P R O F E S S I O N A L D E V E L O P M E N T T H A T W O R K S
What Makes Professional
Development Effective?
After an examination of 13 recent lists of
characteristics of “effective professional
development,” Mr. Guskey concludes that they
vary widely and that the research that supports
them is inconsistent and often contradictory.
Nonetheless, he argues, we need to seek
agreement on criteria for effectiveness, along
with clear descriptions of contextual factors.
BY THOMAS R. GUSKEY
O WE KNOW what makes professional development effective?
D
Have researchers and practitioners reached consensus about
what factors contribute to a successful professional development
experience? Do we even agree on what criteria should be used to
judge professional development’s effectiveness? A review of
newly developed lists of the characteristics of effective
professional development indicates that the answer to each of
these questions is “Maybe not.”
Recently I analyzed 13 different lists of the charac- Development, Education Development Center, Edu-
teristics of effective professional development, all pub- cational Research Service, Educational Testing Serv-
lished within the last decade. These lists were drawn ice, Eisenhower Professional Development Program,
from publications of the American Federation of National Governors’ Association, National Institute
Teachers, Association for Supervision and Curriculum for Science Education, National Partnership for Ex-
cellence and Accountability in Teaching, National
THOMAS R. GUSKEY is a professor of educational policy stud- Staff Development Council, and U.S. Department of
ies and evaluation, College of Education, University of Kentucky, Education. My goal was to find out the extent to which
Lexington. The research reported in this article was presented at
the annual meeting of the American Educational Research As- these various lists agreed. What I discovered is that
sociation, Chicago, 2003. they were derived in very different ways, used differ-
748 PHI DELTA KAPPAN Photo: Liquid Library
Downloaded from pdk.sagepub.com at UNIV OF CONNECTICUT on May 11, 2015
ent criteria to determine “effectiveness,” and varied uals can collaborate to block change or inhibit progress
widely in the characteristics they identified. I also just as easily as they can to enhance the process.2 For
found that the research evidence regarding most of the collaboration to bring its intended benefits it, too, needs
identified characteristics is inconsistent and some- to be structured and purposeful, with efforts guided by
times contradictory. clear goals for improving student learning.
In considering their development, I concluded that Most of the lists stress the inclusion of evaluation
most of the lists could be described as “research-based.” procedures — an emphasis that probably stems from
But that research rarely includes rigorous investigations growing awareness of the need to gather regular form-
of the relationship between the noted characteristics ative information to guide improvement efforts. Most
and improvements in instructional practice or student also emphasize the need for professional development
learning outcomes. Instead, it typically involves surveys activities to be aligned with other reform initiatives
of the opinions of researchers and educators. In other and to model high-quality instruction. These charac-
words, researchers and practitioners generally favor these teristics undoubtedly come from appreciation of the
characteristics and believe they are important, despite large-scale and systemic nature of most modern edu-
the lack of verifying evidence. Only a National Insti- cation reforms and increased awareness of similarities
tute for Science Education (NISE) analysis and an Ed- between the learning patterns of adults and children.
ucational Testing Service (ETS) study show a direct The majority of lists stress that professional devel-
link between their identified characteristics and spe- opment should be school- or site-based, even though
cific measures of student achievement.1 significant research evidence suggests otherwise. A re-
Of the 21 characteristics distinguished in the lists, cent review by the Consortium for Policy Research in
the most frequently cited was enhancement of teach- Education, for example, found that when decisions
ers’ content and pedagogical knowledge. Helping teach- about professional development were primarily school-
ers to understand more deeply the content they teach based, staff members paid only lip service to research
and the ways students learn that content appears to be and were more interested in programs similar to what
a vital dimension of effective professional development. they were already doing than in those producing re-
At present, however, nearly all of the studies relating sults.3 In such instances the decentralization of deci-
this characteristic to improvements in student learning sion making appeared to be undermining the use of
focus on achievement in mathematics or science. Wheth- knowledge rather than promoting it. A carefully or-
er the same is true for achievement in language arts, so- ganized collaboration between site-based educators, who
cial studies, or other subject areas has yet to be thor- are keenly aware of critical contextual characteristics,
oughly investigated. and district-level personnel, who have broader perspec-
Most of the lists mention the provision of sufficient tives on problems, seems essential to optimize the ef-
time and other resources as essential to effective pro- fectiveness of professional development.
fessional development. Obviously, educators need time Despite the current emphasis on student perform-
to deepen their understanding, analyze students’ work, ance, less than half the lists mention the importance
and develop new approaches to instruction. But sig- of using analyses of student learning data to guide pro-
nificant contrary evidence exists. The NISE analysis, fessional development activities. And amazingly, on-
for example, showed that differences in time spent in ly four lists stress that professional development should
professional development activities were unrelated to be based on the best available research evidence. This
improvements in student outcomes. Similarly, the ETS is particularly striking given the long-standing criti-
study found that the amount of time spent in profes- cism of professional development that focuses on fads
sional development was unrelated to achievement. So and bandwagon movements rather than on solid evi-
while effective professional development surely requires dence of what works with students. Other noted char-
time, it’s clear that the time must be well organized, acteristics appear on only one or two lists.
carefully structured, and purposefully directed. This analysis led me to three related conclusions.
