The Early Development Instrument EDI: An Examination of Validity

1
The Early Development Instrument (EDI) An
Examination of Validity

• Shelley Hymel, UBC
• Lucy LeMare, SFU
• William McKee, UBC
• Edudata Canadas Research Forum, 5 May 2006
• Morris J. Wosk Centre for Dialogue, Vancouver, BC

2
Acknowledgements

• Our sincere thanks to
• To the BC Ministry of Education and UBCs Human
  Early Learning Partnership (HELP) for their
  financial support
• The Vancouver, Surrey and New Westminister school
  districts for their participation
• To student assessors Karyn Audet, Joan Broto,
  Carly Glanzberg, Andrew Higgs, Twila MacDonald,
  Kaitlin McKee and Jared Reid for their many hours
  of testing
• To graduate students Nikki Hearle, Lisa Kihl,
  Carrie Strangway and Sarah Van Leeuvwan for their
  help in data collection, scoring and entry
• To Dr. Laurie Ford and Dr. Bruno Zumbo for their
  consultation on this work and to Victor Glickman
  and Clyde Hertzman for their support and
  especially their patience.

3
Early Development Instrument (EDI)Five Domains

• Physical Health and Well Being
• Social Competence
• Emotional Maturity
• Language and Cognitive Development
• Communication and General Knowledge

4
Types of Validity

• Sampling Validity
• Is the sample large enough?
• Is the sample representative of the population?
• Measurement Validity
• Content Validity
• Does the scale assess the entire domain?
• Construct or Criterion Validity (concurrent)
• Does the scale measure the underlying
  characteristic it is supposed to measure?
• Predictive Validity
• Does the scale allow researchers to make
  accurate predictions about future?

5
Child-Based Validity Measures

• Early Screening Inventory Kindergarten Version
• (ESI-K Revised, Meisels et al. 1997)
• Bracken Basic Concept Scale- School Readiness
  Composite (Bracken SRC -Revised, Bracken, 1998)
• Comprehensive Test of Phonological Processing
• (CTOPP, Wagner, Torgesen, Rashotte, 1999
• Group for the Study of Interpersonal Development
• Relationship Questionnaire
• (GSID REL-Q, Shultz Selman, 1998

6
Sample Characteristics

• N267 Kindergarten students
• Data from 27 teachers in 17 schools, in 3
  districts
• Gender 53 Girls (n 125) 47 Boys (n142 )
• First Nations 97 Non-Aboriginal 3
  Aboriginal
• First Language 71 English 28 Other
• ESL 70 No 29 Yes
• Special Needs 99 No 1 Yes

7
Method

• Participants were normally involved in 2 testing
  sessions of approx. 30 minutes each
• Nine different testers, graduate and
  undergraduate students from SFU and UBC
• EDI ratings completed within approx. 3 months of
  child assessments (some before, some after)

8
Correlations Between EDI Scores and Overall
Validity Measures (N249-257)
9
Correlations Between EDI Scores and Overall
Validity Measures (N249-257)
10
EDI Subscales

• Physical Well Being
• Social Competence
• Emotional Maturity
• Language Cognition
• Communication General Knowledge

• Physical Readiness for the school day
• Physical independence
• Gross Fine Motor Skills
• Overall Social Competence
• Responsibility and Respect
• Approaches to Learning
• Readiness to explore new things
• Prosocial and Helping Behavior
• Anxious and Fearful Behavior
• Aggressive Behavior
• Hyperactivity and Inattention
• Basic Literacy Skills
• Interest in literacy/numeracy and Memory
• Advanced Literacy Skills
• Basic Numeracy Skills

11
EDI Physical Scale Subscales (N254-267)
12
EDI Communication/Gen.Knowledge Scale Subscales
(N252-257)
13
EDI Social Competence Scale Subscales
(N245-249)
14
EDI Emotional Maturity Scale Subscales
(N246-249)
15
EDI Language Cognition Scale Subscales
(N249-267)
16
Regression Predicting EDI Scores From Validity
Measures (N237)R.60, R2.36, F
(4,232)32.76, plt.001(all four predictors
significant)
17
Classroom Level Aggregation?

• Correlations between EDI Total Scores and Overall
  School Readiness Scores among 12 teachers for
  whom we had data on at least 10 students (range
  10-21)
• EDI Total and ESI-K Totals ranged from .17 to
  .95
• EDI Total and Bracken SRC ranged from -.04 to
  .66

18
Conclusions

• Results support the concurrent validity of the
  EDI at the level of overall composite scores and
  interpretation of scores at the population level.
• Evidence for the convergent and discriminant
  validity of EDI domain and subscale (factor
  scores) was less compelling.
• Interpretation at the level of individuals and
  classrooms is questionable given the variability
  observed across teachers in the correspondence
  between EDI scores and standardized readiness
  measures.
• Given these findings, the most appropriate use
  of the EDI with regard to individuals or
  classrooms would be as a marker variable,
  flagging, but not identifying potential problems
  indicating a need for further assessment.
• The EDI remains a viable index of readiness
  when interpreted at the population level.