ERIC ED461693: Class Size: What Does Research Tell Us? Spotlight on Student Success.

7 Pages·1997·0.14 MB·English

by  ERIC

DOCUMENT RESUME
UD 032 398
ED 461 693
Finn, Jeremy D.
AUTHOR
Class Size: What Does Research Tell Us? Spotlight on Student
TITLE
Success.
Mid-Atlantic Lab. for Student Success, Philadelphia, PA.
INSTITUTION
Office of Educational Research and Improvement (ED),
SPONS AGENCY
Washington, DC.
REPORT NO LSS-Ser-207
PUB DATE 1997-00-00
NOTE
6p.
Laboratory for Student Success, 1301 Cecil B. Moore Ave.,
AVAILABLE FROM
Philadelphia, PA 19122-6901. Tel: [phone] (Toll Free).
Information Analyses (070)
PUB TYPE
MF01/PC01 Plus Postage.
EDRS PRICE
*Academic Achievement; *Class Size; Disadvantaged Youth;
DESCRIPTORS
*Educational Research; Elementary Secondary Education; *High
Risk Students; *Small Classes; *Teacher Student Ratio; Urban
Schools; Urban Youth
*Student Teacher Achievement Ratio Project TN
IDENTIFIERS
ABSTRACT
This report summarizes the findings of some recent pivotal
studies of class size, especially as they relate to students at risk. Of
particular note is a large-scale study of class size, which was designed to
test the conclusions of G. Glass and M. Smith (1978) and G. Robinson (1990)
about the advantages of small class size. This study, Project Student-Teacher
Achievement Ratio (Project STAR) was conducted in Tennessee beginning in
1985. Project STAR was a controlled scientific experiment that studied a
large number of teachers (over 300 each year) and a large number of students
(more than 10,000 in all) over 4 years. Both norm-referenced and
curriculum-referenced tests were administered each year. Because of the tight
experimental control of Project STAR and the magnitude of the study,
(1) small classes (17
educations can be confident of certain principles:
pupils or fewer) are more effective academically than larger classes (22 and
(2) the advantage of small
above) in the primary grades in all subject areas;
classes is greater for minority students and students in inner-city schools
than for white students; and (3) the advantage of small classes can probably
be attributed to the fact that students are more actively engaged in learning
compared to their peers in larger classes. Even these strong findings do not
fully resolve questions of the cost effectiveness of small classes. In
addition, there is still not enough information to tell whether small classes
might be especially advantageous if the teachers received special training to
take advantage of small class size. There is little information to tell
whether small classes could be especially beneficial when used in conjunction
with other compensatory or accelerated programs. Current research does not
tell much about the effects of small class size in middle school or high
school. (Contains 13 references.)
(SLD)
Reproductions supplied by EDRS are the best that can be made
from the original document.
tit
Class Size: What Does Research Tell Us?
Jeremy D. Finn
LSS Spotlight on Student Success
A Digest of Research from the Laboratory for Student Success, No. 207
OF EDUCATION
U.S. DEPARTMENT
and Improvement
Office of Educational Research
INFORMATION
EDUCATIONAL RESOURCES
CENTER (ERIC)
reproduced as
1:1 This document has been
organization
received from the person or
originating it.
)1( Minor changes have been made to
improve reproduction quality.
in this
Points of view or opinions stated
represent
document do not necessarily
official OERI position or policy.
2
BESTCOPY AVAILABLE
/
LSS
an'
A digest of research from the Laboratory for Student Success
No. 207
Class Size: What Does Research Tell Us?
by
Jeremy D. Finn
Professor of Education, SUNY-Buffalo
Visiting Scholar, Laboratory for Student Success
might occur as well; and we are
The debate about class size is
current state of knowledge. The first
uncertain whether any of these
not new. In fact, the Babylonian
is Glass and Smith's (1978) statisti-
cal reviewa "meta-analysis"of
benefits are worth the additional
Talmud, completed in the sixth cen-
expense.
tury, contains class size principles for
the findings of over 80 empirical stud-
The intent of this report is to
ies. While their review identified
study of the Bible: "The number of
summarize the findings of some
many weaknesses in these studies,
pupils assigned to each teacher is 25.
