachers’ability to meet their needs (Brock,
Nishida, Chiong, Grimm, & Rimm-Kaufman, 2008;
Kane & Staiger, 2012). Surveying students is more
common in secondary and higher education; how-
ever, the use of student perception measures has
gained credibility as a tool for understanding upper
elementary classrooms as well (Downer, Stuhlman,
Schweig, Mart
ınez, & Ruzek, 2014; Polikoff, 2014).
Students’views of classrooms reflect the actual
characteristics of the classroom experience, as well
as students’expectations and past experiences. The
student perspective yields valuable information
about pedagogical practices and facilitates under-
standing of the diversity of views students have
about their schooling (Howard, 2001). In the pre-
sent study, the Tripod was used to measure three
elements of warm demander teaching: challenge,
control, and care (Ferguson, 2010).
Challenge
Challenge, one aspect of demand, refers to the
rigor of instructional content as well as the amount
of effort students are expected to expend on learn-
ing (Ferguson, 2010). Related terms include teacher
expectations,academic press,ordemandingness (God-
dard, Sweetland, & Hoy, 2000; Lee, 2012; Lee &
Smith, 1999). One common element among these
various constructs is teachers’facilitation of aca-
demically rigorous learning environments in which
students are expected to exert the effort needed for
academic success and to persist in the face of diffi-
culty (Ferguson, 2010). When teachers challenge
their students successfully, the teachers’expectation
for thoughtful work and high effort will convey to
students they are all competent and capable of per-
forming to a high standard (Fulmer & Turner, 2014;
Stipek, 2002). Prior research indicates that teachers’
expression of high expectations and facilitation of a
challenging environment influence academic perfor-
mance directly (Brophy & Good, 1970; Goddard
et al., 2000; Rosenthal & Jacobsen, 1968). For exam-
ple, elementary and middle school-age students
whose teachers held high expectations for the entire
class made twice as much reading growth in a year
compared to students of similar ability in class-
rooms with low-expectation teachers (Rubie-Davies,
2007). In a large sample of urban elementary
schools with more than 50% of the student sample
being low income and African American, teachers’
self-reported academic press accounted for approxi-
mately half of the variance in reading and math
achievement between schools (Goddard et al.,
2000).
Students’perceptions of teachers’performance
expectations may be integral to the relation between
academic challenge and outcomes. Good (1981) out-
lined a model of teacher expectations and student
perceptions: (a) the teacher expects certain student
behaviors or academic performance, (b) the teacher
behaves in specific ways toward students based on
those expectations, (c) teacher behavior alters stu-
dents’academic self-concept, (d) over time students
conform to the behaviors that are expected of them.
Thus, when teachers demonstrate class-wide expec-
tations for academic excellence, students will be
more likely to exhibit behavior consistent with
those expectations.
Existing research also suggests that the associa-
tion between teacher expectations and student
achievement may be stronger for African American
students than for their Caucasian peers (Jussim,
Eccles, & Madon, 1996). By fourth grade, students
have the ability to recognize that teachers can hold
differential beliefs about their academic ability and
that those expectations may relate to students’
group membership (McKown & Weinstein, 2003).
Warmth and Demand 3
For example, one study examined elementary stu-
dents’academic trajectories to determine if their
present teacher had lower expectations for perfor-
mance than was reasonable given their previous
academic track record. Results indicated that teach-
ers who held lower expectations, or underestimated
student ability, had a more deleterious effect on the
academic performance of African American stu-
dents than their Caucasian peers (McKown & Wein-
stein, 2002). As the Good (1981) model posited,
expectations can have a powerful influence on stu-
dent performance. The findings from McKown and
Weinstein (2002) underscore the ways in which
high expectations may be especially supportive for
African American students. In addition to chal-
lenge, it is important to understand how classroom
control, the second aspect of demand in warm
demander teaching, influences outcomes.
Control
Teachers who demonstrate effective management
or control of the classroom establish clear routines
and have higher levels of student productivity
(Pianta & Hamre, 2009). Classroom control is char-
acterized by teachers’ability to redirect off-task or
disruptive behavior in ways that do not upset stu-
dents’focus and the flow of a lesson (Ferguson,
2010). Generally, teachers who take a proactive
approach to management (i.e., use clear guidance
about behavioral expectations, take time to establish
routines and classroom rules) tend to have more
well-managed classrooms across grade levels and
school contexts (Woolfolk Hoy & Weinstein, 2011).
Positive management strategies can help to improve
students’ability to regulate their own behavior
(Rimm-Kaufman, Curby, Grimm, Nathanson, &
Brock, 2009), which, in turn, supports academic
achievement (e.g., Ponitz, Rimm-Kaufman, Grimm,
& Curby, 2009). Students’perception of teachers’
classroom management is critical to examine with
ethnic minority student populations given that
research has revealed inequitable disciplinary out-
comes for African American students, and for male
students in particular (Gregory & Weinstein, 2008;
Silva, Langhout, Kohfeldt, & Gurrola, 2015). In light
of these disparities, it is valuable to gain deeper
insight into African American students’perceptions
of classroom management and how those percep-
tions relate to outcomes. However, limited research
has been done in this area.
Most research examines teachers’perceptions of
their own classroom management (e.g., Clunies-
Ross, Little, & Kienhuis, 2008; Mitchell, Bradshaw,
& Leaf, 2010) or observed indicators of high-quality
classroom management strategies (e.g., Pianta &
Hamre, 2009; Rimm-Kaufman et al., 2009). A few
studies have explored students’perceptions of
classroom management and their relation to aca-
demic outcomes. In one qualitative study of middle
and high school students, aspects of classroom
management that were most salient to students
included their teachers’ability to clearly convey
rules, expectations, and consequences for misbehav-
ior. The degree to which the teacher was caring also
emerged, warranting further discussion below
(Cothran, Kulinna, & Garrahy, 2003).
Care
Caring teachers, by definition, establish a class-
room climate in which students feel emotionally
safe and sense that their teacher is concerned for
their well-being and future success. Caring teachers
not only demonstrate positive interpersonal interac-
tions with students, but they also foster an environ-
ment in which students feel they can take academic
risks (Ferguson, 2010; Reyes, Brackett, Rivers,
White, & Salovey, 2012; Rimm-Kaufman & Chiu,
2007).
Caring teachers may be particularly beneficial to
children who experience social inequities in the
education system (e.g., low income background,
membership in a minority group). Howard (2001)
found that African American elementary students
preferred learning environments in which they per-
ceived their teacher as caring. Roorda, Koomen,
Spilt, and Oort (2011) found that the magnitude of
the relation between teacher–child relationships and
achievement in a classroom increased as the num-
ber of students from low-socioeconomic status and
ethnic minority backgrounds increased. Students’
perceptions of teachers’care has also been linked to
engagement and motivation in elementary and sec-
ondary grades (e.g., Klem & Connell, 2004; Went-
zel, 1997). Less is known about the direct relation
between students’perceptions of teachers’caring
behaviors, or warmth, and students’academic
growth in math and reading in upper elementary
grades, particularly when examined simultaneously
with teachers’level of demand (i.e., challenge and
control).
