The Measurement of Teacher Efficacy

1
The Measurement of Teacher Efficacy
Jessica Robyn Koehler, Ed.S., NCSP Department of
Counseling and Personnel Services, University of
Maryland

• Introduction
• Background
• In 1976, the RAND corporation initiated the
  investigation of teacher efficacy by asking
  teachers two questions, one measuring personal
  teaching efficacy (PTE) and the other general
  teaching efficacy (GTE). Personal teaching
  efficacy refers to teachers beliefs in their
  personal teaching skills, while general teaching
  efficacy is belief in the power of teaching to
  overcome external student factors.
• Since the RAND study, various researchers have
  investigated teacher efficacy and developed a
  variety of longer measures to tap this construct.
  Based on three decades of research, teacher
  efficacy has proven to be an important variable
  related to educational outcomes (e.g., student
  achievement, reduced stress among teachers, and
  teachers willingness to stay in the field) .
• While much has been learned about teacher
  efficacy, the measurement of this construct has
  been problematic due to the unstable psychometric
  properties of the instruments used. Both
  conceptual and statistical concerns regarding the
  most commonly used teacher efficacy measure,
  Gibson and Dembos (1984) Teacher Efficacy Scale
  (TES), were reported.
• In response to the questions regarding the TES,
  Tschannen-Moran and Hoy (2001) developed the
  Teacher Sense of Efficacy Scale (TSES). Several
  studies indicate that the TSES has a unified and
  stable factor structure that measures the
  following efficacy for instructional strategies,
  efficacy for classroom management, and efficacy
  for student engagement. The three factors
  structure of the TSES provides useful information
  regarding teachers sense of personal teaching
  efficacy, but general teaching efficacy is not
  measured.
• Focus of Poster
• An additional set of questions, intended to
  measure the GTE construct, has been created and
  added to the TSES (short-form) as part of an
  Institute of Education Sciences (IES) funded,
  randomized control trial of Instructional
  Consultation Teams (An Experimental Study of the
  Effectiveness of Instructional Consultation
  Teams). These new questions aim to evaluate
  teachers beliefs in their abilities to overcome
  student factors, such as students family
  background, previous education, and special
  education status.
• The factor and internal consistency item analysis
  results and mean scores from the TSES with the
  new additional questions related to general
  teaching efficacy are reported.

• Results
• Mean Scores on Scales
• Mean total efficacy and factor scores were
  calculated for participants and compared to two
  demographic variables years of teaching
  experience and level of education.

• Methods
• Procedure
• 1922 teachers in 45 elementary schools in a large
  suburban school district were asked to complete a
  large survey that included the teacher efficacy
  questions. A link to this web-based survey was
  delivered to the participants via email. At the
  time of this poster, the response rate was 86.

• Results
• Factor analysis and reliabilities
• A calibration and validation sample were
  (randomly) derived from the data set in order to
  confirm the analyses. Because analyses on the
  two subsets of the data were similar, the
  calibration sample results are described on this
  poster.
• Principle-axis factoring with varimax rotation of
  the 20 items yielded four factors with
  eigenvalues over one, accounting for 52 of the
  total variance in the respondents scores. A
  scree test appeared to show that four factors
  could be extracted. Rotation converged after
  seven iterations.
• Factor analysis results suggested that two of the
  three TSES scales (Efficacy for Instructional
  Strategies and Efficacy for Classroom Management)
  were intact. Four of the six new questions
  loaded with three of the TSES student engagement
  questions. Two new questions related to efficacy
  for teaching special education loaded separately
  with an original instructional strategies
  question.

• Conclusions and Future Research
• Results suggest that the Teacher Sense of
  Efficacy Scale with the additional six questions
  yields a four factor solution accounting for 52
  of the total variance. The full scale internal
  consistency was .93 and the factor internal
  consistency scores ranged from .76 to .87.
• Means analysis results indicate differences
  between total efficacy and individual factor
  scores based on years of teaching experience and
  level of education.
• Future Directions
• Evaluation of the teacher efficacy scores in
  relation to the teacher demographic variables,
  school variables, and student achievement within
  a multi-level model.
• Evaluation of changes on the teacher efficacy
  instrument in treatment versus control schools
  will be analyzed over a four year period as part
  of the randomized control trial.

• Methods
• Participants
• 1922 elementary school teachers from a large,
  suburban school district
• Gender 92.5 female, 7.5 male
• Race 88.8 White, 7.2 Black or African
  American, lt1 American Indian or Alaska Native,
  1.1 Asian, lt1 Native Hawaiian or Pacific
  Islander, 2.6 Other
• Ethnicity 96.2 Not Hispanic or Latino, 3.8
  Hispanic or Latino
• Primary Job Responsibility 59.8 General
  Classroom Teacher, 11.2 Special Education
  Teacher, 4.4 Reading Specialist, 1.3 Math
  Specialist, 5.7 ESOL Teacher, 11 Specialist,
  6.7 Other
• Level of Education 10.8 Bachelors Degree,
  37.7 Bachelors Degree , 21.2 Masters Degree,
  29.7 Masters Degree , lt1 Doctorate
• Years of Teaching Experience 8.6 1 year or
  less, 26 2 to 5 Years, 21 6 to 10 years, 21.8
  11 to 20 Years, 22.6 More than 20 years
• Measures
• The Teacher Sense of Efficacy Scale
  (Tschannen-Moran and Hoy, 2001) with six
  additional questions was administered.
• Response category values for the survey ranged
  from 1 to 5 on a Likert scale
• 1Nothing/Not At All 2Very Little 3Some
  4Quite A Bit 5 A Great Deal

References Tschannen-Moran, M., Hoy, A. W.
(2001). Teacher efficacy capturing an
elusive construct. Teaching and Teacher
Education, 17, 783-805. Gibson, S., Dembo, M.
H. (1984). Teacher efficacy a construct
validation. Journal of Educational Psychology,
76, 569-582.
Acknowledgements This research was supported by
an Institute of Education Sciences grant awarded
to Sylvia Rosenfield, Ph.D. and Gary
Gottfredson., PhD. The author would also like to
acknowledge the contributions of Gary
Gottfredson, Ph.D., Sylvia Rosenfield, Ph.D., and
Amy Silverman, Ph.D.
For further information Please contact Jessica
Koehler at jes17_at_umd.edu.