Sound Partners Kindergarten

Study: Vadasy & Sanders (2010)

Vadasy, P.F., & Sanders, E.A. (2010). Efficacy of supplemental phonics instruction for low-skilled kindergarteners in the context of language-minority status and classroom phonics instruction. Journal of Educational Psychology, 102, 786-803.
Descriptive Information Usage Acquisition and Cost Program Specifications and Requirements Training

Sound Partners Kindergarten is a supplemental Tier 2 intervention that provides integrated and explicit instruction in phonemic and alphabetic skills, including phonemic decoding skills and assisted oral reading practice in decodable texts. The program consists of a set of 70 scripted lessons (with 7-8 activities per lesson) with matched decodable texts that are used during 30-minute tutoring sessions (typically provided 4 days a week, for about 20 weeks). Tutoring is typically conducted during the school day, outside the classroom in a quiet school space. On average, instruction includes 20 minutes of phonics activities and 10 minutes of oral reading practice in decodable texts. Tutors adjust the rate of progress through the phonics content to meet students’ needs. Lesson activities include the following:

  • Letter-sound correspondences
  • Syllable and phoneme segmenting
  • Decoding and spelling
  • Irregular word instruction
  • Phoneme blending
  • Alphabet naming practice
  • Assisted oral reading practice

The effectiveness of Sound Partners Kindergarten tutoring depends upon careful implementation. To ensure a successful Sound Partners Kindergarten tutoring program, schools should have: space for tutors to work with students; paraprofessional tutors available to work with students (one-on-one, 30 minutes/day, 4 days/week, for the entire school year); and a program supervisor to oversee program implementation.

A program supervisor (reading specialist, resource teacher, or tutor coordinator) is essential to successful program implementation. Responsibility for supervising Sound Partners Kindergarten tutoring can usually be incorporated into the supervisor’s regular school job description, and may require 1 to 3 hours per week of time.

Sound Partners Kindergarten is intended for use in kindergarten. It is designed for use with students with learning disabilities, English language learners, and students at risk for reading difficulties. The academic area of focus is reading (including phonological awareness, alphabetic knowledge, and phonics/word study).

In the Seattle School District, where all of the developer’s studies have been based, it is estimated that about 20 schools use Sound Partners Kindergarten as a supplemental intervention. With funds from a U.S. Department of Education OSEP outreach grant and local private funding from 2004-present, the developer conducted over 70 Sound Partners Kindergarten trainings. These trainings served about 300 paraeducators, 170 teachers, and other parent organizations. The developer has also conducted trainings in Nebraska (serving 27 school districts); Oakland, CA; Holly Springs, MS; Tucson, AZ; and Palm Springs, CA. 

Where to obtain: 

Cambium Learning Group

17855 Dallas Parkway, Suite 400

Dallas, TX 75287

Phone: 303-651-2829

Website:www.voyagersopris.com/curriculum/subject/literacy/sound-partners

Cost:

$89.95 for master set including three copies of lesson materials.

Initial cost per student for implementing program: lesson notebooks available from the publisher, and approximately $25 for two sets of decodable Bob Books storybooks (Maslen: Scholastic Press) available from a bookstore/publisher.

Replacement cost per student for subsequent use: Lessons can be copied. Storybooks can be replaced as needed.

The estimated cost for program implementation will vary, based upon the type of implementer (hourly tutor, salaried paraeducator with benefits, or certificated teacher), the number of students a school plans to serve (and consequently how many students each tutor will serve). Based on average hourly tutor rates at Seattle elementary sites, the estimated Sound Partners Kindergarten cost per student is $800, annually.

School staff interested in assistance with training can contact the developers at Washington Research Institute.

Sound Partners Kindergarten is designed for use with individual students. In a randomized study, dyad tutoring has also been found to be an effective implementation.

Sound Partners Kindergarten takes 30 minutes per session with a recommended four sessions per week for 20 weeks.

