Using Data for Decision Support and Planning

1
Using DataforDecision SupportandPlanning
2
Presenters

• Dr. Marie Loftus
• Dr. Barbara R. Sadowski
• Marywood University
• www.marywood.edu

3
Session Overview

• Group activity followed by discussion
• Starting the research process
• Break
• Group activity questionnaire analysis and
  discussion
• Research results
• Question and answer

4
Group Task 1

• Develop 2-3 questions of interest
• For each question list
• a) what data could be used
• b) factors that might affect the process of
  obtaining data
• _ c) potential problems with existing data
• d) environmental factors that might bias the
  data or interpretation of the results
• e) relevant existing data sources

5
Research Purpose

• To determine if a teachers Thinking Math
  training has an effect on student achievement in
  Mathematics.

6
Questions Considered

• Does the mathematics training of teachers have an
  effect on the math achievement of students they
  teach?
• Does professional development training of
  teachers translate into higher student
  achievement?

7
Practicality of Research Options

• Question 1
• Difficult to get data on teachers background in
  mathematics
• Measurement difficulty in defining college course
  work with in-service professional development
• Question 2
• Teacher training data available
• Non-transient teacher population
• Well-documented training program
• Student achievement data available

8
Other Variables Considered

• What constituted teacher training? How many TM
  workshops?
• Did TM trained teachers actually implement the
  principles? Which ones were implemented? How
  confident did the teachers feel about their
  ability to implement?
• Was years of teaching experience a factor in how
  well TM was implemented?

9
Other Variables Considered

• Was number of years teaching at grade level a
  factor?
• Was the math background of the teacher a factor
  in how well TM was implemented?
• How much of the time did the teacher actually
  implement TM when teaching math?

10
Other Variables Considered

• What student variables might affect the score on
  the achievement test? Reading level? SES? Title
  1 support? The number of TM trained teachers they
  had before reaching fifth grade?
• Would school climate affect how the teachers felt
  about implementing TM math principles? Did the
  principal support their use of TM?
• Ability grouping in the assignment of students

11
Research Question 1

• Is the number of years students have a TM
  trained math teacher a predictor of how they will
  perform on a math achievement test?

12
Scranton PA School District

• Total enrollment 8,673 students
• Economically disadvantaged 54
• Mobility 25.2
• Schools 18
• 13 elementary (Pre K-5)
• Teachers 712
• strong collective bargaining history with the
  Scranton Federation of Teachers

13
Data Collection

• Survey instrument developed collaboratively by
  Thinking Math trainers and researchers
• Information session held for teachers and
  administrators
• Surveys administered to TM teachers
• Student data collected at school sites

14
Sample

• The study sample was teachers and 5th grade
  students in
• 4 of 13 elementary schools
• 3 schools with TM trained teachers at the 5th
  grade level.
• 1 school with TM trained teachers, none at the
  5th grade level

15
Sample Schools

• All offer Title 1 reading, two also offer math
• Fifth grade enrollments are 70, 55, 44, and 34

16
Sample Schools

• Percentages of low income students
  28, 36, 46, and 77 (PA Profiles,
  1999- 2000)
• Class organization
• heterogeneous and homogeneous
• departmentalized and self-contained

17
Group Task 2

• Review/analyze teacher questionnaire results
• Consider possible reasons for the results of the
  scale item means
• Are the scales correlated? Which ones?
• What other variables are related to the scales?
• What changes would you make to improve the
  instrument?

18
Teacher Survey Results
19
Teachers Surveyed

• 33 of 37 are female
• Pre-K - 5, Title 1, Learning Support
• 10 have lt ten years teaching experience
• 70 have twenty or more years of experience
• average of 1.5 ERD courses beyond TM

20
Teacher Survey Scales

• Confidence Scale ( 7 items)
• Rate yourself on your confidence in... e.g.,
  accepting multiple solutions.
• 1 Not confident 10 Very confident
• Scale Mean 8.57 s.d. 0.80

Scranton SD - Marywood University
NCTM 4/ 10/ 03
21
Teacher Survey Scales

• Implementation Scale (7 items)
• To what extent do you... e.g., have your
  students explain and justify their solutions to
  problems?
• 1 Not at all 10 To a great extent
• Scale Mean 8.1 s.d. 0.89

22
Influence of TM Scale Items

• 1. How frequently do you incorporate the TM
  principles in math instruction?
• 2. To what degree has your delivery of math
  instruction changed as a result of TM training?
• 3. How has the depth of the math content you are
  teaching changed since your TM training?
• 4. How much has TM research added to your content
  knowledge of mathematics?
• Scale Mean 8.19 s.d. 0.97

23
Correlations

• Confidence Implementation
• Influence of TM .395
  .429
• Confidence
  .877
• plt.05
• p lt .01
• n 37, two-tailed

24
Regression

• Implementation as the dependent variable
• Influence of TM as independent variable
• Confidence in TM Principles as independent
  variable
• 89 of the variance in implementation is
  accounted for by the two independent variables

25
Experience and Implementation

• Teacher Experience N Mean s.d.
• Less than 20 years 12 7.86
  0.81
• More than 20 years 25 8.33
  0.95

26
Teacher Comments on TM

• I now change problems to reflect my class,
  linking school math to real world experiences.
• I now use more manipulatives.
• It has given me new insight for teaching
  multiple strategies and meeting individual needs.

27
Student Achievement Results
28
Correlations SAT PSSA Math
29
Correlations Reading - Math
30
Research Hypotheses

• Students with 5th grade TM trained teachers will
  perform significantly better on PSSA 5th grade
  Math tests compared to the control group.
• Students with 5th grade TM trained teachers will
  perform significantly better on PSSA 5th grade
  Math tests than the district and state.

31
Math Performance Levels
_2 14.6 df 3, p lt .005
32
Math Performance Levels
Chi-square is not significant.
33
Mean Comparisons
t 2.40 df 39, p lt .05 t 2.74 df
195, p lt .01 t 4.0 df 41, p lt .000
34
Mean Comparisons
t 2.91 df 166, p lt .005 t gt 20 df
167, p lt .000
35
Mean Comparisons
Mean difference is not significant.
36
Teacher-Student Results
37
Mean Comparisons
t -1.76 df 197, p .081 t -0.91
df 197 p .081 t -2.04 df 197, p lt
.05
38
Research Question

• Can math achievement be predicted from teacher
  experience, professional development and
  implementation of the TM principles?

39
Analysis

• A multiple regression was run with SAT math
  procedures as the criterion variable, and three
  independent variables years of experience at
  grade level, professional development and scale
  score on implementation of TM principles.
• A second multiple regression used SAT problem
  solving as the dependent variable.

40
Math Procedures Regression
p .066 p lt .001
41
Math Problem Solving Regression
p lt .05
42
Reading Comprehension Regression
p .023 p lt .000
43
Conclusions

• Teacher confidence in TM principles and
  perception of the importance of TM are predictors
  of implementation which is in turn a predictor of
  student mathematics achievement.
• The number of years a student has a TM trained
  teacher is a predictor of student achievement in
  mathematics.

44
Conclusions

• Students who have a 5th grade TM trained teacher
  performed significantly better on the PSSA 5Th
  grade Math Test than the control group and the
  district.

45
Research Purpose

• The research results suggest that TM training
  does have a significant effect on student
  achievement.

46
Topics for Further Research

• Does TM training affect student achievement in
  Reading Comprehension?
• What is the relationship between TM training,
  student reading comprehension and student
  mathematics problem-solving achievement?

47
Questions?