Investigating the Role of Human Resources in School Turnaround:

hardtofindcurtainΠολεοδομικά Έργα

16 Νοε 2013 (πριν από 3 χρόνια και 11 μήνες)

81 εμφανίσεις

Investigating the Role of Human
Resources in School Turnaround:


A
Decomposition of Improving Schools in Two States








Acknowledgements: This research draws upon work performed under contract with the
Institute of Education Sciences (ED
-
04
-
CO
-
0025/0020). This work does not necessarily
represent the views of any affiliated institutions, and any and all errors are mine.

Michael Hansen

CALDER at the
American Institutes for Research


6
th

Annual
CALDER Conference

February 21, 2013

Washington, DC

2

The presumed role of human
resources

in turnaround


Turnaround, transformation
models


Prescribe principal and/or teacher turnover


Teacher and principal quality are
most consequential schooling inputs


Assume teacher/principal quality are static





3

Workforce turnover or human
capital development?


Which of the two models dominates in
past turnaround schools?


Results:


Evidence of elements of both models playing a
role


Strong improvements among stable teachers


Strong incoming teachers, no evidence on weak
outgoing teachers





Longitudinal Data Sources

Florida



Math FCAT
-
SSS


Student
-
teacher
linked


Spans 2002
-
03 to
2007
-
08 years





North Carolina



Math EOG tests


Student
-
teacher
linked


Spans 2002
-
03 to
2007
-
08 years


Principals

4

How is School Performance
Identified?

Time Span of Observation Window

Monitoring Period

Baseline Period

CLPs

TA

MI

NI

Descriptive Means of the Sample

of Low
-
performing Schools

6

State

Florida

North Carolina

School sample

Elementary

Middle

Elementary

Middle

Proportion of African American
students

52.6%

40.4%

55.5%

62.3%

Proportion of Hispanic students

21.6%

31.6%

12.9%

9.2%

Proportion of students ever eligible
for free or reduced
-
price lunch
program

89.8%

83.2%

74.4%

69.5%

Mean Student Achievement in Math

-
0.37

-
0.11

-
0.46

-
0.33

Unique CLP, Non
-
TA Schools

87

22

66

37

Unique CLP, TA Schools

17

3

8

5

Total student
-
year observations

43,553

15,398

37,371

24,505

Note: Samples limited to student
-
teacher linked observations in math in chronically
low
-
performing schools identified in Hansen and Choi (2012) using the 2005
turnaround point in math.

7

Decomposing Performance
Improvements across Workforce


Pre
-

vs. post
-
period


Turnaround (TA) vs. non
-
TA


3 types of teachers in workforce:


Outgoing


Stable


Incoming





Identifying Teacher Groups
Contributing to Performance

8

Outgoing

Incoming

Stable

Pre
-
period

Post
-
period

What Workforce Dynamics
Turnaround Schools?

-1.5
-1
-0.5
0
0.5
1
1.5
2003
2005
2007
Group Value Added

Year

Workforce Turnover Model

outgoing
stable
incoming
What Workforce Dynamics
Turnaround Schools?

-1.5
-1
-0.5
0
0.5
1
2003
2005
2007
Group Value Added

Year

Human Capital
Development
Model

outgoing
stable
incoming
Teachers: Evidence Suggestive

of Human Capital Development

11

State

Florida

North Carolina

School sample

Elem.

Middle

Elem.

Middle

TA*Post (
β
5
)

0.139**

0.153**

0.187**

0.092**

Outgoing*TA (
β
7
)

0.027

0.031

0.019

-
0.026

Incoming*TA*Post
(
β
9
)

0.022

-
0.010

-
0.063*

0.014







Observations

43,553

15,398

37,371

24,505

R
-
squared

0.575

0.623

0.644

0.681

Observed performance in

NC Schools

-0.15
-0.1
-0.05
0
0.05
0.1
2003
2004
2005
2006
2007
2008
Group Value Added

Year

NC Elementary Schools

outgoing
stable
incoming
Observed performance in

NC Schools

-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
2003
2004
2005
2006
2007
2008
Group Value Added

Year

NC Middle Schools

outgoing
stable
incoming
Principals: Similar Evidence of

Human Capital Development

14

School sample

NC Elem.

NC Middle

TA*Post (
β
5
)

0.154**

0.065**

Outgoing*TA (
β
7
)

0.013

-
0.000

Incoming*TA*Post (
β
9
)

-
0.005

0.034





Observations

39,394

37,353

R
-
squared

0.640

0.682

15

Results are Robust to

Alternative Specifications


Are these results sensitive to:


How teacher groups are categorized?


How TA schools identified?



No qualitative changes to the
estimated relationships





16

Same Patterns of Improvement
Observed in Other Schools?


What about middling schools with low
growth? How do they improve?


Replicate
identification and estimation in
schools that have higher levels of status,
but quick improvement in school growth


Improvement
of stable teachers most
prominent in elementary schools;
turnover in middle schools


17

Summary of Findings


Results show strong, robust gains
associated with stable teachers


Evidence of high
-
performing
incoming teachers, but not outgoing


Does not necessarily vindicate either
of two workforce models, but
suggests mix or spillover





18

Important Study Limitations


Descriptive investigation of outlier
schools


Not causal or representative


Improvements are absorbed into
staff,
though other interventions may be at
work


Not an evaluation of specific treatment;
not predictive of current
efforts





19

Policy Implications


Current policy emphasizes human
capital turnover


Best use of intervention efforts?


Can these successes be replicated?


Feeds into larger debate about
teacher quality


Costs of improvement vs. replacement


Individual or context
-
specific effectiveness