Developmental Significance
The fourth and fifth grades, which span middle
childhood to early adolescence, are important years
in which to examine perceptions of the learning
4 Sandilos, Rimm-Kaufman, and Cohen
environment. During this phase of development,
students transition physically, emotionally, and psy-
chologically into adolescence and also prepare for
an environmental transition from elementary to
middle school. Children’s sense of identity and
need for autonomy advance, and social relation-
ships and interactions with peers and adults
change, as well (Eccles, 1999). As children forge
their individuality, race and culture often come to
the forefront of their experiences and they begin to
understand their own cultural background within a
broader social context (McKown & Weinstein,
2003).
Children’s academic self-concept is also forming
rapidly during this stage and their interactions
with teachers and experiences of success or failure
within the school system have greater leverage in
shaping their long-term identity as a student
(Eccles, 1999). For some children, ethnic identity
may be signaling one message of what is expected
of them in school while educators may be convey-
ing a different message, particularly through the
expression of academic expectations. As such,
teachers’warm demander style has the potential
to be vital in helping students at this developmen-
tal stage establish a positive academic identity
and, in turn, support greater achievement. Taken
together, this sensitive phase of development
includes a complex interplay of personal develop-
ment, ethnic background, and classroom experi-
ences.
Goals of the Current Study
The present study tests two propositions of the
warm demander theory: (a) that warm and
demanding interactions relate to positive develop-
mental outcomes for fourth- and fifth-grade African
American students, and (b) that African American
teachers may be particularly effective warm deman-
ders (Irvine, 2002). We test the warm demander
theory by examining the combination of challenge,
control, and care and its relation to achievement
growth on both high- and low-stakes math and
reading assessments. In doing so, we rely on the
students’perception of challenge, control, and care
because of the unique perspective they bring to
understanding day-to-day classroom life. We pur-
sued three research questions: (a) Do students’per-
ceptions of teachers as warm demanders, as
measured by challenge, control, and care relate to
student achievement growth? (b) Is the relation
between students’perceptions of teachers’warmth
and demand and student achievement growth
moderated by the proportion of African American
students in the classroom? (c) Is the relation
between students’perceptions of teachers’warmth
and demand and student achievement growth in a
subsample of classrooms with predominantly Afri-
can American students (>80%) moderated by the
ethnic match or mismatch between an African
American teacher and students?
We hypothesized that students’perceptions of
their teachers as warm demanders would relate
positively to student achievement growth in fourth-
and fifth-grade elementary classrooms. Consistent
with the warm demander theory, we hypothesized
that the proportion of African American students in
the classroom would moderate this relation, such
that the warm demander characteristics would be
more important for African American students’aca-
demic growth than non-African American students.
We also hypothesized that the presence of an Afri-
can American teacher (ethnic match) would
strengthen the relation between warm demander
characteristics and academic growth in classrooms
with predominantly African American students. We
expected ethnic mismatch would attenuate this rela-
tion.
Method
Participants
This study used data from the 1st year of the 2-
year Measures of Effective Teaching (MET) Project,
a large-scale observational study of classroom
teaching conducted by the Bill & Melinda Gates
Foundation (2009–2011). The primary purpose of
MET study was to evaluate existing measures of
teaching quality and instructional effectiveness
(Kane & Staiger, 2012).
Full Sample
Participants in the present study included a total
of 634 fourth- (n=320) and fifth-grade (n=314)
teachers from five large districts located in New
York, North Carolina, Florida, Tennessee, and Col-
orado. Teachers included in the sample were con-
sidered generalists because they taught all major
subjects to their classroom students. The majority of
teachers were female, and a little more than half
had a master’s degree or higher, with a wide range
of experience teaching in their district (M=6 years,
range =<1–34). The ethnic composition of teachers
was largely African American and Caucasian
(Table 2).
Warmth and Demand 5
African American Subsample
A subsample of 223 fourth- (n=107) and fifth-
grade (n=116) classrooms identified as having pre-
dominantly (80%–100%) African American students
were used to explore the interaction of teacher–stu-
dent ethnic match or mismatch with student per-
ception ratings. For the subsample, the majority of
teachers were female, nearly three-quarters had a
master’s degree or higher with an average of
5 years of experience teaching in their district
(range =1–19 years). The ethnic composition of
teachers was three-quarters African American and
nearly one-quarter Caucasian (Table 2).
Demographic characteristics of the classroom
compositions of the full sample and African Ameri-
can subsample are also included in Table 2. Aver-
age age of students in a classroom ranged from 9 to
11.5 years (M=9.97) in the full sample, and 9.20–
11.5 years (M=10.04) in the African American sub-
sample.
Measures
Tripod 7Cs Student Perceptions Survey
The Tripod 7Cs (Ferguson, 2008) is a 36-item sur-
vey designed to capture students’perspective of the
classroom environment through seven composites:
challenge, control, care, clarify, captivate, confer,
and consolidate. All items are coded as 1 (no, never),
2(mostly not), 3 (maybe/sometimes), 4 (mostly yes),
and 5 (yes, always). Composite scores are calculated
by averaging item-level scores. We focus on chal-
lenge, control, and care composites because of their
alignment with characteristics of warm demander
teachers. Correlations among the three composites
ranged from .44 to .61 (Table 3).
The challenge composite consists of four items
tapping into the rigor of instructional content and
the level of hard work expected of students in the
classroom. Item examples include, “In this class,
my teacher accepts nothing less than our full effort”
and “My teacher pushes everybody to work hard.”
The challenge items demonstrated high internal
consistency (a=.79) and moderate item-level corre-
lations (r=.37–.69). The control composite consists
of four items inquiring about teachers’ability to
maintain a well-behaved and productive classroom.
Item examples include, “Students in this class
behave the way my teacher wants them to”and
“Our class stays busy and doesn’t waste time.”One
item, “Students behave so badly in this class it
slows down our learning,”was reverse scored. The
control composite exhibited high internal consis-
tency (a=.80) and moderate item-level correlations
(r=.35–.64). The care composite consists of seven
items that ask students about their teachers’level of
care and emotional supportiveness in the class-
room. Examples of items include, “I like the way
my teacher treats me when I need help”and “My
teacher in this class makes me feel that he or she
really cares about me.”The care composite had
high internal consistency reliability (a=.92) and
moderate to high correlations among items
(r=.44–.80). A full list of items is included in
Table 1. Individual student scores were aggregated
to create a classroom level score for each of the
challenge, care, and control composites to prepare
for analyses.
Achievement Growth Measured by Value-Added
Estimates
Student achievement growth on low- and high-
stakes reading and math assessments was measured
through value-added models (Kane & Staiger,
2012). MET researchers used student assessment
data and background information to create value-
Table 1
Tripod Items for Challenge, Control, and Care
Items
Challenge
My teacher pushes everybody to work hard.
My teacher pushes us to think hard about things we read.
In this class, my teacher accepts nothing less than our full
effort.
In this class, we have to think hard about the writing we do.
Control
Students in this class behave the way my teacher wants them
to.
Our class stays busy and does not waste time.
Students behave so badly in this class that it slows down our
learning (reversed).
Everybody knows what they should be doing and learning in
this class.
Care
I like the way my teacher treats me when I need help.
My teacher in this class makes me feel that he or she really
cares about me.
The teacher in this class encourages me to do my best.
My teacher gives us time to explain our ideas.
My teacher seems to know if something is bothering me.
If I am sad or angry my teacher helps me feel better.