If there are 50, we appoint two
including the fact that no scientifi-
recent pivotal studies of class size,
teachers. If there are 40, we appoint
especially as they impact on students
cally controlled experiment had ever
an assistant, at the expense of the
at risk. Three themes are apparent
been performed on the question of
town" (translation from Epstein,
class size, Glass and Smith arrived
in this work:
Small classes are academically
1976, p. 214, emphasis added). The
at two major conclusions. First, "re-
duced class size can be expected
advantageous, particularly for
two iirinciples embodied in this state-
to produce increased academic
minority students or students
ment have confronted one another
ever sincethe desirability of limit-
achievement" (p. iv), and second,
living in poverty.
"the major benefits from reduced
ing the number of students working
Small classes promote the devel-
with one teacher, and the real eco-
opment of classroom behaviors
class size are obtained as the size is
nomic considerations to be faced.
reduced below 20 pupils" (p. v).
that are important to learning and
Several years later, Educational
Fourteen centuries later, we continue
may provide significant long-term
to debate the benefits and costs of
benefits.
Research Service (Robinson, 1990)
small classes.
The advantages of small classes
synthesized a much larger set of stud-
Assessing the immediate costs
ies. The review concluded that re-
are best realized when there are
sults were "mixed," that is, it is
no major obstacles to implemen-
of small classes is relatively straight-
difficult at best to decide whether
forward, although some schools are
tation.
finding ways to minimize the expense
small classes are more effective than
RECENT RESEARCH ON
through principles of "substitution"
large classes, but to the extent that
SMALL CLASSES
they are, two conclusions can be
rather than "add-on." Assessing the
benefits is more complex because of
drawn. First, small classes are prob-
The extensive research on small
the uncertainties it involves. We are
classes has been reviewed time and ably most beneficial in reading and
mathematics in the early primary
uncertain whether real achievement
time again. Of the many reports that
gains will be realized from smaller
grades, and second, "students who
have been produced, however, three
classes; we are uncertain whether
are economically disadvantaged or
deserve particular attention because
important non-achievement gains
they created the foundation for the
from some ethnic minorities perform
Spotlight on Student Success is an occasional series of articles highlighting findings from the Laboratory for Student Success (LSS), the
Mid-Atlantic Regional Educational Laboratory, that have significant implications for improving the academic success of students in the
mid-Atlantic region. For information about the LSS and other LSS publications, contact the Laboratory for Student Success, 1301 Cecil
B. Moore Avenue, Philadelphia, PA, 19122-6091; telephone: [phone]; e-mail: [email protected]. Also visit our World Wide
Web site at http://www.temple.edu/LSS.
3
better academically in smaller
No differences were found on
classes had fewer than 20 students;
classes" (p. 85).
the study's design enabled research-
student self-concept or motivation.
In interpreting these findings,
ers to look at the effects on minority
The magnitude of the small-
one caveat must be noted. The con-
as well as majority students. Project
class advantage was about 2/10 of
clusions drawn by Glass and Smith
STAR was a controlled scientific ex-
and by Educational Research Ser-
a standard deviation, or about .1GE
periment with a number of additional
features: it was based on a large
vice were not based on the analysis
in kindergarten in each subject area.
of pupil:teacher ratios for a school
number of teachers (over 300 each
The advantage was about 3/10 of a
or district, since the pupil:teacher
year) and students (over 10,000 in
standard deviation, or about .2GE,
ratio can camouflage many features
all); students were followed for four
in each subject in subsequent grades.
