abstract in a response to intervention (rti) model, educators in both general and special education...

1
Abstract In a response to intervention (RTI) model, educators in both general and special education use curriculum- based measurement (CBM) oral reading fluency (ORF) assessments across grades to monitor progress of students receiving reading intervention. Researchers have explored average growth in ORF and found decelerating, nonlinear growth rates across grade-levels (e.g., Christ et al., 2010; Nese et al., 2012; Nese et al., 2013). In this study, we examine growth for those students in Grades 1-8 specifically identified for reading intervention and receiving regular progress monitoring, and explore two growth models to determine (a) the within-year ORF growth, and (b) the growth across Tier II/III intervention. Results Table 1: Trajectory fixed effect parameters (intercept and slopes) were similar across Within-year and Tier II/II growth models. Figure 1: The first testing occasion for Tier II/II students was nearly always in the early fall for all grades but first (winter). Figure 2 • Decelerating Tier II/II growth in Grades 1-5. • Linear Tier II/II growth in Grades 6-8. Growth of progress monitored students similar to that of easyCBM 20 th percentile group, but with lower intercepts and greater yearly gains in most grades. Describing the Reading Fluency Growth of Progress Monitored Students Joseph F. T. Nese, Julie Alonzo, Gerald Tindal National Center on Assessment and Accountability for Special Education, 2013 References Christ, T. J., Silberglitt, B., Yeo, S., & Cormier, D. (2010). Curriculum-based measurement of oral reading: An evaluation of growth rates and seasonal effects among students served in general and special education. School Psychology Review, 39, 447-462. Muthén, L. K., & Muthén, B. O. (1998-2013). Mplus user’s guide (7th ed.). Los Angeles, CA: Muthén & Muthén. Nese, J. F. T., Biancarosa, G., Anderson, D., Lai, C. F., & Tindal, G. (2012). Within-year oral reading fluency with CBM: A comparison of models. Reading and Writing, 25, 887-915. DOI: 10.1007/s11145-011-9304-0 Nese, J. F. T., Biancarosa, G., Cummings, K., Kennedy, P., Alonzo, J., Tindal, G. (2013). In search of average growth: Describing within-year oral reading fluency growth for Grades 1-8. DOI: 10.1016/j.jsp.2013.05.006 For More Information Please contact Joe Nese: [email protected] . More information on this and related projects can be obtained at http://ncaase.com . Funding Source This project was funded through the National Center on Assessment and Accountability for Special Education (NCAASE) grant (R324C110004) from the U.S. Department of Education, Institute of Education Sciences. Opinions expressed do not necessarily reflect the opinions or policies of the U.S. Department of Education. Selection Rules for ORF Progress Monitoring Sample Ste p Rules Studen ts Score s Original data 139,16 5 495,1 74 1) Delete out of range scores (e.g., < 0 and > 365) - 495,1 12 2) Delete students in grades other than 1-8 137,84 8 - 3) Delete students with any instance of off- grade-level testing 129,61 7 - 4) Delete students who scored > 30th percentile on first benchmark or progress monitoring assessment 35,923 - 5) Delete scores from assessments > 20th testing occasion - 145,4 78 Delete “invalid” scores - - If two scores are less than two weeks (14 days) apart AND are different by > 35 WCPM, delete the score that is least like adjacent scores. 