My teacher is nice to me when I ask questions.
Note. Items are scored on a range of 1 (no, never)to5(yes,
always).
6 Sandilos, Rimm-Kaufman, and Cohen
added scores for teachers. A teacher’s value-added
estimate was computed for each of the four student
assessment measures and standardized across the
entire sample of teachers. Value-added scores for
each assessment were used as key outcomes. The
data included within a value-added estimate con-
sisted of (a) students’current assessment scores in a
particular measure, (b) their scores in a measure of
that content domain from a year prior, (c) the
aggregate classroom score on that content domain
assessment in the students’classroom a year prior,
(d) individual student background variables, and
(e) student background variables averaged across
the classroom (White & Rowan, 2013). Student
background variables included ethnicity, English
language learners (ELL) status, age, gender, special
education status, gifted status, and receipt of free or
reduced lunch. Students’prior year state test scores
were used as an estimate of prior achievement for
value-added scores in the MET-selected assessments
(Stanford 9 Open-Ended Reading Assessment
[SAT-9] and Balanced Assessment in Mathematics
[BAM]) because those measures were administered
exclusively during the MET study. Teachers received
an aggregate value-added score for each of the four
assessment measures. More information regarding
the calculation of MET value-added estimates can be
found in the MET User’s Guide (White & Rowan,
2013).
Value-added scores are interpreted as the teach-
ers’ability to facilitate growth in an academic area,
conditional on the composition of students in a
classroom. A standardized value-added score of
zero would indicate that students are performing as
expected given prior achievement and background
data, a negative score would indicate that students
are performing lower than expected and a positive
score would indicate that students are performing
higher than expected. As with other measures of
educational accountability and effectiveness, there
are cautions surrounding value-added models, such
as the limitations of using a single number to quan-
tify teacher effectiveness and the potential instabil-
ity of scores (Braun, 2005). Despite these cautions,
these models are commonly used in teacher evalua-
tions and have been linked empirically to long-term
student outcomes (e.g., college attendance, job sal-
ary; Chetty, Friedman, & Rockoff, 2011).
The following section describes the four individ-
ual assessments used to create the value-added
scores. MET researchers designated the BAM and
SAT-9 as low-stakes assessments and the state math
and English tests as high-stakes assessments. The
latter are high-stakes assessments because data
from these tests are used to make important deci-
sions about students and teachers (von der Embse,
Kilgus, Solomon, Bowler, & Curtiss, 2015).
Balanced Assessment in Mathematics. The BAM is
an open-ended mathematics assessment that
requires written mathematical responses from stu-
dents. The BAM is designed to test students’con-
ceptual knowledge, higher level reasoning, and
problem-solving skills. The BAM measures stu-
dents’skill at modeling or formulating problems,
transforming and manipulating mathematical for-
mulas, inferring or drawing conclusions from tasks,
and communicating about mathematics findings.
The assessment contains four to five tasks that take
approximately 1 hr to complete.
Stanford Open-Ended Assessment in Reading (SAT-
9). The SAT-9 is a measure of students’reading
comprehension of various types of text (fictional,
nonfictional, and practical material such as newspa-
per stories or advertisements). Students participat-
ing in the MET study were presented with one
narrative reading passage and asked to respond to
nine open-ended comprehension questions. The
SAT-9 was administered within a single class per-
iod.
State standardized tests in mathematics (MATH) and
English language arts. State standardized tests in
MATH and English language arts (ELA) varied by
district. The standardized test formats were typi-
cally multiple choice. Only rank-based zscores for
the state standardized test measures are provided
in the MET data set.
Demographic Data
A demographic survey was used to collect rele-
vant background data on participating teachers.
Local district administrative data were used to
derive aggregate classroom demographic informa-
tion for this study (Kane & Staiger, 2012). Analyses
included the proportion of students identified as
African American students (i.e., selected to test the
warm demander theory), male, receiving special
education, and ELL.
Procedure
Participating districts were recruited through an
opportunity sampling procedure, which resulted in
six large districts across the country agreeing to
participate in the MET study. One district was
excluded from the present study because it did not
have participating fourth- and fifth-grade class-
rooms. Once districts were identified, school
Warmth and Demand 7
principals were recruited and the principals identi-
fied eligible teachers. To be an eligible participant,
the teacher had to be part of a grade level or sub-
ject area that had at least two other teachers to
form an “exchange group.”The requirement of an
exchange group was established so that students
could be randomly assigned to one of three teach-
ers in the 2nd year of the study. Alternative schools
(special education, vocational schools, etc.) were
excluded. Teachers who engaged in team-teaching
situations in which it would be difficult to link stu-
dents to one teacher also were excluded.
Incentives were used for recruitment and reten-
tion. Schools received a total of $1,500, the use of
which could be determined by the schools’admin-
istration, as well as $500 a year to pay for a project
coordinator. Video recording equipment used for
MET classroom observations was donated to the
schools at the conclusion of the study. Each partici-
pating teacher received a $1,500 incentive for partic-
ipating in the study ($1,000 at the start of the study
and $500 at the conclusion).
Data used in the present study were collected
during spring 2010 of the 1st year of the MET
study. No randomization occurred during the 1st
year of the MET study. The BAM and SAT-9 were
administered between April and June 2010. The
state standardized testing occurred between March
and June 2010. Students were administered the Tri-
pod 7Cs survey between February and June 2010.
The measures selected by MET researchers (i.e.,
BAM, SAT-9, and Tripod) were administered to all
consenting students within participating teachers’
classrooms. The MET data set is restricted because
it contains sensitive and potentially identifying
information about school districts. The findings
reported in this article follow the secure data
reporting policies outlined by MET. Additional
information about the MET data set can be found
in MET reports (Kane & Staiger, 2012).
Data Analysis
Regression models were conducted in MPlus ver-
sion 7 (Muth
en & Muth
en, 1998–2012) with all data
aggregated to the classroom level. The average
number of schools included in each analysis ranged
from 104 to 108, with an average cluster size of
approximately six classrooms per school. Intraclass
correlations (ICCs) were examined for the outcomes
and variables of interest to determine the need for
nesting classrooms within schools. ICCs ranged
from .08 to .22 revealing a shared variance among
teachers within the same school. Thus, the
TYPE =COMPLEX function in Mplus was utilized
in each analysis to account for the nonindepen-
dence of observations through the use of robust
standard errors.
Of the 634 generalist classrooms included in the
analysis, 60 teachers had some amount of missing
data. Of the 223 majority African American class-
rooms, 37 teachers had some amount of missing
data. Based on the missing at random (MAR)
assumption (Enders, 2010), all models were esti-
mated using full information maximum likelihood.
Although the MAR assumption could not be
directly tested, it was believed to be reasonable for
these data because no significant differences were
found between participants with missing data and
participants with complete data. One exception to
this was the proportion of students within a class-
room receiving free or reduced price lunch, a proxy
for socioeconomic status, which was not provided
for one entire district. As such, this variable was
included as an auxiliary variable in the analyses to
adjust standard errors for systematically missing
data. Grade level and district fixed effects were
included in the analysis as dummy variables. Fol-
lowing the MET secure data requirements, district
fixed effects estimates were not reported in the
Results section or Tables. Other covariates consisted
of the proportion of students identified as male,
ELL, and special education status. These variables
were included because they demonstrated signifi-
cant correlations with either the Tripod focal pre-
dictors or outcome variables.