The effect was slightly smaller
of instruction. The studies were
years; and both norm-referenced
among White students and some-
based on actual counts of the num-
and curriculum-based achievement
ber of pupils in a classroom. That
tests were administered each year.
what larger among minority students
each year. The minority-White
these two ways of looking at class
Because of these features, Project
STAR provided educators with
sizeteacher:pupil ratios and actual
achievement gap was reduced sub-
student countsdo not have the
definitive answers about the impact
stantially in small classes, especially
of small classes in the primary
on the criterion-referenced tests.
same impact has been demonstrated
grades.
For example, in regular classes the
repeatedly (see, for example, Boozer
The results of Project STAR
& Rouse, 1995). Small classesto
pass rate for African-American stu-
the extent that they are effective
are presented in numerous reports
dents on the first-grade reading test
(e.g., Finn & Achilles, 1990;
was 14% below the rate for White
owe their effectiveness to the small
Mosteller, 1995; Word et al., 1990).
students; in small classes, the differ-
number of pupils in a particular class-
Although the study had many
ence was reduced to just 4%.
room, with one or more teachers.
facets, six primary findings stand
To the credit of Project STAR
The third noteworthy study is a
researchers, the students were fol-
large-scale experiment designed to
out:
lowed in subsequent years after
Statistically significant differ-
test the four propositions of Glass and
ences were found among the
Smith and RobinsonTennessee's
returning to regular-size classes;
three class types on every
the continued effort was called
Project STAR (Student-Teacher
Achievement Ratio). In 1985, all
the Lasting Benefits Study (LBS).
achievement measure every year
The carryover effects of small
pupils entering kindergarten in 79
of the study;
schools across the state were
classes continued through at least
The differences were always be-
grade 7' although the small-class
tween the performance of small
assigned at random to a small class
classes and other class types
advantage was diminished to about
(13-17 pupils), a regular class (22-26
not between teacher-aide and
1/6 of a standard deviation on each
pupils), or a regular class with a full-
time teacher aide. Teachers were
regular classes;
test. To date, minority-White differ-
The same benefits were found
also assigned at random to the class
ences have not been examined in the
for boys and girls alike;
groups; no special instructions of any
follow-up study.
In each grade, there was a
sort were given to the teachers or
An important component was
aides. The class groupings were
added to the grade-4 assessment
greater small-class advantage for
that may provide the key to under-
minorities or for students attend-
maintained through third grade, with
assessments of achievement and
standing why a consistent small-
ing inner-city schools on some or
class advantage was found each
all measures;
self-concept administered annually.
Project STAR was built on the
The small-class advantage began
in kindergarten, increased by
principles identified by Glass and
the end of first grade, and
Smith and by Educational Research
' Data collection continued through grade 1 1,
Service. The intervention began in
remained stable through grades
but results for grades 7-11 have not been
the early primary grades; small
2 and 3;
analyzed to date.
Spotlight on Student Success Page 2
0 Temple University Center for Research in Human Development and Education
districts where their effects are
year. In grade 4, teachers were
more protracted interactions with
needed most (e.g., Odden, 1990).
their students. Kiser-Kling (1995)
asked to rate each pupil on the Stu-
And some schools are currently
found that small classes are char-
dent Participation Questionnaire,
a 26-item scale that reflects the
reconsidering their budgets with the
acterized by a greater percentage of
perspective that small classes
effort the student puts forth in the
"task-oriented events" and a smaller
reduce the need for other special-
percentage of "institutional events"
classroom, the initiative the student
takes with regard to learning tasks,
ized services.
