Sean M' Lennon

1
PERCEPTION AND PRACTICE A TRIANGULATION STUDY
OF FIVE IDENTIFIED TRAITS RELATED TO SUCCESS IN
A DELAWARE AND A MARYLAND PUBLIC HIGH SCHOOL

• Sean M. Lennon

Skip Intro
2
Introduction to the Study

• To determine possible traits that relate to
  success for a public high school in Delaware and
  a public high school in Maryland
• Define possible traits that may, or may not lead
  to success
• Determine if literature, defining possible traits
  or concepts relates to success

3
Purpose of This Study

• To sample two successful schools
• Determined through published test data
• State web sites
• Delaware Dept. of Education
• Maryland Dept of Education
• Utilizing
• Separate school systems
• Separate states
• Separate accountability systems
• Isolate common practices, traits and/or
  ideologies
• Find commonalities related to success
• Confer these commonalities to the general
  educational system

4
Need For the Study

• Define success
• To meet testing (new) benchmarks
• Delineate traits that relate to success
• Determine traits that may have no relation to
  success
• Utilize modern statistical methodologies
• Study
• Combines delimited traits with the new testing
  accountability standards while utilizing mixed
  methods research techniques
• The utilization of this methodology to traits
  associated with assessments appears to be unique
• Internet search yielded no correlations

5
Determining Traits

• Cumulative trait chart. doc
• Delineated 5 traits from literature
• Collaborative work
• High expectations
• Professional Development
• School culture
• Strong leadership
• Meta Analysis (Frequency chart)
• Utilized literature containing multiple
  indicators
• Not limited to educational
• Counted the most frequencies, or times a trait
  was mentioned
• Culture was defined by component parts as
  described by Deal Peterson (1999) than weighted
  as 0.5 or ½ a point

6
Dependent Variable

• Success is the dependent variable
• Defined through the assessment approach
• Federal standards (NCLB)
• Continued growth (AYP)
• Delineated to specific demographic groups meeting
  success (AMO)
• Safe Harbor utilize other criterion standards
  to meet AYP
• 100 success rate by 2014
• State standards
• Delaware Student Testing Program
• Compares districts and schools
• Offers dual diplomas
• Maryland High School Assessment
• Compares districts and schools
• Required successful completion for graduation
  (class of 2009)
• Both schools have met AYP in the previous testing
  block and are in the top tier for state testing

(U.S. Department of Education, n.d.)
(Delaware Department of Education, 2003)
(Maryland State Department of Education, 2003)
7
Independent Variables

• Variables that (probably) cause, influence or
  affect outcomes
• Variables controlled by the experimenter
• Independent variables are the five indicators or
  domains
• Collaborative work
• School culture
• High expectations
• Professional development
• Strong leadership
• Discrete variables
• Variables that take on a small set of possible
  values

(Creswell, 2003, p. 94)
(Howell, 2004, p. 22)
(Howell, 2004, p. 22)
8
Other VariablesDefinitions

• Intervening Variables
• Can be referred to as a mediating variable
• A variable that comes between the dependent and
  independent variable
• Can mediate the effects of the independent
  variable on the dependent variable
• Confounding Variables
• Can be referred to as a spurious variable
• Influence not directly measured in a study
• Usually commented as an influence after the study
  has been completed

(Creswell, 2003, p. 94)
(Creswell, 2003)
Link to triangulation
9
Sample Frame

• A sampling frame is the list or quasi list of
  elements from which a probability sample is
  selected.
• Sample Frame needs to include
• Public high schools
• 2 states
• Deemed successful or high performing
• Faculty from two selected schools
• Administration from two selected schools
• Faculty and administration of a successful public
  high school in Delaware and a successful public
  high school in Maryland

(Babbie, 2002)
10
Sample Design

• Nonprobability
• Specific constructs
• Convenience
• Schools near each other
• Schools near the researcher
• Purposive
• It can be appropriate to select a sample on the
  basis of knowledge of a population, its elements
  and the purpose of the study
• Public high school faculty and administrators
• Faculty and administration of Easton High School
  (MD) and Sussex Technical High School (DE)

