Instructional Research and Research Synthesis

1
Instructional Research and Research Synthesis

• Russell Gersten
• Instructional Research Group
• University of Oregon

2
Focus

• Findings that challenge folk wisdom
• Multiplicities of interpretation
• The human factor (J. Singer)

3
Interventions for Struggling Readers Torgesen
Research

• Addressed common myth work on word reading is
  all struggling readers need
• Most importantly, addresses a wild card long
  term fix for struggling readers in grades 3-9
• A professional issue for me since 1969

4
A few issues that bubble over

• Homogeneous groups seem to work best for building
  foundational skills/ Heterogeneous for
  comprehension
• The right texts for students to read as they work
  on comprehension
• Limits of the strategy instruction model (Beck
  McKeown) limited understanding of need to
  possess relevant background knowledge (before
  activating)
• We can only hope for increments

5
Carlo et al Cognates

• Context common feature of texts on teaching
  methods
• Most useful for technical words and Latin roots
  (for Latin languages)
• Ironies English learners born in U. S. dont
  have large tech vocabularies
• Catch 22 They dont always work
• Does one teach awareness? Or really work on this
  ability

6
Thus the study

• Findings open up new issues
• Again, impact on words taught, NS on untaught
  cognates
• Implication More to build broad awareness? Not a
  big deal?
• Deeper implications Common finding for
  vocabulary instruction (depth, but not transfer)

7
The Research Synthesis Part 1 Instructional
Interventions

• Strength Does not overshoot the data
• Points out that the adjustments are NOT CLEARLY
  EXPLAINED
• Helps point out that these need to be focus of
  next wave of instructional research experimental
  and observational/ descriptive

8
The Research Synthesis Part 2 Language of
Instruction

• Politically controversial (unlike the first)
• Raises issue of how each meta-analysis is an act
  of interpretation and decision rules (defining
  terms, classifying interventions, methodological
  inclusion) affect outcomes

9
My Two Cents Worth

• RCT as gold standard is not enough
• Though some methodological issues stressed,
  others ignored
• Total confounding of teachers with intervention
• Differential attrition
• Use of covariates that are salient predictors
  (e.g. oral language has zero predictive validity)
• Control for instructional time

10
A very different picture would emerge