(e.g., discipline; organization).
and students' non-participatory
On the benefit side, critics note
WHERE DOES THIS LEAVE
that the size of the effects obtained
behavior (disruptive behavior and
in Project STAR are not large in
US TODAY?
passive/withdrawn behavior). Previ-
ous research had shown that these
Because of the tight experimen-
comparison to other interventions, for
tal control exercised in Project
example, peer tutoring or Success
"engagement behaviors" are essen-
for All. Proponents note that there
STAR, and because of the magni-
tial keys to learning in the classroom
setting, and that poor engagement
is more to Project STAR/LBS than
tude of the study, educators can be
meets the eye. In particular, the
confident of certain principles:
behaviors are especially problematic
benefits associated with small
among students at risk (Finn,
Small classes (17 pupils or below)
classes were realized in every sub-
are more effective academically
Pannozzo, & Voelkl, 1995). The ques-
than larger classes (22 and
ject area in every year of the study;
tionnaire included such behaviors as
the "total" impact is much greater
above) in the primary grades in
paying attention, completing assign-
than any one statistic would reveal.
all subject areas.
ments on time, persisting when con-
The benefits of small classes were
The advantage of small classes
fronted with difficult tasks, and asking
realized by boys and girls alike and
questions to get more information.
is greater for minority students or
students attending inner-city
by minority and White students alike.
The Project STAR follow-up showed
schools than for White students.
Small classes are not "band-aids" for
that students who had been in small
particular students having specific
The advantage of small classes
classes were significantly above their
peers in regular-size classes on all
can probably be attributed to the
problems, but impact all students.
fact that students are more
behavior dimensions.' Small-class
More significantly, small classes
actively engaged in learning
participation increased students'
may reduce the need for other costly
services. For example, the LBS
active engagement in learninga
compared to their peers in larger
grade-4 study documented that
set of constructive behaviors that
classes. It also seems that teach-
ers spend more time in direct
persisted after small classes were
students in small classes were less
disruptive than their peers in regular
instruction and less time in class-
disbanded.
At the same time, Project
classes. Likewise, a small-class
room management when the
initiative in a high-poverty school
number of students is small.
STAR and LBS provided only lim-
ited information about teachers'
district in North Carolina resulted
Questions About Implementing
in reduced disciplinary problems
behavior in small classes. Observa-
Small Classes
tions of Project STAR classrooms
over successive years (Kiser-Kling,
indicated that small classes had
1995). In short, the time and effort
Question 1: Is it worth it?
greater numbers of students "on
needed to control discipline problems
Even the strong findings of
may be reduced. Likewise, there is
task" and that small-class teachers
were more consistent in managing
Project STAR and LBS leave open
some indication from Project STAR
that grade retentions are reduced
the classroom (Evertson & Folger,
the question of whether the benefits
of smaller classes offset the costs.
1989). Other non-STAR studies sup-
On the expenditure side of the
port these conclusions. For example,
Bourke (1986) found that teachers
ledger, economists and others have
2 The Student Participation Questionnaire
of small classes spend less time on
recommended that small classes be
was completed again in grade 8 but, to date,
targeted to those schools and
classroom management and have
the data have not been analyzed.
Spotlight on Student Success
Page 3
© Temple University Center for Research in Human Development and Education
feature is essential. Second, the
when students are in small classes
class: What do teachers do differ-
teachers were all fully certified
(Harvey, 1993). Finally, there is the
ently? Paper presented at the meet-
ing of the American Educational
elementary teachers in Tennessee,
real possibility, though undocumented,
Research Association, San Francisco.
that the need for special education
qualified for the positions they were
Finn, J. D., & Achilles, C. M. (1990).
and other remedial programs (for
given. It is an open question whether
Answers and questions about class
example, Reading Recovery) would
less qualified staff members could
size: A statewide experiment.
counteract the positive effects of
be reduced. Small classes are a fun-
American Educational Research
small classes. Third, no regular
damental ingredient of all special
Journal, 27, 557-577.
education programs, not to mention
school resources were "traded in"
Finn, J. D., Pannozzo, G. M., &
other interventions to serve students
for the small classes (e.g., libraries
Voelkl, K. E. (1995). Disruptive and
having academic difficulty. Unfortu-
or bilingual programs).
inattentive-withdrawn behavior and
nately, thorough analyses of these
achievement among fourth graders.
CONCLUDING REMARKS
and other trade-offs have not been
The Elementary School Journal,
This report emphasizes re-
performed to date.