(Babbie, 2002, p. 178)
11
Research Design

• Mixed Method
• Suggested/conceptualized in 1959
• Not accepted until the 1980s
• Utilization of qualitative and quantitative
  concepts and methodologies
• The third wave or third research movement
• Recent increase in use or popularity is due to
  its pluralism or eclecticism, which frequently
  results in superior research
• Utilizes two or more modes of inquiry
• Garner a higher level of validity or
  corroboration
• To limit or find fault possibly missed in a
  monomethod study

(Rocco, et al, 2003)
(Johnson Onwuegbuzie, 2004, p. 17)
(Johnson Onwuegbuzie, 2004, p. 14)
12
Research Design
(Creswell, 2003)
13
Research Question

• Question
• Are there commonalities, or common attributes for
  success, in two different, geographically
  separated, high achieving public schools that can
  be inferred to the general public high school
  population?
• Sub-questions
• Is collaborative work an indicator of success and
  a common attribute in successful schools?
• School culture?
• High expectations?
• Professional development?
• Strong leadership?

Back to conclusion
14
Limitations/Delimits

• Limitations
• Traits may not infer to other schools
• Ambiguity in defining/assessing success
• Bias
• Specific demographics to the localized region
• Limited demographic variances in target schools
• Delimitations
• Limit to two schools
• Quantitative sample group limited to 100
  respondents
• Qualitative sample group limited to 4 respondents
• Traits, or indicators limited to 5 variables
• Instrument limited to less than 4 questions per
  domain/variable
• Qualitative questionnaire limited to 1 central
  question and 5 sub-questions

Back to closure
15
Testing Matrix
16
Testing Methodology

• Testing Design 1
• Answering the Research
• Question
• Tests 1, 2, 4, and 6
• External comparison of
• sample groups
• Testing Design 2
• Triangulation
• Tests 3 and 5
• Internal analysis of sample
• groups

Quantitative
Inferential
Descriptive
Quantitative
Qualitative
and
Inferential
Data Analysis
(Creswell, 2003)
17
Answering the Research Question

• Comparison of the two sample groups
• Testing Design 1
• Tests conducted
• Domain to domain
• Traits
• Answered Research Question
• As designed within the quantitative methodology
• Spearmans Correlation Coefficient for Ranked
  Data
• Question to question
• Consisted of two different tests
• Spearmans Correlation Coefficient for Ranked
  Data
• Kruskal-Wallis one-way analysis of variance
• Sample to sample
• Construct
• Spearmans Correlation Coefficient for Ranked
  Data
• Whole sample frame comparison

Test 4
(Howell, 2004, p. 22)

Tests 1 and 2
Test 6
Skip Stat Info
18
Test of Association

• Spearman rank correlation coefficient
• Spearman Rho
• Corresponds to the Pearson correlation
  coefficient
• Utilizes ranking not variables
• Correlates ranked data
• Measure of monotonic relationship
• Line is always falling or rising though not
  always straight
• Tests
• Each question from Easton High group to Sussex
  Technical High group
• Each domain from Easton High group to Sussex
  Technical High group
• Each complete sample group, including both
  qualitative and quantitative questions (Easton
  High School) to the other complete sample group
  (Sussex Technical High School)

(Dallal, 2000)
(Howell, 2004)
Test 2
Test 4
Test 6
19
Test for Association
Spearman Rho
rs strength of association chart
Basic generalization
(Correlational Research, n.d.)
20
Test for Association
Spearman Rho
Spearmans Correlation Coefficient for Ranked Data
Determining Significance
Students t distribution
Gaussion (normal) distribution
21
Tests of Differences

• Kruskal-Wallis
• Corresponds to the One way analysis of variance
  test (F test)
• 3 or more samples
• Mann Whitney
• Common test
• Corresponds to the t test (for independent
  samples)
• 2 independent samples

Test 3
Test 5
(Dallal, 2000)
(LTDI, 1999)
(Dallal, 2000)
Test 1
(Howell, 2004)
Back to closure
22
Testing Issue

