MethodParticipants. • Approximately 64 second (43.8%) • and third (56.3%) grade students. • 51.6% female and 48.6% maleMeasures• Oral Reading Fluency (ORF)

• Administered by school personnel 3 times per year (fall, winter, and spring). Spring scores were used for analyses.

• Students read from 3 one minute grade level passages, and the median score was recorded.

• Fountas & Pinnell Benchmark Assessment System (BAS)• Instructional level determined by spring scores based on fluency,

accuracy, and comprehension.• Curriculum Based Assessment for Instructional Design (CBA-ID)

• Administered by researchers one time in the spring. • Students read from 3 books (1 minute each) based on their BAS

instructional level.• The number of words read correctly divided by the total words

read was recorded and the median score was used for analyses.

UNIVERSITY OF MINNESOTADepartment of Educational Psychology

Sandra M. Pulles, Kathrin E. Maki, & Matthew K. Burns

Relationship Between Reading Inventory Instructional Level

and Student Reading Performance

School Psychology250 Education Sciences Building 56 E. River Road, Minneapolis, MN 55455

Contacts: Sandra M. Pulles: moran122@umn.edu Kathrin E. Maki: makix312@umn.edu Matthew K. Burns: burns259@umn.edu

College of Education + Human Development

Introduction• Instructional match is closely associated with improved student

learning (Burns, 2007; Daly, Martens, Kilmer, & Massie, 1996)

• Instructional level occurs when students have sufficient background knowledge to interact with the material, yet still experience some level of challenge (Betts, 1946).

• Frustrational: <93%• Instructional: 93-97%• Independent: 97-100%

• Two ways to assess instructional level1. Informal Reading Inventories (IRI)

• Fountas & Pinnell (1996)

• 20-30 minutes per student; low psychometric properties• Provide students

2. Curriculum Based Assessment for Instructional Design (CBA-ID;

Gickling & Havertape, 1981).

• Students read from instructional level text for 1 minute and the number of words read correctly is recorded and accuracy is computed.

• 5 minutes per student; high psychometric propertiesResearch Questions1. What level of agreement is there between instructional level

estimates from reading three books from the same reading level? 2. To what extent does the estimate of instructional level from a

reading inventory agree with instructional level estimates from reading the corresponding leveled book?

3. How do reading skills affect agreement between estimate of instructional level from a reading inventory and estimates from reading the corresponding leveled book?

Discussion

• Students did not consistently read with accuracy from books rated at their IRI instructional level

• Students read with 93 to 97% accuracy about 28% of the time• Struggling readers frequently failed to read with 93% accuracy• High readers were not challenged enough by their IRI instructional level• Psychometric issues associated with IRIs make it difficult to obtain an accurate

student instructional level• Reliability-inconsistency across books • Validity-use of IRIs for determining instructional level

• Matching instructional material with student skill level results in improved student outcomes (Burns, 2007)

• Students should therefore be reading at their instructional level to ensure adequate reading growth

Limitations

• Many students were higher readers therefore limiting generalizability to other skill levels

• No direct measure of comprehension was used• There was no control over prior exposure thus it is unknown whether or not students

were familiar with the material

 

Mean

(SD)

Frustration n

(%)

Reading Level

Instructional n

(%)

Independent n

(%)

Spring Benchmark Oral Reading Fluency

128.08

(48.71)

NA NA NA

Reading 1 Accuracy

96.7%

(3.27%)

8

(12.5%)

23

(35.9%)

33

(51.6%)

Reading 2 Accuracy

96.4%

(4.91%)

10

(15.6%)

15

(23.4%)

39

(60.9%)

Reading 3 Accuracy

96.1%

(5.9%)

11

(17.2%)

17

(26.6%)

36

(56.3%)

Median Accuracy 96.7%

(3.6%)

10

(15.6%)

18

(28.1%)

36

(56.3%)

  Reading 1 Accuracy Reading 2 Accuracy Reading 3 Accuracy

Reading 1 Accuracy   r = .47* r = .61*

Reading 2 Accuracy r = .67* r = .68*

Reading 3 Accuracy r = .67* r = .59*  

  Reading 1 Accuracy Reading 2 Accuracy Reading 3 Accuracy

Reading 1 Accuracy   70.3% 68.8%

Reading 2 Accuracy k =.49* 67.2%

Reading 3 Accuracy k = .47* k = .42*  

  Median Percent AccurateCBA-ID Categorical

Score

IRI Instructional Level r = .65 tau = .65

Group Frustration n

(%)

Instructional n

(%)

Independent n

(%)

Low – 25th Percentile or Less 7

(58%)

5

(41.7%)

0

(0.0%)

Middle – 26th to 75th Percentile

2

(9.5%)

4

(19.0%)

15

(71.4%)

High – 76th Percentile or Higher

1

(3.2%)

9

(29.0%)

21

(67.7%)

Table 1

Descriptive Statistics and Frequency Data

Table 2 Correlations Among Accuracy Measures from Three Reading Performance Assessments

Table 3 Percent Agreement and Kappa Among Accuracy Measures from Three Reading Performance Assessments

Table 4

Correlation between IRI Instructional Level and CBA-ID Accuracy and Categorical Score

Table 5 Number and Percentage of Median Accuracy Scores from Three Reading Performance Assessments that Fell within the Frustration, Instructional, and Independent Level by Skill Group

ResultsResults

Ethnicity

Caucasian

African American

Hispanic

Asian-American

Low Average High

2nd Grade < 69 69-120 > 120

3rd Grade < 84 84-140 > 140