57,78 7 Based on the following rule of growth: [(Expected growth per week+SE g ) * 2 weeks)] + ( SEM * 2 weeks) - - [(1.5wcpm+1.0SE g ) * 2] + (15*2) - - 5 + 30 - - Take the median of scores that are within Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun 0 10 20 30 40 50 60 70 80 First ORF assessment (%) Method We conducted two latent growth models for each grade using Mplus version 7.11 (Muthén & Muthén, 1998-2013): (a) Within-year growth, where the time metric was the calendar year, such that the intercept was either late September or early October; (b) Tier II/III growth, where the time metric was the assessment occasion, such that the intercept was the first assessment occasion. Correlation G rade Param eters Correlation W ithin-m odel Betw een-m odels ( n ) M odel Intercept Linear Quadratic I-L I-Q L-Q I-I L-L Q-Q 1 W ithin-year 2.38** 0.42** 0.01** .52 .76 -.01 .60 .66 -.58 (540) TierII/III 8.22** 1.16** -0.01* .66** -.42 -.79** 2 W ithin-year 24.50** 1.22** 0.00 .34** -.15** -.89** .70 .85 .73 (1170) TierII/III 28.03** 1.55** -0.01** .59** -.54** -.92** 3 W ithin-year 44.06** 2.39** -0.04** .15** -.11* -.97** .86 .88 .84 (1180) TierII/III 48.70** 2.44** -0.04** .42** -.43** -.96** 4 W ithin-year 69.36** 1.50** -0.02** .36** -.28** -.97** .96 .95 .92 (1165) TierII/III 71.17** 1.63** -0.02** .47** -.42** -.97** 5 W ithin-year 96.68** 0.82** 0.00 .56** -.53** -.72** .98 .89 .76 (941) TierII/III 96.61** 1.26** -0.02** .29* -.24* -.91** 6 W ithin-year 92.09** 0.68** -0.01 .35 -.28 -.91** .97 .82 .69 (197) TierII/III 94.67** 0.57** 0.00 .13 .01 -.93** 7 W ithin-year 109.99** 0.25 0.01 .64** -.64** -.86** .98 .97 .92 (107) TierII/III 110.30** 0.45* 0.00 .63** -.71** -.95** 8 W ithin-year 114.44** -0.50* 0.03** .20 -.07 -.98** .98 .64 .49 (72) TierII/III 114.37** 0.11 0.01 .40* -.42 -1.00** Discussion Tier II/III interventions begin in the fall; less identification of at-risk students beyond fall. Implications for winter/spring benchmarks in RTI system: Lost opportunities for educators and students for lack of RTI system training? As hypothesized, growth rates are lower than 50 th percentile and average empirical trajectories (e.g., Nese et al., 2013), except for Grades 4 & 5. In all grades but sixth, linear growth rates greater in magnitude for the Tier II/II metric than the Within-year metric. Tenuous evidence for effective interventions. Limitations Selected sample very specific. Intervention information (if any) unknown. Table 1. Fixed Effects and Within- and Between-Model Correlations for the Quadratic Within-year and Tier II/II Growth Models Figure 1. Percent of Students First ORF Assessment for Each Month Across Grades 1-8 Figure 2. Predicted mean ORF scores in words correct per minute (wcpm) by grade based on the Within-year and Tier II/II time metrics, and observed mean ORF benchmark scores at the 20 th and 50 th percentiles for the easyCBM system. 0 30 60 90 120 150 180 26-Jul 14-Sep 3-N ov 23-D ec 11-Feb 1-Apr 21-M ay 10-Jul Fluency (w cpm ) G rade1 0 30 60 90 120 150 180 26-Jul 14-Sep 3-N ov 23-D ec 11-Feb 1-Apr 21-M ay 10-Jul Fluency (w cpm ) G rade2 0 30 60 90 120 150 180 26-Jul 14-Sep 3-N ov 23-D ec 11-Feb 1-Apr 21-M ay 10-Jul Fluency (w cpm ) G rade3 0 30 60 90 120 150 180 26-Jul 14-Sep 3-N ov 23-D ec 11-Feb 1-Apr 21-M ay 10-Jul Fluency (w cpm ) G rade4 0 30 60 90 120 150 180 26-Jul 14-Sep 3-N ov 23-D ec 11-Feb 1-Apr 21-M ay 10-Jul Fluency (w cpm ) G rade5 0 30 60 90 120 150 180 26-Jul 14-Sep 3-N ov 23-D ec 11-Feb 1-Apr 21-M ay 10-Jul Fluency (w cpm ) G rade6 0 30 60 90 120 150 180 26-Jul 14-Sep 3-N ov 23-D ec 11-Feb 1-Apr 21-M ay 10-Jul Fluency (w cpm ) G rade7 0 30 60 90 120 150 180 26-Jul 14-Sep 3-N ov 23-D ec 11-Feb 1-Apr 21-M ay 10-Jul Fluency (w cpm ) G rade8

Upload: daniella-kelley

Post on 25-Dec-2015

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Abstract In a response to intervention (RTI) model, educators in both general and special education use curriculum-based measurement (CBM) oral reading

AbstractIn a response to intervention (RTI) model, educators in both general and special education use curriculum-based measurement (CBM) oral reading fluency (ORF) assessments across grades to monitor progress of students receiving reading intervention.