• Trait Analysis by Question
• Discrepancy with 5 statements
• Familywise error rate
• Avoiding Type I Error
• Confidence level implications
• Removing Test of Significance
• Reducing bias
• Testing manipulation
• Utilizing RHO statistic only

Test 1
Test 2
(Howell, 2004)
23
Probability of a Type I error as a function of
the number of pairwise Comparisons where a 0.05
for any one comparison
Probability point
Probability of Type I error
1.0
0.9
0.8
0.7
0.6
Number of pairwise comparisons in tests 1 and 2
(36)
0.5
0.4
0.3
0.2
0.1
5
10
15
20
25
30
35
40
45
50
(Howell, 2004, pg 376)
Number of pairwise comparisons
Back to closure
24
Trait Analysis by Domain
Test 4
Spearman Rho

• Trait 1 - Collaborative Work
• Questions 1-4
• Trait 2 - Culture
• Questions 5-7
• Trait 3 High Expectations
• Questions 8-10
• Trait 4 - Leadership
• Questions 11-14
• Trait 5 Professional Development
• Questions 15-18

• Trait 1 - Collaborative Work
• Questions 1-4
• Trait 2 - Culture
• Questions 5-7
• Trait 3 High Expectations
• Questions 8-10
• Trait 4 - Leadership
• Questions 11-14
• Trait 5 Professional Development
• Questions 15-18

Sample Group Easton High School Faculty
Sample Group Sussex Technical High School Faculty
Test Spearmans Correlation Coefficient for
Ranked Data (Test 4) Defined Sample to sample
comparison of association by domain Collaborative
Tests Sample to sample comparison by questions
25
Trait Analysis by Domain
Test 4
Spearman Rho
Test Spearman Rho Defined Sample to sample
comparison of association by domain Collaborative
Tests Sample to sample comparison by question
26
Trait Analysis by Question
Test 2
Question 1 Question 2 Question 3 Question
4 Question 5 Question 6 Question
7 Question 8 Question 9 Question
10 Question 11 Question 12 Question
13 Question 14 Question 15 Question
16 Question 17 Question 18
Question 1 Question 2 Question 3 Question
4 Question 5 Question 6 Question 7 Question
8 Question 9 Question 10 Question 11 Question
12 Question 13 Question 14 Question
15 Question 16 Question 17 Question 18
Sample Group Sussex Technical High School Faculty
Sample Group Easton High School Faculty
Test Kruskal-Wallis One-Way Analysis of Variance
(Test 1) Defined Sample to sample comparison of
difference by question Collaborative Test
Spearmans Correlation Coefficient for Ranked
Data (Test 2)
27
Trait Analysis by Question
Test 2
Spearman Rho
rs Statistic for Individual Statements
28
Trait Analysis by Construct
Test 6
Spearman Rho
Faculty Response Survey Instrument Adminis
trative Response Questionnaire Instrument
Faculty Response
Survey Instrument
Administrative Response Questionnaire
Instrument
Sample Group Easton High School Faculty
Sample Group Sussex Technical High School Faculty
0.13 Rho Statistic
Test Spearmans Correlation Coefficient for
Ranked Data (Test 6) Defined Sample to sample
comparison of association
29
Descriptive Statistics

• Description or illustration of data
• 2 tests or data compilations
• Illustration by mean
• Individual questions compared between the two
  sample groups
• Difference in response between faculty and
  administration within sample groups

(Howell, 2004)
Illustrated the possible influence of mediating
variables
Illustrated differences in two traits for both
samples
30
School Means by Individual Statements
31
Test 5
Mean Difference by Domain for Easton High School
and Sussex Technical High School
High Expectations
Leadership
Difference in mean response between faculty and
administration
Note
32
Triangulation

• Testing Design 2
• Multiple methods to confirm, cross validate, or
  corroborate findings within a single study
• 2 tests with multiple data compilations
• Test 3
• Test 5
• Comparison of means
• Qualitative data

