Transcript Hobbs empowerment_presentation_drk12_2009

Growth of Empowerment
in Career Science Teachers Project Instrument Development
Presented at the DRK-12 P.I. Meeting, 2009
Dr. Mary Hobbs, P.I.
Washington, D.C.
Amy Moreland, Graduate Research Assistant
The University of Texas at Austin
Problem
• Much of the recent education research has
focused on new teachers and why large numbers
of teachers are leaving the profession (Ingersoll &
Smith, 2003).
• Few studies have focused on experienced
teachers in an attempt to identify factors that may
have contributed to their professional growth and
retention.
Purpose
To develop an instrument to identify those
pivotal experiences of career science
teachers that have promoted their
advancement along the teacher professional
continuum and have helped them to persist in
their careers
Plan
•
The researchers in this study used techniques
of narrative inquiry, as well as behavior over
time (BOT) graphing, to capture the
experiences fifty-two teachers identified as
having positively or negatively impacted their
feelings of empowerment.
•
The researchers created a multi-domain, online
survey instrument to collect statewide and
national data:
The Online Teacher Empowerment Survey.
Project timeline Years 1 - 2
Project timeline Years 3 - 4
Perspective
•
What is empowerment and why is it important?
•
Empowerment is most often viewed as a
process through which people become
powerful enough to engage in, share control of,
and influence events and institutions affecting
their lives.
•
Short (1992) presented six empirically derived
dimensions underlying the construct of teacher
empowerment – autonomy, self-efficacy,
decision-making, status, impact, and
professional growth.
Methods
 In this study, data collection was accomplished
through the use of one-on-one interviews (Clandinin
& Connelly, 2000) in conjunction with the Systems
Dynamics (Anderson & Johnson, 1997) technique of
utilizing behavior-over-time graphing (Year 1).
Sample Behavior-Over-Time Empowerment Graph
from Nancy, a 17-year science teacher
Methods
• The researchers categorized 52 science teacher
experiences, identified patterns, and examined the
data for congruence with the growth of
empowerment model developed by Hobbs (2004)
during a previous study
(Year 2).
Hobbs’
Empowerment
Model
• Phase 1: Initiating
Empowerment
(years 0-3)
• Phase 2: Growth
of Empowerment
(years 4-8)
• Phase 3:
Design adapted from Senge’s The Dance of Change.
Sustaining
Empowerment
(9+ years)
Methods
• The researchers then developed the survey
questions as based on these 52 teacher
empowerment interviews, congruence with the
empowerment model, and regional focus groups
(Year 3).
Construct Validity Map
Data Collection
• After many design iterations, as informed by teachers, focus
groups, external consulting, and peer feedback, the
researchers created a multi-domain, online survey instrument
to collect statewide and national data in more efficient ways:
The Online Teacher Empowerment Survey
(Years 3 - 4).
Snowball Data Collection Timeline (to date) of the Online Teacher Empowerment
Survey
The Online Teacher Empowerment Survey
Online Teacher Empowerment Survey Example Question 5: Professional Development
With the definition of professional growth in mind, which of
the following best describes your current sense of
Professional Growth? (Choose one) [N=252]
48%: “Teaching is a
profession and professions
require continuing
education.”
46%: “I’ve always had a big
commitment to professional
development - I think that’s
just been a big consistent
for me .”
Data Analyses Methods
• Upon receiving the first 250 initial respondents (requirement of this
grant), the project’s statistician ran a Rasch rating scale program to
access the psychometric properties of the survey. It is important to
create linear continuous variables from ordinal responses when using
surveys. Too often, researchers use ordinal data with parametric
procedures to analysis data (analysis of variance; t-test, etc.).
Parametric analyses were intended for use with means and standard
deviations that meet statistical analysis assumptions.
• In this study, a linear continuous variable was created
(Empowerment) that met homogeneity of variance, skewness,
kurtosis, missing data, and outlier assumptions.