To address the warm demander hypothesis, the
percent of African American students per classroom
was included in analyses as a continuous propor-
tion variable. To explore the influence of ethnic
match or mismatch between teachers and students,
a subsample of classrooms with 80% or more Afri-
can American students was examined. If the teacher
in this subsample was also African American, then
that classroom was coded as a dichotomous
teacher–student ethnic match variable (match
1=yes, 0 =no). If the teacher was not African
American, then the classroom was coded as
dichotomous ethnic mismatch variable (mismatch
1=yes, 0 =no).
For each regression model, variables were
entered in a stepwise fashion to first determine if
main effects were present and then, to determine
the presence of significant interaction effects. To
address the first research question, the three focal
Tripod composites were used to predict value-
added scores. Classroom proportion of African
American students was then included as a
8 Sandilos, Rimm-Kaufman, and Cohen
moderator in each model to explore the second
research question. To examine the third research
question, Tripod composites were used to predict
value-added scores in a smaller subsample of class-
rooms with predominantly African American stu-
dents, and match or mismatch variables were
entered as moderators to determine if there was a
significant interaction between Tripod predictors
and teacher–student ethnic match or mismatch.
Results
Raw score means, standard deviations, and ranges
are reported for the Tripod composites and BAM
and SAT-9 value-added outcomes (Table 2). State-
standardized testing data were only available in
standardized form. Correlations between Tripod
composites and value-added outcomes ranged from
.08 to .22 (Table 3). Within each regression model,
all data were zscored to aid with interpretation of
the findings.
To explore the influence of challenge, control,
and care on value-added scores in the full sample,
the main effects of those variables were tested in an
initial set of regression models (Model 1, Table 4).
In the model predicting BAM, the classroom aggre-
gate rating of control related to higher value-added
scores (b=.14, p<.01). For SAT-9, the classroom
aggregate ratings of challenge (b=.11, p<.05) and
control (b=.16, p<.05) related to higher value-
added scores. Similarly for the MATH value-added
model, average student perceptions of challenge
(b=.18, p<.001) and control (b=.15, p<.01)
related to higher value-added scores. In the model
predicting ELA, aggregate classroom ratings of
challenge (b=.18, p<.01) and control (b=.18,
p<.01) related to higher value-added scores.
Notably, when included in models with chal-
lenge and control variables, the main effect of care
was consistently nonsignificant. However, when
care was entered into the model alone, it was sig-
nificantly and positively related to the value-added
outcomes. This finding indicates that the influence
of the care composite was diminished in the pres-
ence of challenge and control composites. Because
the warm demander theory posits that these teach-
ing qualities should be present simultaneously in
the classroom, models with all three Tripod com-
posites entered together were selected as the final
models.
To explore the moderation effect of the propor-
tion African American students in a classroom on
the relation between Tripod ratings and value-
added scores, interaction effects were tested in a sec-
ond set of regression models (Model 2, Table 4). For
BAM, there was a significant positive interaction
between challenge and the proportion of African
American students in a classroom (b=.13, p<.05).
Simple slopes were examined to determine the nat-
ure of the interaction. At high levels (+1SD above
the mean) of proportion of African American stu-
dents, the slope was statistically significant (b=.05,
SE =.02, t=2.47, p=.014) such that the association
between challenge and BAM value-added scores
was stronger in classrooms with higher numbers of
African American students (Figure 1).
Likewise for SAT-9 value-added scores, a signifi-
cant positive interaction was found between chal-
lenge and the proportion of African American
students in a classroom (b=.12, p<.05). An exam-
ination of simple slopes revealed significant slopes
at average (b=.03, SE =.02, t=2.06, p=.040) and
high proportions (b=.07, SE =.03, t=2.50,
p=.013) of African American students (Figure 1).
The findings indicated that the strength of the asso-
ciation between challenge and SAT-9 value-added
scores increased as the proportion of African Amer-
ican students increased.
No significant interactions were identified in the
model prediction MATH value-added scores. As
such, only the main effects of this model were inter-
preted.
Table 2
Percentages of Sociodemographic Characteristics for Full Sample
(N=634) and African American Subsample (n=223)
Characteristic
Full
sample (%)
African
American
subsample (%)
Teacher ——
Female 91 91
African American 40 75
Caucasian 54 22
Latino(a) 5 <1
Master’s degree or higher 54 75
Classroom ——
Female 50 50
African American 52 93
Caucasian 17 1
Latino(a) 23 5
Asian 5 <1
English language learners 15 4
Special education 9 5
Free and reduced price lunch 44 27
Note. Asian and Native American teachers comprised <1% of
the full teacher sample. Native American students comprised
<5% of the full classroom sample.
Warmth and Demand 9
For ELA, a significant positive interaction was
found between challenge and the proportion of
African American students (b=.14, p<.01) in a
classroom. Simple slopes showed significance at
average (b=.04, SE =.01, t=2.96, p=.003) and
high (b=.07, SE =.02, t=3.11, p=.002) propor-
tions of African American students. The association
between challenge and ELA value-added scores
increased as the number of African American stu-
dents increased (Figure 1).
In a subsample with 80% or more African Amer-
ican students per classroom (Table 5), analyses
revealed a main effect of challenge for BAM
(b=.19, p<.05), SAT (b=.21, p<.01), MATH
(b=.23, p<.01), and ELA (b=.32, p<.001). No
significant interactions were found between Tripod
scores and teacher–student ethnic match or mis-
match; only the main effects of the care, control,
and challenge on value-added scores were inter-
preted.
An alternative angle for predicting achievement
growth is to consider variability in students’per-
ceptions of the classroom environment, in addition
to central tendency. For example, if 20 children
reporting on their teacher all have disparate views,
that classroom experience is potentially different
from a classroom in which 20 children convey simi-
lar views of their teacher. As such, we explored
this angle through the use of the coefficient of vari-
ation. The coefficient of variation is a statistical
technique used to measure variation in an observed
variable (Bedeian & Mossholder, 2000) and was
used as an indicator of variability in students’rat-
ings of challenge, control, and care. Main effect
findings indicated that higher variability in student
perceptions of demand (challenge and control) was
associated with lower achievement. Regarding the
interactions, as the variability in students’
perceptions of their teacher’s level of challenge
increased, value-added achievement scores
decreased. The negative relation between variability
in challenge and achievement was stronger in class-
rooms with a higher proportion of African Ameri-
can students.
Discussion
The results of this study partially support the warm
demander theory, with demand emerging as a par-
ticularly important construct as it significantly and
positively related to students’academic growth.
Main effect analyses indicated that students who
perceived their teachers as more challenging and
higher in control showed greater growth on both
high- and low-stakes math and reading assess-
ments. Moderation analyses indicated that the posi-
tive relation between challenge and academic
growth was stronger in classrooms with greater
numbers of African American students. Findings
showed that teacher–student ethnic match or mis-
match was not associated with student academic
growth in this sample.
Consistent with our hypotheses, student percep-
tions of an academically rigorous and well-mana-
ged classroom environment contributed to
academic growth in fourth- and fifth-grade class-
rooms on both high- and low-stakes achievement
measures in math and reading. Prior research cor-
roborates the importance of challenge and control
(e.g., Pianta & Hamre, 2009; Rubie-Davies, 2007).
