Athena R' Lentini

1
FALSE-POSITIVE RISK IDENTIFICATIONS IN
KINDERGARTEN LITERACY ASSESSMENT A
RESPONSE-TO-INTERVENTION SOLUTION

• Athena R. Lentini
• University of Connecticut

Council for Exceptional Children, DLD Reception,
2008
2
Who is at risk?

• Factors associated with reading risk at fall of
  kindergarten
• Low letter name knowledge (e.g., Catts, Fey,
  Zhang, and Tomblin, 2001 Scarborough, 1998)
• Low phonological awareness (e.g., Catts, et al.,
  2001 OConnor and Jenkins, 1999 Scarborough,
  1998)
• When these measures are used assign risk status,
  false positive identification rates can be high
  (average 45 Scarborough,1998)

3
Who is at risk?

• K student 1 LNF 0

• K student 2 LNF 0

• Two years of academic preschool
• Many childrens books
• Read together every night

• No preschool
• Limited family interaction around print

Low knowledge despite multiple opportunities to
learn
Low knowledge, but may learn appropriately given
the opportunity
4
Who is at risk?

• Schools using these risk factors may include many
  students who arent really at risk.
• Can students who are likely not truly at risk
  (false positive risk classifications) be
  identified early in the intervention process and
  successfully returned to general classroom
  instruction?

5
Data Year 1

• The Early Reading Intervention Examination of
  Curriculum Efficacy, Durability, Replicability,
  and Intensity, Institute of Education Sciences
  (2006 2010)
• Deb Simmons, Texas AM University
• Michael Coyne, University of Connecticut

IES Grant R305G050121 to Texas AM University
6
Participants Year 1

• 61 kindergarten students
• Three New England school districts
• IDd as at risk in Fall 2006
• Low phonemic awareness, M 7.72, SD 1.19
  (CTOPP Sound Matching, standard score)
• Low letter naming fluency, M 4.02 SD 5.51
  (DIBELS Letter Naming Fluency)
• Early Reading Intervention, 2006 2007
• Focuses on Big Ideas PA, alphabetic principle
• 30 min/day, 5 days/week, small groups of 3 5
• Remained in intervention for entire academic year

7
False positive risk classification

• Evidence of sustained high achievement
• Scores gt 94 on all 4 curriculum mastery tests
• Outcome scores outside of risk status 30th ile
  or higher
• CTOPP Sound Matching
• WRMT Word ID, Word Attack, Passage Comprehension
• Met benchmark goals on DIBELS PSF and NWF

11/61 students were identified as possible false
positives
8
Research question 1

• Do these students differ from other students
  receiving reading intervention services on
  pretest measures administered in early
  kindergarten?

• CTOPP SM BW No differences
• DIBELS LNF Possible false-positive (10) gt At
  risk (3)

Score Range on Pretest Measures
9
Research question 2

• Are these students identifiable based on early
  mastery of the intervention curriculum?

• First curriculum mastery test
• 97 correct or higher captures 9 of 11 students
• Misclassifies 5 students
• Accuracy of Mastery Test Scores In Identifying
  False Positive Students

a Students are not classified as false-positives
using the cut-off score, but did meet
false-positive criteria. b Students are
classified as false-positives using the cut-off
score, but did not meet false-positive criteria.
10
Research question 3 Year 2

• Can students who have possible false-positive
  risk classifications be successfully returned to
  general classroom instruction?
• Participants
• 63 kindergarten students
• Same three New England school districts
• IDd as at risk in Fall 2007

11
Research question 3 Year 2

• Early Reading Intervention, 2007 2008
• 12 met 97 criteria score on first mastery test
  (after 9 weeks of intervention instruction) and
  were identified as possible false positives
• 8 were returned to general classroom instruction
  during January or February 2008
• 1 was returned to general classroom instruction
  in November 2007
• Current mean scores on DIBELS progress
  monitoring
• PSF M 49, SD 21.4
• NWF M 32, SD 11.8

12
Progress monitoring
13
Progress monitoring
14
Discussion

• Year 1 Students with possible false positive
  classifications can be IDd early in the
  intervention process
• High LNF scores identified two of 11students
  (18) at pretest
• Early curriculum mastery score of 97 or higher
  identified nine of 11students (81)
• Year 2 Most students with possible false
  positive classifications returned to general
  classroom instruction after 8 weeks of
  intervention meet or exceed DIBELS
  end-of-Kindergarten goals

15
Thank you

• Athena R. Lentini
• University of Connecticut
• Center for Behavioral Education and Research
• Athena.Lentini_at_UConn.edu
• For a copy of this presentation, please visit
• www.cber.uconn.edu