The program includes a detailed tutor manual. The program is not affiliated with a basal text, but it is coordinated with a set of decodable storybooks that include the Bob books (Maslen) published by Scholastic. No special technology is required.

Typically a half-day training is provided for non-teacher tutor trainees.

This program is designed to be implemented by paraeducator tutors. Trainers model use of the scripted lessons and supervise practice in use of each component. Follow-up coaching is typically provided either locally by our trainers or by the school contact who is trained to be the on-site coach/supervisor. The program includes a coaching guide as well as a format for recording fidelity of use.

Instructors must be professionals (in the case of the tutor trainer) or paraprofessionals (in the case of the typical tutor implementer). The tutors often have little background in beginning reading instruction, but the tutor trainers and supervisors should have a background in beginning reading acquisition and instruction.

The training manual and training procedures were used in several RCT studies of the program as well as in an earlier first- grade version of the program, and in a second-third- grade version of the program (tutoring in advanced word reading skills).

Although on-site technical support in training local tutors is not typically needed, Seattle-area trainers are available to provide long-distance support to schools beginning to use the program.

 

Participants: Unconvincing Evidence

Sample size: 148 kindergarten students across six schools in four districts in third grade (67 students in the treatment group [38 LM + 29 non-LM] and 81 students in the control group [46 LM + 35 non-LM]; 169 students initially pretested)

Risk Status: Screening for students at risk included three measures. The first two were measures of alphabetic knowledge, the number of letter sounds and letter names produced out of 52 randomly ordered upper-case English letters (Fuchs et al., 2001). The third was a test of phonological awareness (Sound Matching subtest from the Comprehensive Test of Phonological Processing (CTOPP) Wagner, Torgesen, & Rashotte, 1999). Both alphabetic knowledge (Chall, 1967; Share, Jorm, Maclean, & Matthews, 1984; Tunmer, Herriman, & Nesdale, 1988) and phonological awareness  (Adams, 1990; Bus & van IJzendoorn, 1999; Shankweiler & Liberman, 1989; Share, 1995; Snowling, 1991; Stanovich, 2000) are well-established early predictors of reading outcomes in monolingual English speakers, and phonological awareness and alphabetic skills are related to English word reading in LM children (Chiappe, Siegel, & Wade-Wooley, 2002; Lesaux, Koda, Siegel,& Shanahan, 2006; Lipka & Siegel, 2007).

Demographics:

 

Program

Control

p of chi square

Number

Percentage

Number

Percentage

Grade level

  Kindergarten

 67 [38+29]

 100%

 81 [46+35]

100% 

 

  Grade 1

 

 

 

 

 

  Grade 2

 

 

 

 

 

  Grade 3

 

 

 

 

 

  Grade 4

 

 

 

 

 

  Grade 5

 

 

 

 

 

  Grade 6

 

 

 

 

 

  Grade 7

 

 

 

 

 

  Grade 8

 

 

 

 

 

  Grade 9

 

 

 

 

 

  Grade 10

 

 

 

 

 

  Grade 11

 

 

 

 

 

  Grade 12

 

 

 

 

 

Race-ethnicity

  African-American

21 [7+14]

31.3%

13 [6+7]

16.0%

 

  American Indian

 

 

 

 

 

  Asian/Pacific Islander

13 [9+4]

19.4%

17 [11+6]

21.0%

 

  Hispanic

23 [20+3]

34.3%

36 [29+7]

44.4%

 

  White

8 [0+8]

11.9%

13 [0+13]

16.0%

 

  Other

2 [2+0]

3.0%

2 [0+2]

2.5%

 

Socioeconomic status

  Subsidized lunch

56 [35+21]

83.6%

65 [43+22]

80.2%

 

  No subsidized lunch

11 [3+8]

16.4%

16 [3+13]

19.8%

 

Disability status

  Speech-language impairments

1 [0+1]

1.5%

4 [1+3]

4.9%

 

  Learning disabilities

 

 

 

 

 

  Behavior disorders

 

 

 

 

 

  Intellectual disabilities

 