Researchers have explored average growth in ORF and found decelerating, nonlinear growth rates across grade-levels (e.g., Christ et al., 2010; Nese

et al., 2012; Nese et al., 2013).

In this study, we examine growth for those students in Grades 1-8 specifically identified for reading intervention and receiving regular progress monitoring, and explore two growth models to determine (a) the within-year ORF growth, and (b) the growth across Tier II/III intervention.

AbstractIn a response to intervention (RTI) model, educators in both general and special education use curriculum-based measurement (CBM) oral reading fluency (ORF) assessments across grades to monitor progress of students receiving reading intervention.

Researchers have explored average growth in ORF and found decelerating, nonlinear growth rates across grade-levels (e.g., Christ et al., 2010; Nese

et al., 2012; Nese et al., 2013).

In this study, we examine growth for those students in Grades 1-8 specifically identified for reading intervention and receiving regular progress monitoring, and explore two growth models to determine (a) the within-year ORF growth, and (b) the growth across Tier II/III intervention.

Results Table 1: Trajectory fixed effect parameters

(intercept and slopes) were similar across Within-year and Tier II/II growth models.

• Figure 1: The first testing occasion for Tier II/II students was nearly always in the early fall for all grades but first (winter).

Figure 2• Decelerating Tier II/II growth in Grades 1-5.• Linear Tier II/II growth in Grades 6-8.• Growth of progress monitored students

similar to that of easyCBM 20th percentile group, but with lower intercepts and greater yearly gains in most grades.

Results Table 1: Trajectory fixed effect parameters

(intercept and slopes) were similar across Within-year and Tier II/II growth models.

• Figure 1: The first testing occasion for Tier II/II students was nearly always in the early fall for all grades but first (winter).

Figure 2• Decelerating Tier II/II growth in Grades 1-5.• Linear Tier II/II growth in Grades 6-8.• Growth of progress monitored students

similar to that of easyCBM 20th percentile group, but with lower intercepts and greater yearly gains in most grades.

Describing the Reading Fluency Growth of Progress Monitored Students

Joseph F. T. Nese, Julie Alonzo, Gerald TindalNational Center on Assessment and Accountability for Special Education, 2013

ReferencesChrist, T. J., Silberglitt, B., Yeo, S., & Cormier, D. (2010). Curriculum-based

measurement of oral reading: An evaluation of growth rates and seasonal effects among students served in general and special education. School Psychology Review, 39, 447-462.

Muthén, L. K., & Muthén, B. O. (1998-2013). Mplus user’s guide (7th ed.). Los Angeles, CA: Muthén & Muthén.

Nese, J. F. T., Biancarosa, G., Anderson, D., Lai, C. F., & Tindal, G. (2012). Within-year oral reading fluency with CBM: A comparison of models. Reading and Writing, 25, 887-915. DOI: 10.1007/s11145-011-9304-0

Nese, J. F. T., Biancarosa, G., Cummings, K., Kennedy, P., Alonzo, J., Tindal, G. (2013). In search of average growth: Describing within-year oral reading fluency growth for Grades 1-8. DOI: 10.1016/j.jsp.2013.05.006

ReferencesChrist, T. J., Silberglitt, B., Yeo, S., & Cormier, D. (2010). Curriculum-based

measurement of oral reading: An evaluation of growth rates and seasonal effects among students served in general and special education. School Psychology Review, 39, 447-462.

Muthén, L. K., & Muthén, B. O. (1998-2013). Mplus user’s guide (7th ed.). Los Angeles, CA: Muthén & Muthén.