Appropriately challenging learning contexts, in
which teachers convey high expectations and offer
academically rigorous content, may contribute to
students’self-perceptions and internalizations of
their own ability to exert effort and achieve
Table 3
Correlations Among Tripod Predictors and Value-Added Outcomes
1234 567
1. Challenge —.44*** .44*** .10* .15*** .21*** .22**
2. Control —.61*** .15** .19*** .20*** .22***
3. Care —.08 .12** .14** .13**
4. Balanced Assessment in Mathematics —.44*** .36*** .34***
5. SAT-9 —.21*** .35***
6. MATH —.56***
7. English language arts —
Raw M(SD) 16.68 (1.22) 13.94 (1.60) 29.06 (2.42) 17.97 (5.88) 106.23 (86.97) ——
Range 12.50–16.68 8.87–19.67 19.51–34.63 3.90–33.54 5.95–521.97 ——
*p <.05. **p<.01. ***p<.001.
10 Sandilos, Rimm-Kaufman, and Cohen
Table 4
Main Effects of Tripod and Moderation Effects of Proportion African American Students to Predict Value-Added Scores
Low-stakes MET assessments High-stakes state assessments
Model 1
BAM (n=575)
R
2
=.03*
Model 2
BAM (n=574)
R
2
=.04**
Model 1
SAT-9 (n=578)
R
2
=.05*
Model 2
SAT-9 (n=577)
R
2
=.06**
Model 1
MATH (n=594)
R
2
=.07***
Model 2
MATH (n=593)
R
2
=.07***
Model 1
ELA (n=594)
R
2
=.07**
Model 2
ELA (n=593)
R
2
=.09**
Fourth grade .00 (.05) .01 (.05) .02 (.04) .02 (.04) .00 (.05) .01 (.05) .02 (.05) .03 (.05)
Proportion special education .02 (.11) .06 (.04) .03 (.05) .03 (.05) .02 (.04) .02 (.04) .02 (.04) .02 (.05)
Proportion ELL .03 (.07) .02 (.06) .03 (.05) .03 (.05) .04 (.04) .04 (.04) .03 (.05) .02 (.04)
Proportion Male .00 (.05) .02 (.05) .02 (.05) .00 (.04) .02 (.04) .03 (.04) .01 (.04) .03 (.06)
Challenge .06 (.06) .07 (.06) .11 (.05)* .12 (.06)* .18 (.05)*** .19 (.01)*** .18 (.06)** .19 (.01)***
Control .14 (.05)** .14 (.05)** .16 (.07)* .16 (.07)* .15 (.05)** .15 (.01)** .18 (.06)** .17 (.01)**
Care .03 (.06) .04 (.06) .03 (.05) .04 (.05) .04 (.06) .05 (.02) .06 (.07) .07 (.02)
African American —.02 (.08) —.02 (.09) —.01 (.02) —.01 (.01)
Challenge 9African American —.13 (.06)* —.12 (.06)* —.07 (.02) —.14 (.01)**
Control 9African American —.01 (.05) —.07 (.06) —.01 (.01) —.03 (.01)
Care 9African American —.04 (.06) —.02 (.05) —.02 (.02) —.04 (.01)
Note. Standardized estimates are reported with the standard error in parenthesis. District was also controlled for as a fixed factor in the analysis. Model 1 =main effects model;
Model 2 =moderation model; African American =classroom proportion variable; BAM =Balanced Assessment in Mathematics; ELA =English language arts; ELL =English lan-
guage learners; MET =Measures of Effective Teaching. *p<.05. **p<.01. ***p<.001.
Warmth and Demand 11
academically (Lee, 2012; Rubie-Davies, 2007). More-
over, classroom environments that are high on con-
trol, such that time is used productively and there
are minimal behavioral disruptions, are more con-
ducive to learning (Woolfolk Hoy & Weinstein,
2011).
Low-Stakes Math Assessment: BAM Low-Stakes Reading Assessment: SAT9
High-Stakes Reading Assessment: ELA
Figure 1. Interaction between challenge and proportion of African American (AA) students to predict the low- and high-stakes assess-
ments.
*p<.05. **p<.01.
Table 5
Main Effect of Tripod in the African American Student Subsample
Low-stakes MET assessments High-stakes state assessments
BAM (n=191)
R
2
=.04
SAT-9 (n=192)
R
2
=.09
MATH (n=199)
R
2
=.10**
ELA (n=199)
R
2
=.14**
Fourth grade .09 (.09) .02 (.07) .05 (.09) .05 (.09)
Proportion special education .07 (.06) .00 (.08) .01 (.06) .01 (.09)
Proportion ELL .03 (.25) .22 (.18) .11 (.17) .08 (.19)
Proportion male .03 (.10) .02 (.06) .02 (.07) .07 (.06)
Challenge .19 (.09)* .21 (.10)* .23 (.08)** .32 (.08)***
Control .01 (.06) .02 (.11) .12 (.08) .12 (.08)
Care .02 (.09) .01 (.08) .07 (.12) .05 (.10)
Note. Standardized estimates are reported in the table with standard errors in parenthesis. District was controlled for as a fixed factor in
the analysis. BAM =Balanced Assessment in Mathematics; ELA =English language arts; ELL =English language learners; MET =Mea-
sures of Effective Teaching. *p<.05. **p<.01. ***p<.001.
12 Sandilos, Rimm-Kaufman, and Cohen
The present findings extend existing work by
using student reports to indicate that control and
challenge are important and complementary
instructional practices. Given these encouraging
and consistent findings, teacher preparation and
professional development providers should think
critically about how to best support novice and
experienced teachers alike in developing these prac-
tices in tandem. Additionally, the instructional
methods employed by teachers likely serve as the
foundation for students’perceptions of their envi-
ronment. Thus, the positive influence that teachers’
demand has on the classroom should be considered
in the context of instructional practices that are
regarded as high quality, such as scaffolding, con-
cept development, and proactive redirection of
behavior. Research on teachers’demand could be
expanded by examining this construct in the pres-
ence and absence of high-quality teaching strate-
gies.
Results regarding students’perception of teacher
caring were more complex to interpret. Analyses
examining the contribution of challenge, control,
and care simultaneously showed no statistically sig-
nificant association between care and value-added
scores. However, as noted in the Results section,
analyses that considered the contribution of care
alone showed a statistically significant contribution
of care to achievement growth. These nuanced find-
ings require careful interpretation involving both
methodological and substantive considerations.
Challenge, control, and care correlate with one
another and thus, the teaching practices situated at
the nexus between care and challenge, or between
care and control, may be associated with improved
achievement. Within a warm demander framework,
care and challenge are particularly difficult to tease
apart because one way for teachers to implicitly
demonstrate care for their students is to hold high
expectations for those students (Gay, 2002; Ladson-
Billings, 2009). Yet another explanation stems from
the selection of achievement on standardized tests
as the study outcome. High performance on such
assessments requires academic challenge in an
orderly environment. Simply demonstrating caring
behaviors without providing time and opportunity
to learn, as well as academic work well-matched to
the student about to take the test, may create a pos-
itive classroom climate, but may not facilitate the
academic learning essential for achievement. Care
may play a larger role when examining outcomes
proximal to student learning behaviors, such as stu-
dents’engagement in learning (e.g., Hughes &
Kwok, 2007; Rimm-Kaufman, Baroody, Larsen,
Curby, & Abry, 2015; Roorda et al., 2011). The find-
ings raise important questions for future research.