(Creswell, 2003, p. 217)
EHS
STHS
Mediating or confounding variables
Table
(Creswell, 2003)
33
Collaborative Work
Trait 1
34
Trait 1
Administrative Response for Collaborative Work
Domain
35
STHS Faculty divided in response Administration
not divided in response
EHS Faculty divided in response Administration
divided in response
Slight difference as illustrated in means
Slight difference as illustrated in means
Collaborative Work
36
Culture
Trait 2
37
Trait 2
Administrative Response for School Culture Domain
38
STHS Faculty divided in response Administration
not divided in response
EHS Faculty divided in response Administration
divided in response
No difference as illustrated in means
No difference as illustrated in means
Illustrates sample wide discrepancy towards
culture
Possible mediating variable
Culture
39
High Expectations
Trait 3
40
Trait 3
Administrative Response for High Expectations
Domain
41
Trait 3
Administration and Faculty Domain Mean Chart for
Easton High School
High Expectations
42
STHS Faculty not divided in response Administra
tion not divided in response
EHS Faculty divided in response Administration
not divided in response
No difference as illustrated in means
Statistical Difference as illustrated in means
and test 3
Illustrates an administrative belief inconsistent
towards faculty
Possibly influenced by a mediating variable
High Expectations
43
Leadership
Trait 4
44
Trait 4
Administrative Response for School Leadership
Domain
45
Trait 4
Administration and Faculty Domain Mean Chart for
Sussex Technical High School
Leadership
46
STHS Faculty somewhat divided in
response Administration divided in response
EHS Faculty divided in response Administration
divided in response
Difference as illustrated in means
Little difference as illustrated in means
Possible mediating variable
Possibly influenced by a mediating variable
Corroborated with qualitative data
Leadership
47
Professional Development
Trait 5
48
Trait 5
Administrative Response for Professional Developme
nt Domain
49
STHS Faculty not divided in response Administra
tion not divided in response
EHS Faculty not divided in response Administrat
ion divided in response
No difference as illustrated in means
No difference as illustrated in means
Illustrates sample wide discrepancy towards
professional development
Possibly influenced by a mediating variable
Professional Development
50
Conclusion

• Research Question
• Sub Questions
• Collaborative Work
• Culture
• High Expectations
• Leadership
• Professional Development
• Triangulation

Not defined or corroborated with this study
No association between the two sample schools for
all five traits
Defined individually and separate
Illustrated possible mitigating and confounding
variables
The traits interacting together
51
Conclusion

• No strong trait correlated to both schools
• Schools are too complex to differentiate
  specific, individual traits
• Schools are unique in culture and needs
• Evidence of traits being interconnected
• Culture and leadership probable mediating
  variables
• Collaborative work, high expectations and
  professional development influenced by mediating
  variables
• Evidence of issues or events in samples
• Previous for Sussex Technical High School
• Present for Easton High School

If not defining leadership under the context of
the cultural domain
52
Implications for EducationalLeadership

• Educational leaders
• Must not focus on these traits as individualistic
  factors to be addressed
• No definable, individual correlation to testable
  outcomes
• No definable correlation in a fixed, or short
  time period or testing block
• Need to understand the traits appear to be
  associated or have a commonality or influence to
  each other
• Difficult to ascertain the effects on the other
• Should think of the traits as combined elements
  or factors

Incorporated within or defined as school culture
Leadership
And possibly
53
Recommendations

• From this study
• Understand the vagaries of the five traits as
  distinct functions and to be aware of these
  limitations
• Understand the dualistic impact of the traits as
  a collaboration of school culture and concept
• Understand the importance of data in determining
  or isolating issues or problems
• For further research
• Extend the research question in incorporating the
  mitigating variables
• Extend the sample frame
• Establish mixed method as an effective research
  methodology for education

54
Closure

• Thank You
• Questions
• Links to sections within PowerPoint
• Study definitions and methodology
• Statistical test information and description
• Statistical issues and information

Chapters 1 through 3
Tests 1 and 2
Familywise error rate