95, 421-434.
search on the effects of small
Glass, G. V, & Smith, M. L. (1978).
Meta-analysis of research on the
Question 2: What conditions are
classes in the early grades. There
relationship of class size and
necessary for positive outcomes to
are two reasons for this. First, most
achievement. San Francisco: Far
be realized?
of the current (and best) research
West Laboratory for Educational
to date has been conducted in kin-
At present, about half the states
Research and Development.
in the U.S. are considering or imple-
dergarten through grade 3. Second,
Harvey, B. (1993, December). An
menting small-class initiatives for
the early years lay the foundation for
analysis of grade retention for
some or all of their school districts.
much that follows. At the same time,
pupils in K-3. Unpublished doctoral
research to date does not clearly
The questions that will be raised about
dissertation. University of North
implementation are uncountable.
delineate the conditions and conse-
Carolina, Greensboro.
For example, there is little or no
quences of small class size in the
Kiser-Kling, K. (1995). Life in a small
information available to tell us
middle- or high-school grades. This
teacher-pupil ratio class. Unpub-
whether still greater benefits would
work has yet to be undertaken.
lished Ed.D. Dissertation. Univer-
sity of North Carolina, Greensboro.
accrue if teachers received special
Mosteller, F. (1995). The Tennessee
training to take advantage of the
REFERENCES
study of class size in the early
small-class setting. And there is little
Boozer, M., & Rouse, C. (1995, June).
school grades. The Future of Chil-
Intraschool variation in class size:
or no information to tell us whether
dren, 5(2), 113-127.
Patterns and implications. (Work-
small classes might be especially
Odden, A. (1990). Class size and
ing paper No. 344). Washington,
efficacious if used in conjunction
student achievement: Research-
DC: National Bureau of Economic
with other compensatory or accel-
based policy alternatives. Educa-
Research, Industrial Relations
erated programs.
tional Evaluation and Policy
Section.
At the same time, the Project
Analysis, 12, 213-227.
Bourke, S. (1986). How smaller is
STAR/LBS studies were conducted
Robinson, G.E. (1990). Synthesis of
better: Some relationships between
research on effects of class size.
under conditions that, if disregarded,
class size, teaching practices and
Educational Leadership, 47(7),
are likely to counteract some or all
student achievement. American
80-90.
of the small-class advantage. First,
Educational Research Journal, 23,
Word, E., Johnston, J., Bain, H., Fulton,
the classes were comprised of a
558-571.
D. B., Boyd-Zaharias, J., Lintz, M.
small number of students (12-17)
Epstein, I. (1976). Hebrew-English
N., Achilles, C. M., Folger, J., &
translation of the Babylonian
with a single teacher in a single class-
Breda, C. (1990). Student-Teacher
room. This is not simply a ratio of
Talmud (Baba Bathra, Vol. I).
Ratio
Achievement
(STAR):
London: Soncino Press.
15 to 1 (that might be attained with
Tennessee's K-3 class-size study.
Evertson, C. M., & Folger, J. K.
30 students and 2 teachers) but truly
Nashville, TN: Tennessee State
a small class environment; this
(1989, March). Small class, large
Department of Education.
Spotlight on Student Success
Page 4
© Temple University Center for Research in Human Development and Education
6
E KLC
U.S. Department of Education
Office of Educational Research and Improvement (OERI)
MOW Imo Woman CM
National Library of Education (NLE)
Educational Resources Information Center (ERIC)
NOTICE
Reproduction Basis
This document is covered by a signed "Reproduction Release
(Blanket)" form (on file within the ERIC system), encompassing all
or classes of documents from its source organization and, therefore,
does not require a "Specific Document" Release form.
This document is Federally-funded, or carries its own permission to
reproduce, or is otherwise in the public domain and, therefore, may
be reproduced by ERIC without a signed Reproduction Release form
(either "Specific Document" or "Blanket").
EFF-089 (5/2002)