ASSESSING TEACHER KNOWLEDGE:
USING RESEARCH TO INFORM LEGISLATION AND POLICY
written by Murray
Mitchell, University of South Carolina
"There are three kinds of lies: Lies,
Damned Lies, and Statistics" - Benjamin Disraeli (n.d.)
Assessment
is a high stakes enterprise in education because the consequences
of collecting and interpreting data can change lives. For
students, assessment can make the difference between passing
and failing, or admission to or rejection from a college,
university, or other post-high school program. For teachers,
principals, and school district superintendents, assessment
can result in salary adjustments and keeping or losing a job.
Because of the high stakes involved, many fear what statistics
may be used to inform important decisions, including the creation
of legislation and to make policies.
The assessment of teacher performance is a current issue
in many states. For example, there is a bill in the South
Carolina legislature (House Bill 3716) designed in part to
allow school districts to decide how to provide raises for
teachers, and directing the Department of Education to develop
an "incentive compensation system based on teacher performance"
(South Carolina Legislature, 2011). A concern for any teacher
must be tied to exactly how teacher performance will be assessed.
Student Outcomes as a Measure of Teacher Performance
Examining some measure of student outcome data is one possible
approach to measuring teacher performance. Most teachers working
in affluent schools with no behavior issues, lots of resources,
and a supportive parent/teacher environment might be attracted
to this kind of proposal. Teachers working in challenged environments,
with sparse resources, a transient student population, minimal
contact with parents and guardians, violence and substance
abuse issues, might be less attracted to this kind of proposal.
In spite of these kinds of discrepancies, there is evidence
to support the notion that really good teachers can teach
regardless of these kinds of obstacles, and that bad teachers
still can't teach even with all the support in the world (c.f.,
Rink & Stewart, 2003).
Content Knowledge as A Measure of Teacher Performance
Another approach to assessing teacher performance is to examine
their content knowledge; but, this is not as simple as it
might seem. For example, the health-related fitness (HRF)
content knowledge, using samples of middle school physical
education teachers, inservice teachers, and preservice teachers,
has been found to be less than optimal on artificial paper
and pencil assessments when measuring content knowledge (Ayers,
2002; Castelli & Williams, 2007; Miller & Housner,
1998). In contrast, Mitchell (2010) used an alternative assessment
strategy to assess what he described as "task-relevant
knowledge" (p. A-67) and found it to be far more flattering.
Rather than an out-of-context paper and pencil assessment,
Mitchell analyzed the actual contracts submitted by 2,828
students to 126 teachers.
Contracts were content analyzed in a 6-step process:
- Step 1, Identify the type of activity indicated on the
contract, to ensure that the activity would contribute to
a health-related fitness variable;
- Step 2, Identify the health-related
fitness variables, looking for two details: (1) the HRF
variable matches the activity; and (2) the variable identified
is one of the accepted HRF variables (cardiovascular endurance,
muscular endurance, muscular strength, flexibility and/or
body composition);
- Step 3, Confirm that the dates
indicated span a minimum of 6 weeks;
- Step 4, Confirm that a minimum
of 20 minutes per episode is indicated;
- Step 5, Confirm that a minimum
of 3 episodes per week is indicated; and
- Step 6, Identify if the teacher
has indicated whether or not the information presented on
the contract is or is not appropriate.
In Mitchell’s (2010) study, the mean percent of correctly
accepted contracts was 80.22% (SD=25.2: Median =88.35). There
were no incorrectly rejected contracts. This is a much different
success rate or indicator of teacher performance than was
evidenced in the other studies. See Table 1 for a contrast
of these assessments of teacher performance.
Table 1: Comparison of Physical Educator Competence
Scores
Study |
Competence
Score |
Mitchell
(2010) |
80.22% |
Castelli
& Williams (2007) |
63.93% |
Ayers
(2002) |
48-64% |
Miller
& Housner (1988) |
5.22% |
Conceptual Clarity
Content knowledge is an integral part of effective teaching.
Identifying what knowledge is necessary for effective instruction,
how thoroughly that knowledge must be "known," and
how to measure content knowledge are disputed in the literature.
Anderson (1995) makes a distinction between declarative knowledge
(descriptions of facts) and procedural knowledge (ability
to "do").
Landy (1974) made a comparable distinction when he sought
objective measures for performance obsolescence in work with
engineers. Kelley and Lindsay (1977) tried to make a similar
distinction between knowledge obsolescence and performance
obsolescence when examining physical educators, when they
noted poor performance on a paper and pencil test does not
necessarily mean that "physical educators as a group
do not have performance capabilities that enable them to fulfill
effectively their responsibilities" (p. 473).
In recent work on physical educators, paper and pencil tests
have been used to examine content knowledge. Inservice and
preservice teachers have been found to be wanting. Such findings
have been used to castigate the professionalism of many teachers,
and to identify poor content knowledge as a possible contributing
factor to the health crisis facing our nation.
In the Mitchell (2010) study, an alternative approach was
used to examine knowledge in the context of a real (or authentic)
performance task for which teachers are responsible - grading
student contracts. Correctly grading the contracts requires
knowledge of the HRF content for teachers to make professionally
appropriate decisions. Using this authentic approach, much
higher competence scores are reported for teachers.
Methods for measuring performance require more theoretical
clarity than has been evident in recent studies. It is possible
that measurement error, rather than incompetence, may explain
some previous findings assailing physical education teacher
reputations. So, in preparation for the predictable needs
for more effective legislation and policy decisions, it will
be important to design research studies that can help to ensure
the best teachers are in classrooms, and that they are appropriately
rewarded. Equally important will be the need for careful interpretation
of data to inform important decisions in this high stakes
assessment environment.
click for
references
Murray Mitchell is an Associate Professor and Graduate Director
for the Department of Physical Education and Athletic Training
at the University of South Carolina. He also serves as the
High School Assessment Director for the South Carolina Physical
Education Assessment Program. Mitchell is also a Fellow in
the Research Consortium of AAHPERD and a Fellow in the South
Carolina Education Policy Fellowship Program, sponsored by
the Institute for Educational Leadership.
(back to pelinks4u homepage) |