Nese, J. F. T., Biancarosa, G., Anderson, D., Lai, C. F., & Tindal, G. (2012). Within-year oral reading fluency with CBM: A comparison of models. Reading and Writing, 25, 887-915. DOI: 10.1007/s11145-011-9304-0

Nese, J. F. T., Biancarosa, G., Cummings, K., Kennedy, P., Alonzo, J., Tindal, G. (2013). In search of average growth: Describing within-year oral reading fluency growth for Grades 1-8. DOI: 10.1016/j.jsp.2013.05.006

For More InformationPlease contact Joe Nese: [email protected]. More information on this and related projects can be obtained at http://ncaase.com.

For More InformationPlease contact Joe Nese: [email protected]. More information on this and related projects can be obtained at http://ncaase.com.

Funding SourceThis project was funded through the National Center on Assessment and Accountability for Special Education (NCAASE) grant (R324C110004) from the U.S. Department of Education, Institute of Education Sciences. Opinions expressed do not necessarily reflect the opinions or policies of the U.S. Department of Education.

Funding SourceThis project was funded through the National Center on Assessment and Accountability for Special Education (NCAASE) grant (R324C110004) from the U.S. Department of Education, Institute of Education Sciences. Opinions expressed do not necessarily reflect the opinions or policies of the U.S. Department of Education.

Selection Rules for ORF Progress Monitoring Sample

Selection Rules for ORF Progress Monitoring Sample

Step Rules Students ScoresOriginal data 139,165 495,174

1) Delete out of range scores (e.g., < 0 and > 365) - 495,1122) Delete students in grades other than 1-8 137,848 -3) Delete students with any instance of off-grade-level testing 129,617 -

4)Delete students who scored > 30th percentile on first benchmark or progress monitoring  assessment 35,923 -

5) Delete scores from assessments > 20th testing occasion - 145,478Delete “invalid” scores - -If two scores are less than two weeks (14 days) apart AND are different by > 35 WCPM, delete the score that is least like adjacent scores. 57,787

Based on the following rule of growth:

[(Expected growth per week+SEg) * 2 weeks)] + ( SEM * 2 weeks) - -

[(1.5wcpm+1.0SEg) * 2] + (15*2) - -5 + 30 - -

6)Take the median of scores that are within 7 days keep scores that are >7 and <28 days of each other, and delete scores that > 28 days apart.

34,550

7) Delete students with < 3  progress monitoring tests 5,373 30,628

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun0

10

20

30

40

50

60

70

80

Firs

t OR

F as

sess

men

t (%

)

MethodWe conducted two latent growth models for each grade using Mplus version 7.11 (Muthén & Muthén,

1998-2013):

(a) Within-year growth, where the time metric was the calendar year, such that the intercept was either late September or early October;

(b) Tier II/III growth, where the time metric was the assessment occasion, such that the intercept was the first assessment occasion.

MethodWe conducted two latent growth models for each grade using Mplus version 7.11 (Muthén & Muthén,

1998-2013):

(a) Within-year growth, where the time metric was the calendar year, such that the intercept was either late September or early October;

(b) Tier II/III growth, where the time metric was the assessment occasion, such that the intercept was the first assessment occasion.

