Reliability and Validity of the Reading Level Assessment and the “Flash” Word Recognition Automaticity Measure Grace T. Craig [email protected] Kathleen.

Reliability and Validity of the Reading Level Assessment

and the “Flash” Word Recognition Automaticity

MeasureGrace T. Craig [email protected]

Kathleen J. BrownMatthew K. Fields

University of Utah Reading ClinicR. Darrell Morris

Appalachian State University

Methods

• 4 schools– 2 = Title 1 1 = public, 1 = parochial– 2 = non-Title 1 both = public & mixed SES

• 192 students in G2-G5 in March, 2006

• Rank ordered DIBELS or QRI of each grade within a school, then sampled 12 students: 4 high, 4 average, 4 poor to achieve a representative distribution for testing

Methods

• 135 minutes of assessment in 3 sessions

• Two forms of Reading Level Assessment (RLA) and a standardized test- Gray Oral Reading Test

• Manual and computer presentations of Flash

• Tests and presentations were counterbalanced

• Manual flash interrater differences = n.s.

Reading Level Criteria

Acc.

(%)

Rate

(wpm)

Comp

(%)

Mid G1 90 30 60

End G1 90 40 60

End G2 93 90 60

End G3 93 110 60

End G4 95 120 60

End G5 95 130 60

End G6 95 150 60

End G7 95 150 60

End G8 95 150 60

Flash Percentage 80 (for all graded lists)

Alternate Form Reliability

• To what extent are RLA Form A scores equivalent to RLA Form B scores?

• To what extent are computer presentation Flash scores equivalent to manual presentation Flash scores?

Results: Alternate Form Reliability

Spearman’s Rho Correlations

RLA Form B

Flash Manual

Flash Comp

RLA A .906** .774** .830**

RLA B .829** .869**

Flash Manual .820**

**p < .01

Wilcoxon Signed Ranks Test(Non-parametric paired hypothesis test)

Z p

RLA Form B – RLA Form A

-1.738a .082

Flash Comp – Flash Manual

-.610a .542

a. Based on positive ranks.

Criterion Validity

• To what extent are Flash scores and RLA scores consistent with scores achieved on a “flagship” standardized reading measure, the Gray Oral Reading Test (GORT)?

Results: Criterion Validity

Spearman’s Rho Correlations

RLA B

GORT

Flash

RLA A .906** .871** .836**

RLA B .887** .885**

GORT .868**

**p < .01

Content Validity

• Is the Reading Level Assessment representative of grade level benchmarks?

• Are the graded passages of the Reading Level Assessment representative of their respective grades?

Results: Content Validity

Median Grade Equivalent by Grade

RLA Flash

GORT

Grade 2 2.9

2.9

3.9

Grade 3 3.9

3.9

5.5

Grade 4 4.9

4.9

7.2

Grade 5 5.9

5.9

7.7

85

88

91

94

97

100

20 50 80 110 140 170 200Rate

Acc

urac

y

93%

90 WPM

Accuracy and rate of second grade students on Grade 2 passage.

85

88

91

94

97

100

50 75 100 125 150 175 200Rate

Acc

ura

cy

93%

110 WPM

90

92

94

96

98

100

65 80 95 110 125 140 155 170 185 200Rate

Acc

ura

cy

95%

120 WPM

88

90

92

94

96

98

100

60 80 100 120 140 160 180 200Rate

Acc

ura

cy95%

130 WPM

Accuracy and rate of third grade students on Grade 3 passage.

Accuracy and rate of fourth grade students on Grade 4 passage. Accuracy and rate of fifth grade students on Grade 5 passage.

52% pass 53% pass

59% pass 51% pass

Conclusions: Alternate Form Reliability

• RLA A and RLA B seem to be equivalent forms

• Manual Flash and Computer Flash seem to be equivalent forms

• Examiners can be trained to mimic a 300 ms eye fixation without significant difference from a computer

Conclusions: Criterion Validity

• RLA Form A and RLA Form B are strongly correlated with a popular standardized oral reading test, the GORT

• The Flash is strongly correlated with the GORT

• These correlations indicate that the Reading Level Assessment and the Flash instrument are, like the GORT, testing reading ability

Conclusions: Content Validity

• The Reading Level Assessment seems to have high validity for identifying students’ instructional reading levels

• The Flash seems to have high validity for identifying students’ instructional levels

• The GORT seems to identify student instructional levels which are inflated by at least a year, and sometimes more than two years