 

 

 

 

  Other

 

 

 

 

 

  Not identified with a disability

 

 

 

 

 

ELL status

  English language learner

37 [37+0]

55.2%

43 [43+0]

53.1%

 

  Not English language learner

30 [1+29]

44.8%

38 [3+35]

46.9%

 

Gender

Female

34 [19+15]

50.7%

32 [18+14]

39.5%

 

Male

33 [19+14]

49.3%

49 [28+21]

60.5%

 

Training of Instructors: All paraeducator tutors (paraeducators) were recruited from their school communities based on their interest in working with children, prior tutoring and school volunteer experience, and scheduling flexibility. The 23 participating paraeducators were mostly non-minority (74%) and female (83%), and varied in their age, educational levels, general tutoring experience, and experience working with kindergartners. Tutors averaged a mode of 35-44 years old (ranging from 18 to over 55), and ranged in educational level from high school diploma to master’s degree, with a modal level of a bachelor’s degree (39%). (The average education level of paraeducators in this study is similar to that recommended under NCLB for supplemental education services.) Prior to the study, the range of paraeducator tutoring experience ranged from 0 (22%) to 10 or more years (26%), with an average of M=4.52 years (SD = 5.06). Most paraeducators (72%) had at least one year previous experience working with early grade levels (K-2) averaging M = 3.22 years (SD = 4.04, Range = 0 to 15 years). All paraeducators were hired as district employees and paid by the schools with funds provided by the research grant.

Researchers provided an initial 2-hour training session to describe each lesson activity, and model paraeducator/student behaviors, errors, and error correction strategies. Trainees were paired together to practice each activity while trainers provided feedback and responded to questions. Follow-up training was provided throughout the intervention, with added coaching for paraeducators with less experience and/or low initial intervention fidelity ratings. Less experienced tutors received from 0.5 to 3.0 hours of coaching during the intervention, averaging 1 hour of additional on site coaching. All coaches also conducted fidelity observations, described below.

Research staff were teachers or experienced tutors with backgrounds in reading instruction. They provided ongoing coaching and modeling of appropriate scaffolding to help paraeducators provide the type of support at-risk students often require to accomplish phonemic segmenting, decoding, and encoding tasks (Foorman & Torgesen, 2001; Juel, 1996). Tutors were instructed that, before students moved beyond lesson 10 (when modeling of the phonemic decoding task decreases), each student should demonstrate at least 70% mastery of all letter sounds introduced, as well as an understanding of phoneme decoding (although not necessarily at full mastery). Tutors who worked with LM students were instructed to provide judicious incidental vocabulary instruction without compromising the intensity of phonics instruction time.

Design: Convincing Evidence

Did the study use random assignment?: Yes.

If not, was it a tenable quasi-experiment?: Not applicable.

If the study used random assignment, at pretreatment, were the program and control groups not statistically significantly different and had a mean standardized difference that fell within 0.25 SD on measures used as covariates or on pretest measures also used as outcomes?: Yes.

If not, at pretreatment, were the program and control groups not statistically significantly different and had a mean standardized difference that fell within 0.25 SD on measures central to the study (i.e., pretest measures also used as outcomes), and outcomes were analyzed to adjust for pretreatment differences?: Not applicable.

Were the program and control groups demographically comparable at pretreatment?: Yes.

Was there attrition bias1 ?: No.

Did the unit of analysis match the unit for random assignment (for randomized studies) or the assignment strategy (for quasi-experiments)?: Yes.