Correlation Grade

Parameters

Correlation Within-model

Between-models

(n) Model Intercept Linear Quadratic I-L I-Q L-Q

I-I L-L Q-Q 1 Within-year 2.38** 0.42** 0.01**

.52 .76 -.01

.60 .66 -.58 (540) Tier II/III 8.22** 1.16** -0.01*

.66** -.42 -.79** 2 Within-year 24.50** 1.22** 0.00

.34** -.15** -.89**

.70 .85 .73 (1170) Tier II/III 28.03** 1.55** -0.01**

.59** -.54** -.92** 3 Within-year 44.06** 2.39** -0.04**

.15** -.11* -.97**

.86 .88 .84 (1180) Tier II/III 48.70** 2.44** -0.04**

.42** -.43** -.96** 4 Within-year 69.36** 1.50** -0.02**

.36** -.28** -.97**

.96 .95 .92 (1165) Tier II/III 71.17** 1.63** -0.02**

.47** -.42** -.97** 5 Within-year 96.68** 0.82** 0.00

.56** -.53** -.72**

.98 .89 .76 (941) Tier II/III 96.61** 1.26** -0.02**

.29* -.24* -.91** 6 Within-year 92.09** 0.68** -0.01

.35 -.28 -.91**

.97 .82 .69 (197) Tier II/III 94.67** 0.57** 0.00

.13 .01 -.93** 7 Within-year 109.99** 0.25 0.01

.64** -.64** -.86**

.98 .97 .92 (107) Tier II/III 110.30** 0.45* 0.00

.63** -.71** -.95** 8 Within-year 114.44** -0.50* 0.03**

.20 -.07 -.98**

.98 .64 .49 (72) Tier II/III 114.37** 0.11 0.01

.40* -.42 -1.00** Discussion

Tier II/III interventions begin in the fall; less identification of at-risk students beyond fall.

• Implications for winter/spring benchmarks in RTI system: Lost opportunities for educators and students for lack of RTI system training?

As hypothesized, growth rates are lower than 50th percentile and average empirical trajectories (e.g., Nese et al., 2013), except for Grades 4 & 5.

In all grades but sixth, linear growth rates greater in magnitude for the Tier II/II metric than the Within-year metric.

• Tenuous evidence for effective interventions.

Limitations

• Selected sample very specific.

• Intervention information (if any) unknown.

Discussion Tier II/III interventions begin in the fall; less

identification of at-risk students beyond fall.

• Implications for winter/spring benchmarks in RTI system: Lost opportunities for educators and students for lack of RTI system training?

As hypothesized, growth rates are lower than 50th percentile and average empirical trajectories (e.g., Nese et al., 2013), except for Grades 4 & 5.

In all grades but sixth, linear growth rates greater in magnitude for the Tier II/II metric than the Within-year metric.

• Tenuous evidence for effective interventions.

Limitations

• Selected sample very specific.

• Intervention information (if any) unknown.

Table 1. Fixed Effects and Within- and Between-Model Correlations for the Quadratic Within-year and Tier II/II Growth Models

Figure 1. Percent of Students First ORF Assessment for Each Month Across Grades 1-8

Figure 2. Predicted mean ORF scores in words correct per minute (wcpm) by grade based on the Within-year and Tier II/II time metrics, and observed mean ORF benchmark scores at the 20th and 50th percentiles for the easyCBM system.

0

30

60

90

120

150

180

26-Jul 14-Sep 3-Nov 23-Dec 11-Feb 1-Apr 21-May 10-Jul

Flu

ency

(w

cpm

)

Grade 1

0

30

60

90

120

150

180

26-Jul 14-Sep 3-Nov 23-Dec 11-Feb 1-Apr 21-May 10-Jul

Flu

ency

(w

cpm

)

Grade 2

0

30

60

90

120

150

180

26-Jul 14-Sep 3-Nov 23-Dec 11-Feb 1-Apr 21-May 10-Jul

Flu

ency

(w

cpm

)

Grade 3

0

30

60

90

120

150

180

26-Jul 14-Sep 3-Nov 23-Dec 11-Feb 1-Apr 21-May 10-Jul

Flu

ency

(w

cpm

)

Grade 4

0

30

60

90

120

150

180

26-Jul 14-Sep 3-Nov 23-Dec 11-Feb 1-Apr 21-May 10-Jul

Flu

ency

(w

cpm

)

Grade 5

0

30

60

90

120

150

180

26-Jul 14-Sep 3-Nov 23-Dec 11-Feb 1-Apr 21-May 10-Jul

Flu

ency

(w

cpm

)

Grade 6

0

30

60

90

120

150

180

26-Jul 14-Sep 3-Nov 23-Dec 11-Feb 1-Apr 21-May 10-Jul

Flu

ency

(w

cpm

)

Grade 7

0

30

60

90

120

150

180

26-Jul 14-Sep 3-Nov 23-Dec 11-Feb 1-Apr 21-May 10-Jul

Flu

ency

(w

cpm

)

Grade 8