For instance, in classrooms that present high chal-
lenge and control, does care relate to achievement
indirectly by boosting student emotional function-
ing and learning behaviors?
As mentioned previously, the demand (i.e.,
challenge and control) component of the warm
demander theory was associated with growth on
high- and low-stakes achievement tests. Challenge
emerged as a particularly vital aspect of teaching
when exploring classrooms with greater numbers of
African American students. Findings related to
teachers’level of demand resemble recommenda-
tions for culturally responsive teaching practices
with African American students, indicating the
value of high expectations for students of color
(Gay, 2002; Ladson-Billings, 2009).
The importance of challenge in classrooms with
greater proportions of African American students
also has implications for contemporary issues in
educational research and policy. Recent work
reveals disparities in the quality of education in
low-income and urban areas with larger minority
student populations compared to other settings
(Pianta et al., 2007). These disparities have sparked
debates in the education system regarding the
equality of access to high-quality education for Afri-
can American and Latino students (National Center
for Education Evaluation, 2011). Results from the
present study indicate that students’perception of
challenging instruction is a particularly powerful
aspect of teaching for African American students.
Although the design of this study does not warrant
causal inferences, the work does suggest the impor-
tance of future work on interventions that boost
expectations and prepare teachers to challenge stu-
dents, particularly African American students.
Warm demander theory developed out of obser-
vations of highly effective African American teach-
ers who could leverage a shared cultural
background to develop strong relationships with
African American students (Irvine & Fraser, 1998;
Ware, 2006). Given that this theory emerged from
classrooms with ethnic match between teachers and
students, it was necessary to explore the influence
of teacher race in the current study. Results indi-
cated that with the current sample, ethnic match or
mismatch between teachers and students did not
interact significantly with challenge, control, or care
in classrooms with predominantly African Ameri-
can students. This finding is not altogether surpris-
ing given that other empirical research regarding
racial or ethnic congruence between teachers and
Warmth and Demand 13
students has been fairly mixed and inconclusive
(Dee, 2005; Kline, Le, & Hamilton, 2001; Rimm-
Kaufman et al., 2000; Saft & Pianta, 2001). The lack
of significance in this study can be viewed posi-
tively in that challenging environments appear to
benefit the African American elementary students
regardless of their teachers’ethnicity. However, the
finding also requires further inquiry. The MET data
are rich, but not exhaustive. For instance, the data
collected do not tap elements of African American
culture embedded in warm demander practices,
such as using culturally specific communication
styles (Ford & Sassi, 2014) or referencing shared
cultural history and racial identity (Ford & Sassi,
2014; Ware, 2006). In the presence of data with
more cultural nuance or more varied outcomes,
such as engagement in learning rather than test per-
formance, it is quite possible that the ethnic match
between teachers and students could play a larger
role in the results.
Limitations and Future Directions
Several limitations should be noted. First, the
MET data set is not a nationally representative
sample of teachers and students (Kane & Staiger,
2012), which limits the generalizability of the find-
ings. Forty percent of the teachers in the full sam-
ple were African American and three quarters of
the teachers were African American in the subsam-
ple compared to roughly 7% in public school
national averages (National Center for Education
Statistics [NCES], 2013). Compared to most
research on U.S. educators, these data afforded a
large and diverse set of teachers and a larger than
typical representation of African American teach-
ers. However, one necessary next step is to use a
nationally representative sample with a range of
grade levels to examine questions of ethnic match
or mismatch. Schools are a critical developmental
context for youth as they establish their ethnic and
racial identity, and ethnic match between teachers
and students may hold different meaning and
importance at varying developmental levels or
grades.
In addition to representativeness, this study
examines challenge, control, and care in a descrip-
tive manner, such that these constructs cannot be
linked causally to value-added outcomes. A future
randomized control trial of an intervention
designed to enhance challenge, control, and care
would facilitate causal inferences. Yet another need
is to understand the contribution of challenge, con-
trol, and care on other important student learning
behaviors including engagement, motivation, and
efficacy.
To our knowledge, there is no existing quantita-
tive measure of warm demander teacher character-
istics and thus, the Tripod measure used in the
MET study was the closest proxy for the constructs
of warmth and demand. Important aspects of cul-
turally specific interactions (e.g., communication
style, discussions of shared cultural history, and
racial or ethnic identity) are missing from the Tri-
pod, however. A new quantitative measure of
warm demander characteristics is needed to mea-
sure culturally informed practices for African Amer-
ican students.
Finally, limitations to the value-added models
used to determine achievement growth must be
acknowledged. For instance, there is limited
evidence for the reliability and stability of value-
added scores. Despite these limitations, value-
added scores are used with increasing frequency in
teacher evaluation nationwide. Many districts use
value-added scores to make high-stakes decisions
about employment and salary increases (Chetty
et al., 2011). A beneficial aspect of value-added
models is that the data included within the esti-
mate (e.g., students’scores in a measure of that
content domain from a year prior, individual stu-
dent and aggregate classroom demographic infor-
mation) helps to control for the potential
limitations of using nonrandomized classroom data,
such as the purposeful sorting of students into cer-
tain classrooms.
As mentioned in the Results section, variability
in student ratings within a classroom also was
examined and findings revealed that higher vari-
ability in ratings was related to lower levels of
achievement growth. Another avenue for future
research is to further explore this question to better
understand the adverse effect that variability in
student perceptions has on achievement. For exam-
ple, it is possible that variability is not a product
of differentiation, but rather it is an artifact of a
teacher interacting more positively with, and pro-
viding more challenging experiences for, some stu-
dents and not others. Our finding that the inverse
relation between variability in perceptions and
achievement outcomes was strengthened as the
proportion of African American students in a
classroom increased is consistent with prior
research indicating that African American students
may be more attuned to differential expectations
conveyed by their teachers than their Caucasian
peers (Jussim et al., 1996; McKown & Weinstein,
2003).
14 Sandilos, Rimm-Kaufman, and Cohen
Conclusion
This study examined the warm demander theory
and showed the salience of demand (challenge and
control) on academic growth in fourth- and fifth-
grade classrooms. The work suggests that challenge
may be even more important for African American
students, regardless of the ethnicity of the teacher.
The work contributes knowledge in education at a
time when the United States is deeply concerned
about improving the academic outcomes for ethnic
minority students. The findings are based on
diverse classrooms and lend insight into how stu-
dents’perceptions of classroom processes relate to
selected high-stakes achievement outcomes used as
policy-relevant benchmarks. In closing, we high-
light three implications of the work. First, in light
of the current findings, teachers should be encour-
aged to hold high expectations for all students, par-
ticularly African American students in their
classrooms. Second, professional development pro-
grams need to emphasize the value of classroom
management and academic rigor without negating
the importance of creating a caring environment.
Third, the field of education has many theories sim-
ilar to the warm demander pedagogy that have
been described well and explored using small quali-
tative data sets. The availability of new national
data sets such as the MET study provides research-
ers with new opportunities to test those theories
with larger and more varied samples.