Another consistently noted characteristic is the pro- First, there appears to be little agreement among pro-
motion of collegiality and collaborative exchange. Ed- fessional development researchers or practitioners about
ucators at all levels value opportunities to work togeth- the criteria for “effectiveness.” Some define it in terms
er, reflect on their practices, exchange ideas, and share of teachers’ self-reports of professional development fea-
strategies. But research on teachers shows that individ- tures that increase their knowledge and lead to changes
JUNE 2003 749
Downloaded from pdk.sagepub.com at UNIV OF CONNECTICUT on May 11, 2015
in their instructional practices. Others look for consen- Take, for example, professional development specif-
sus in the opinions of professional development writ- ically designed to enhance teachers’ content and peda-
ers and researchers. Such efforts, while informative, rep- gogical knowledge. Schools in economically depressed
resent only a starting point. areas that have trouble attracting and keeping well-
To gain authentic evidence and make serious im- qualified teachers and, as a result, have many teachers
provements, we need to push beyond this starting point teaching subjects outside their area of certification, may
and move toward professional development’s ultimate benefit greatly from such programs. Schools in more
goal: improvements in student learning outcomes. These affluent communities, on the other hand, that have suf-
outcomes should be broadly defined to include a va- ficient resources to attract and retain well-qualified teach-
riety of indicators of student achievement, such as as- ers with advanced training in the subject areas they teach
sessment results, portfolio evaluations, marks or grades, may see little improvement from similar programs. These
or scores from standardized examinations. They might real-world contextual differences profoundly influence
also include affective and behavioral outcomes, such as the effectiveness of professional development endeavors.
students’ attitudes, attendance rates, dropout statistics, Third, these results show that while the promise of
and participation in school activities. Significant advances research-based decision making in professional devel-
in professional development will come only when both opment has been largely unfulfilled, it need not remain
researchers and practitioners insist on this fundamental so. Analyses of student learning data typically show
goal as the principal criterion of effectiveness. that greater variation exists between classrooms with-
Second, most of the currently identified character- in a school than between schools or between districts.4
istics of effective professional development seem best In other words, within the unique context of nearly
described as “yes, but . . .” statements. For example, every school there are teachers who have found ways
yes, enhancing teachers’ content and pedagogical knowl- to help students learn well. Identifying the practices
edge is important, but existing research is limited main- and strategies of these teachers and sharing them with
ly to investigations of mathematics and science instruc- their colleagues might provide a basis for highly effec-
tion. Yes, professional development should provide suf- tive professional development within that context.
ficient time and resources, but such time and resources The characteristics that influence the effectiveness
must be used wisely, focusing on activities that posi- of professional development are multiple and highly
tively affect learning and learners. Yes, professional de- complex. It may be unreasonable, therefore, to assume
velopment should include procedures for evaluation, that a single list of characteristics leading to broad-
but evaluations that focus narrowly on educators’ self- brush policies and guidelines for effective profession-
reported satisfaction with professional development ac- al development will ever emerge, regardless of the qual-
tivities offer inadequate guidance and direction to im- ity of professional development research. Still, by agree-
provement efforts. And so on. ing on the criteria for “effectiveness” and providing
The problem with such “yes, but” statements is that clear descriptions of important contextual elements,
they frustrate policy makers and practitioners who want we can guarantee sure and steady progress in our ef-
simple answers to their questions about effective pro- forts to improve the quality of professional development
fessional development. They also diminish the value endeavors.
of “research” evidence in the eyes of those seeking un-
ambiguous statements about “best practices” in pro- 1. Mary M. Kennedy, Form and Substance in Inservice Teacher Educa-
fessional development. Nevertheless, they accurately tion (Madison: National Institute for Science Education, University of
Wisconsin, Research Monograph No. 13, 1998); and Harold Wenglin-
represent the fact that nearly all professional develop- sky, “How Schools Matter: The Link Between Teacher Classroom Prac-
ment takes place in real-world contexts. The complex- tices and Student Academic Performance,” Education Policy Analysis Ar-
ities of these varied contexts introduce a web of fac- chives, vol. 10, no. 12, 2002, available at http://epaa.asu.edu/epaa/v10n12.
2. Judith Warren Little, “The Persistence of Privacy: Autonomy and Ini-
tors that influence whether or not a particular charac- tiative in Teachers’ Professional Relations,” paper presented at the annual
teristic or practice will produce the desired results. These meeting of the American Educational Research Association, San Fran-
nuances of context are difficult to recognize and even cisco, 1989.
more difficult to take into account within the confines 3. Thomas B. Corcoran, Susan H. Fuhrman, and Catherine L. Belcher,
“The District Role in Instructional Improvement,” Phi Delta Kappan,
of a single program. Thus programs that appear quite September 2001, pp. 78-84.
similar may produce different results for subtle and 4. Edward Kifer, Large-Scale Assessment (Thousand Oaks, Calif.: Cor-
unanticipated reasons. win Press, 2001). K
750 PHI DELTA KAPPAN
Downloaded from pdk.sagepub.com at UNIV OF CONNECTICUT on May 11, 2015