• The descriptive statistics indicated the desired range of 0 to 100 for
the Empowerment scaled score. The mean and standard deviation
indicate that Empowerment scores at the 95% confidence level
(mean +/- 1.96 * standard deviation) would fall between 10 and 82.
Empowerment Score Interpretation
•
The Empowerment scores were comprised of 7 items which had
an internal consistency of .99. This implies a very good construct
scale. The person reliability coefficient was .41 indicating that
teachers did not respond consistently on the scale, which is
expected when a construct scale indicates different levels of the
construct.
•
The Empowerment score did not differ significantly by most of the
demographics for this sample (i.e. gender & ethnicity). However,
this does not invalidate the Empowerment scale. The
Empowerment scores do have sufficient range to indicate more or
less of the construct Empowerment.
•
The Empowerment scale can be used as a diagnostic measure to
indicate whether a teacher has low, average, or high self
empowerment.
Preliminary Results – Trends
•
Although the mean Empowerment score indicated
an increasing trend from 38.19 to 48.87 across
the teaching groups, the F = 2.138, p = .096
indicated there was no statistically significant
mean difference between the years of teaching
groups at the .05 level of significance.
Preliminary Results – Trends
•
A visual inspection suggests that PK to Elementary
Teachers had higher Empowerment scores, so
basically the trend is that Empowerment decreases as
grade level increases. However, this trend was not
statistically significant (One way ANOVA of
Empowerment Score: F = 2.047, p = .131).
Preliminary Results – Trends
•
The Empowerment scores for the Rural school districts
were on average 5.5 points lower than the Urban and
Suburban school districts. However, the mean
differences were not statistically significant (One way
ANOVA of Empowerment ScoreF = 2.29, p = .10) at the
.05 level of significance.
Preliminary Results – Trends
•
A visual inspection suggests that traditionally certified
teachers on average had higher empowerment scores
(46.98 versus 44.38), however, this was not statistically
significant (t value = 1.07, p = 0.285).
Preliminary Results – Significant Trends
•
Teachers who have participated in a community of learners
program where they received 60+ hours of professional
development had a statistically significant higher average
Empowerment score than Teachers who did not (t = 2.138, p =
.033) and scored on average 5 points higher. In this analysis,
the number of professional development hours did matter.
Summary
•
Empowerment scores on average were
statistically different for Teachers who received
60+ hours of professional development training
versus those who did not.
•
The Empowerment score differences for the
other demographic variables are in the direction
one would expect, although for this sample of
teachers the differences were not of sufficient
magnitude to indicate statistical significance at
the .05 level of significance.
•
Another sample of teachers might well indicate
differences in average Empowerment scores for
the other demographic variables used in this
analysis.
Further Research
•
Years of teaching experience and number of PD hours
if left as continuous variables could better indicate a
“positive” relationship with Empowerment scores.
Basically, a statistically significant correlation coefficient
would indicate that as number of PD hours and years of
teaching increases, the Empowerment score increases;
similar interpretation for number of hours of professional
development.
•
The Empowerment scores can be used as a diagnostic
measure to indicate which teachers have lower than
average levels of self empowerment. The distribution of
Empowerment scores can be considered normally
distributed. As indicated by the quartile distribution,
scores in the range 0 to 34 have low self empowerment;
scores in the range 35 to 60 have average self
empowerment; and finally scores in the range 61 to 100
have high self empowerment.
Researchers Contact Information
Mary E. Hobbs, Ph.D.(Co-P.I.)
Coordinator for Science Initiatives
Texas Regional Collaboratives
The University of Texas at Austin
[email protected]
James P. Barufaldi, Ph.D.(Co-P.I.)
Amy L. Moreland, M.S.
Principal Investigator
Graduate Research Assistant
Texas Regional Collaboratives
The University of Texas at Austin
The University of Texas at Austin
[email protected]
[email protected]