References
Bedeian, A. G., & Mossholder, K. W. (2000). On the use
of coefficient of variation as a measure of diversity.
Organizational Research Methods,3, 285–297.
doi:10.1177/109442810033005
Bondy, E., & Ross, D. D. (2008). The teacher as warm
demander. Educational Leadership,66,54–58.
Braun, H. I. (2005). Using student progress to evaluate teach-
ers: A primer on value-added models. Princeton, NJ: Edu-
cational Testing Service.
Brock, L. L., Nishida, T. K., Chiong, C., Grimm, K. J., &
Rimm-Kaufman, S. E. (2008). Children’s perceptions of
the classroom environment and social and academic per-
formance: A longitudinal analysis of the contribution of
the Responsive Classroom approach. Journal of School
Psychology,46, 129–149. doi:10.1016/j.jsp.2007.02.004
Brophy, J. E., & Good, T. L. (1970). Teachers’communica-
tion of differential expectations for children’s classroom
performance: Some behavioral data. Journal of Educa-
tional Psychology,61, 365–374.
Cabrera, N. J., Beeghly, M., & Eisenberg, N. (2012). Posi-
tive development of minority children: Introduction to
the special issue. Child Development Perspectives,6, 207–
209. doi:10.1111/j.1750-8606.2012.00253.x
Chetty, R., Friedman, J. N., & Rockoff, J. E. (2011). The
long-term impacts of teachers: Teacher value-added and stu-
dent outcomes in adulthood. National Bureau of Economic
Research. Retrieved from http://www.nber.org/pa-
pers/w17699
Cholewa, B., Goodman, R. D., West-Olatunji, C., & Ama-
tea, E. (2014). A qualitative examination of the impact
of culturally responsive education practices on the psy-
chological well-being of students of color. The Urban
Review,46, 574–596. doi:10.1007/s11256-014-0272-y
Clunies-Ross, P., Little, E., & Kienhuis, M. (2008). Self-
reported and actual use of proactive and reactive class-
room management strategies and their relationship with
teacher stress and student behaviour. Educational Psychol-
ogy: An International Journal of Experimental Educational
Psychology,28, 693–710. doi:10.1080/01443410802206700
Cothran, D. J., Kulinna, P. H., & Garrahy, D. A. (2003).
“This is kind of giving a secret away. . .”: Students’
perspectives on effective class management. Teaching
and Teacher Education,19, 435–444. doi:10.1016/S0742-
051X(03)00027-1
Dee, T. S. (2005). A teacher like me: Does race, ethnicity,
or gender matter? The American Economic Review,95,
158–165. doi:10.1257/000282805774670446
Downer, J. T., Stuhlman, M., Schweig, J., Mart
ınez, J. F.,
& Ruzek, E. (2014). Measuring effective teacher-student
interactions from a student perspective: A multi-level
analysis. Journal of Early Adolescence,35, 722–758.
doi:10.1177/0272431614564059
Eccles, J. S. (1999). The development of children ages 6 to
14. The Future of Children,9,30–44. Retrieved from
http://www.jstor.org/stable/1602703
Enders, C. K. (2010). Applied missing data analysis. New
York, NY: Guilford.
Ferguson, R. F. (2008). The tripod project framework. Cam-
bridge, MA: Harvard University.
Ferguson, R. F. (2010, October). Student perceptions of
teaching effectiveness (Discussion Brief). National Center
for Teacher Effectiveness and the Achievement Gap Ini-
tiative, Harvard University.
Ford, A. C., & Sassi, K. (2014). Authority in cross-racial
teaching and learning (re) considering the transferabil-
ity of warm demander approaches. Urban Education,49,
39–74. doi:10.1177/0042085912464790
Fulmer, S. M., & Turner, J. C. (2014). The perception and
implementation of challenging instruction by middle
school teachers. The Elementary School Journal,114, 303–
326. doi:10.1086/674053
Gay, G. (2002). Preparing for culturally responsive teach-
ing. Journal of Teacher Education,53, 106–116.
doi:10.1177/0022487102053002003
Goddard, R. D., Sweetland, S. R., & Hoy, W. (2000). Aca-
demic emphasis of urban elementary schools and stu-
dent achievement in reading and mathematics: A
multilevel analysis. Educational Administration Quarterly,
36, 682–702. doi:10.1177/00131610021969164
Warmth and Demand 15
Good, T. L. (1981). Teacher expectations and student per-
ceptions: A decade of research. Educational Leadership,
38, 415–422.
Gregory, A., & Weinstein, R. S. (2008). The discipline gap
and African Americans: Defiance and cooperation in
the high school classroom. The Journal of School Psychol-
ogy,46, 455–475. doi:10.1016/j.jsp.2007.09.001
Howard, T. C. (2001). Telling their side of the story: Afri-
can-American students’perceptions of culturally rele-
vant teaching. The Urban Review,33, 131–149.
Hughes, J., & Kwok, O. (2007). Influence of student-tea-
cher and parent-teacher relationships on lower achiev-
ing readers’engagement and achievement in the
primary grades. Journal of Educational Psychology,99(1),
39–51. doi:10.1037/0022-0663.99.1.39
Irvine, J. J. (Ed.). (2002). In search of wholeness: African
American teachers and their culturally specific classroom
practices. New York, NY: Palgrave.
Irvine, J., & Fraser, J. (1998, May 13). “Warm deman-
ders”: Do national certification standards leave room
for the culturally responsive pedagogy of African
American teachers? Education Week, p. 56.
Jussim, L., Eccles, J. S., & Madon, S. (1996). Social percep-
tion, social stereotypes, and teacher expectations: Accu-
racy and the quest for the powerful self-fulfilling
prophecy. In M. P. Zanna (Ed.), Advances in experimen-
tal social psychology (Vol. 28, pp. 281–388). New York,
NY: Academic Press.
Kane, T. J., & Staiger, D. O. (2012). Gathering feedback for
teachers: Combining high-quality observations with student
surveys and achievement gains. Policy and Practice Brief
Prepared for the Bill and Melinda Gates Foundation.
Retrieved from http://metproject.org/downloads/
MET_Gathering_Feedback_Practioner_Brief.pdf
Klem, A. M., & Connell, J. P. (2004). Relationships matter:
Linking teaching support to student engagement and
achievement. Journal of School Health,74, 262–273.
doi:10.1111/j.1746-1561.2004.tb08283.x
Kline, S., Le, V. N., & Hamilton, L. (2001). Does matching
student and teacher racial/ethnic group improve math?
(DRU-2529-EDU). RAND Education. Retrieved from
www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA390470
Ladson-Billings, G. (2009). The Dreamkeepers: Successful
teachers of African American children (2nd ed.). San Fran-
cisco, CA: Jossey-Bass.
Lee, J. (2012). The effects of the teacher-student relation-
ship and academic press on student engagement and
academic performance. International Journal of Educa-
tional Research,53, 330–340. doi:10.1016/j.ijer.2012.
04.006
Lee, V., & Smith, J. B. (1999). Social support and achieve-
ment for young adolescents in Chicago: The role of
school academic press. American Education Research
Journal,36, 907–945. doi:10.3102/00028312036004907
McKown, C., & Weinstein, R. (2002). Modeling the role of
child ethnicity and gender in children’s differential
response to teacher expectations. Journal of Applied
Social Psychology,32, 159–184.
