Examining the Effect of Read Aloud Accommodations on Students Math Test Performance Michael Russell

1
Examining the Effect of Read Aloud
Accommodationson Students Math Test
PerformanceMichael RussellTechnology and
Assessment Study CollaborativeBoston College
2
Math Tests

• Require students to read English proficiently
• Tap into multiple constructs besides math
  knowledge
• Reading Comprehension
• Writing skills
• Populations of concern ELL and IEP

3
Populations of concern

• 1. English language learners
• Additional Constructs
• English proficiency
• Translation from English to L1
• 2. Special Education Students
• Additional Constructs
• Reading Comprehension

4
Read Aloud Accommodation

• Most states provide read aloud accommodations
  for ELL and IEP students, usually delivered by a
  human reader
• Problems with accommodation delivery
• Difficulty standardizing delivery
• Reader influence on students responses

5
Read Aloud Study

• Research questions
• Which type of delivery modeDigitized Voice or
  human readeris more effective?
• Do these read aloud accommodations reduce
  influence of irrelevant constructs?
• Which types of students are harmed or aided by
  the option?

6
Study Design

• Three formats
• Computer-delivery with digital technology read
  aloud accommodation
• Paper-based test with human reader read aloud
  accommodation
• Paper-based with no accommodation

7
Sample

• Study Sample
• 235 10th/11th grade students
• 82 ELL students
• 40 IEP students
• 3 NH High Schools
• Students randomly assigned to groups within
  schools

8
Data Collected

• Math test
• 12 mc questions
• NH 2001 and 2002 released items
• Computer Fluidity
• Computer Literacy - VA NC
• Computer Use Survey

9
Test Interface
10
Tutorial

• 3 minute
• Controlled pace
• Described and demonstrated functionality and
  options

11
Findings

• No statistically significant differences between
  accommodation delivered by computer and
  accommodation delivered by human reader
• Statistically significant differences between
  groups with accommodations and group without
  accommodation (.45 effect size)

12
IEP Status

• Interaction between mode of delivery and IEP
  status on test scores BUT it is not statistically
  significant

13
Multilingual Status

• Interaction between mode of delivery and
  Multilingual status on test scores BUT it is not
  statistically significant

14
Modal Effect by Computer Fluidity
No significant effects within computer fluidity
levels No significant interactions
15
Modal Effect by Computer Literacy
Significant Effect within Medium only No
significant interactions
16
Modal Effect by School Use
Significant Effect within high only No
significant interactions
17
Summary of Findings

• Results indicate that delivering the read aloud
  accommodation by computer appears to be is as
  effective as using a human reader however,
  students indicated that they preferred taking the
  computer test
• Students who received read aloud accommodations
  by the two modes tested performed better than
  students that did not receive accommodations
• No interactions between IEP or Multilingual
  Status and modal effect
• Low computer fluidity and literacy benefit less
  from CBD