McKown, C., & Weinstein, R. (2003). The development
and consequences of stereotype consciousness in mid-
dle childhood. Child Development,74, 498–515.
doi:10.1111/1467-8624.7402012
Mitchell, M. M., Bradshaw, C. P., & Leaf, P. J. (2010). Stu-
dent and teacher perceptions of school climate: A mul-
tilevel exploration of patterns of discrepancy. Journal of
School Health,80, 271–279. doi:10.1111/j.1746-1561.2010.
00501.x
Muth
en, L.K., & Muth
en, B.O. (1998–2012). Mplus user’s
guide (7th ed.). Los Angeles, CA: Author.
National Center for Education Evaluation. (2011, April).
Do low-income students have equal access to the highest-per-
forming teachers? (NCEE 2011-4016). Retrieved from
http://ies.ed.gov/ncee/pubs/20114016/pdf/
20114016.pdf
National Center for Education Statistics (NCES). (2013).
Digest of education statistics. Retrieved from https://
nces.ed.gov/programs/digest/d13/tables/
dt13_209.10.asp
Pianta, R. C., Belsky, J., Houts, R., & Morrison, F. (2007).
Opportunities to learn in America’s elementary class-
rooms. Science,315, 1795–1796. doi:10.1126/science.
1139719
Pianta, R. C., & Hamre, B. K. (2009). Conceptualization,
measurement, and improvement of classroom pro-
cesses: standardized observation can leverage capacity.
Educational Researcher,38, 109–119. doi:10.3102/
0013189X09332374
Pianta, R. C., & Walsh, D. J. (1996). High risk children in
schools. New York, NY: Routledge.
Polikoff, M. S. (2014). The Stability of Observational and
Student Survey Measures of Teaching Effectiveness. The
American Journal of Education,121, 183–212.
doi:10.1086/679390
Ponitz, C. C., Rimm-Kaufman, S. E., Grimm, K. J., &
Curby, T. W. (2009). Kindergarten classroom quality,
behavioral engagement, and reading achievement.
School Psychology, Review,38, 102–120.
Reyes, M. R., Brackett, M. A., Rivers, S. E., White, M., & Sal-
ovey, P. (2012). Classroom emotional climate, student
engagement, and academic achievement. Journal of Educa-
tional Psychology,104(3), 700–712. doi:10.1037/a0027268
Rimm-Kaufman, S. E., Baroody, A. E., Larsen, R. A. A.,
Curby, T. W., & Abry, T. (2015). To what extent do
teacher–student interaction quality and student gender
contribute to fifth graders’engagement in mathematics
learning? Journal of Educational Psychology,107, 170–185.
doi:10.1037/a0037252
Rimm-Kaufman, S. E., & Chiu, Y. J. (2007). Promoting
social and academic competence in the classroom. Psy-
chology in the Schools,44, 397–413. doi:10.1002/pits.20231
Rimm-Kaufman, S. E., Curby, T. W., Grimm, K. J.,
Nathanson, L., & Brock, L. L. (2009). The contribution
of children’s self-regulation and classroom quality to
children’s adaptive behaviors in the kindergarten class-
room. Developmental Psychology,45, 958–972.
doi:10.1037/a0015861
16 Sandilos, Rimm-Kaufman, and Cohen
Rimm-Kaufman, S. E., Pianta, R. C., & Cox, M. J. (2000).
Teachers’judgments of problems in the transition to
kindergarten. Early Childhood Research Quarterly,12,
363–385. doi:10.1016/S0885-2006(00)00049-1
Roorda, D. L., Koomen, H. M. Y., Spilt, J. L., & Oort, F. J.
(2011). The influence of affective teacher-student rela-
tionships on students’school engagement and achieve-
ment: A meta-analytic approach. Review of Educational
Research,81, 493–529. doi:10.3102/0034654311421793
Rosenthal, R., & Jacobsen, L. (1968). Pygmalion in the class-
room: Self-fulfilling prophecies and teacher expectations.
New York, NY: Holt, Rhinehart, and Winston.
Rubie-Davies, C. M. (2007). Classroom interactions:
Exploring the practices or high- and low-expectation
teachers. British Journal of Educational Psychology,77,
289–306. doi:10.1348/000709906X101601
Saft, E. W., & Pianta, R. C. (2001). Teachers’perceptions
of their relationships with students: Effects of child,
age, gender and ethnicity of teachers and children.
School Psychology Quarterly,16, 125–141. doi:10.1521/
scpq.16.2.125.18698
Silva, J. M., Langhout, R. D., Kohfeldt, D., & Gurrola, E.
(2015). “Good”and “bad”kids? A race and gender
analysis of effective behavioral support in elementary
school. Urban Education,50, 787–811. doi:10.1177/
0042085914534859
Skinner, E. A., Kindermann, T. A., Connell, J. P., & Well-
born, J. G. (2009). Engagement and disaffection as orga-
nizational constructs in the dynamics of motivational
development. In K. Wentzel & A. Wigfield (Eds.),
Handbook of motivation at school (pp. 232–241). New
York, NY: Routledge.
Stipek, D. (2002). Good instruction is motivating. In A. Wig-
field & J. Eccles (Eds.), Development of achievement motiva-
tion (pp. 309–332). San Diego, CA: Academic Press.
Vasquez, J. (1989). Contexts of learning for minority stu-
dents. The Education Forum,52, 243–253. doi:10.1080/
00131728809335490
Villegas, A. M., & Lucas, T. (2002). Preparing culturally
responsive teachers: Rethinking the curriculum. Journal
of Teacher Education,53,20–32. doi:10.1177/0022487102
053001003
von der Embse, N. P., Kilgus, S., Solomon, H. J., Bowler,
M., & Curtiss, C. (2015). Initial development and factor
structure of the educator test stress inventory. Journal of
Psychoeducational Measurement,33,223–237. doi:10.1177/
0734282914548329
Ware, F. (2006). Warm demander pedagogy: Culturally
responsive teaching that supports a culture of achieve-
ment for African American students. Urban Education,
41, 427–456. doi:10.1177/0042085906289710
Wentzel, K. (1997). Student motivation in middle school:
The role of perceived pedagogical caring. Journal of
Educational Psychology,89, 411–419. doi:10.1037/0022-
0663.89.3.411
White, M., & Rowan, B. (2013). User guide to the measures
of effective teaching longitudinal database (MET LBD). Ann
Arbor, MI: Inter-University Consortium for Political
and Social Research.
Williams, J. M., & Bryan, J. (2013). Overcoming adversity:
High achieving African American youth’s perspectives
on educational resilience. Journal of Counseling &
Development,91, 291–300. doi:10.1002/j.1556-6676.2013.
00097.x
Woolfolk Hoy, A., & Weinstein, C. S. (2011).
Student and teacher perspectives on classroom man-
agement. In C. M. Evertson & C. S. Weinstein (Eds.),
Handbook of classroom management: Research, practice,
contemporary issues (pp. 181–223). Mahwah, NJ: Erl-
baum.
Zaslow, M., Martinez-Beck, I., Tout, K., & Halle, T.
(2011). Quality measurement in early childhood settings.
Baltimore, MD: Paul H. Brookes.
Warmth and Demand 17