1 NCII follows guidance from the What Works Clearinghouse (WWC) in determining attrition bias. The WWC model for determining bias based on a combination of differential and overall attrition rates can be found on pages 13-14 of this document: http://ies.ed.gov/ncee/wwc/pdf/reference_resources/wwc_procedures_v2_1_standards_handbook.pdf

 

Fidelity of Implementation: Convincing Evidence

Describe when and how fidelity of treatment information was obtained: Six research staff were trained to conduct on-site fidelity observations of paraeducators with their assigned students. Fidelity observations involved a 5-point rating scale ranging from 1 (never implements correctly) to 5 (always implements correctly) for each of the instructional components. After training but prior to field observations, researchers viewed six videotaped tutoring sessions of paraeducators implementing instruction with students; students in the videotaped sessions were selected to represent a range of reading skills we expected to see in the field, with ages ranging from 4- to 7-years-old. To determine interrater reliability, we calculated the internal consistency of the observers’ mean implementation ratings for the videotaped sessions (using observers’ ratings as items and each videotape as subjects): Cronbach’s alpha was 0.97.

Provide documentation (i.e., in terms of numbers) of fidelity of treatment implementation: After establishing reliability, researchers conducted a total of 156 observations for the 23 paraeducators over the course of the intervention, averaging 6.78 observations per tutor. Fidelity ratings averaged M = 4.41 (SD = 0.57).

Measures Targeted: Convincing Evidence

Measures Broader: Convincing Evidence

Targeted Measure Score type & range of measure Reliability statistics Relevance to program instructional content

Peabody Picture Vocabulary Test-IIIA (PPVT-IIIA)

Students select a picture that best illustrates the meaning of an orally presented stimulus word. Testing is discontinued after the student misses 8 out of 12 items. The raw score (number of items correct) is converted to a standard score.

Reliability reported in the test manual is 0.94 for 5-year-olds. For this sample, internal consistency was 0.97 for all students, and 0.95 for LM students.

PPVT-IIIA measures Receptive vocabulary.

Mean of two naming measures: Letter Names and Letter Sounds

Both our letter names and letter sounds measures use all 26 letters of the alphabet twice (once in upper-case and once in lower-case). Letters were randomly sorted (upper and lower-case together) and presented on a single page in six rows in comic sans font, which allowed students to better differentiate between the lower-case letter L and upper-case letter I in particular. For the letter names task, students are asked to name as many letters on the page as they can. For the letter sounds task, students are asked to produce the sounds that each letter on the page represented. The number correctly named or produced, respectively, was divided by the number of seconds the students took to finish the 52 items. This value was then multiplied by 60 to obtain letter names or sounds correct per min. For letter sounds, we considered only hard consonants and short vowels as correct.

For our sample, internal consistency was 0.98 (for all students and LMs only), and at posttest was 0.96 and 0.97 for letter names and sounds, respectively (for LMs, 0.97 for both).

Letter Names and Letter Sounds measure alphabetic knowledge.

Wide Range Achievement Test-Revised (WRAT-R) Spelling subtest

This test requires the student to copy marks/symbols, print his/her name, and print a list of dictated words, with testing discontinued after 10 consecutive incorrect responses. Similar to Fuchs et al. (2001), however, we applied the Tangel and Blachman (1992) developmental scoring rubric to all words attempted (within normal test administration guidelines). This rubric allowed us to credit students for partial and less phonemically sophisticated responses. Items were scored from 0 (random string of letters) to 6 (entire word correctly spelled).

Test-retest reliability reported for ages 7.0-7.5 is 0.97.

Internal consistency for this sample was 0.89 and 0.96 at pretest and posttest, respectively. Reliability for LM students was similar: 0.85 at pretest and 0.95 at posttest.

WRAT-R Spelling subtest measures spelling.

Mean standard score of the Word Attack and Word Identification subtests from the Woodcock Reading Mastery Test- Revised/Normative Update (WRMT-R/NU)

The Word Attack subtest includes 50 nonwords that increase in difficulty. Testing is discontinued after 6 consecutive incorrect responses.

The Word Identification subtest consists of 106 words that increase in difficulty. Testing is discontinued after 6 consecutive incorrect responses.

Word Attack subtest: Split-half reliability for first graders reported in the manual is 0.94. For this sample, internal consistency at posttest was 0.91 (0.86 for LMs). Sample reliability could not be computed at pretest due to extreme floor effects.

Word Identification subtest: Split-half reliability reported in the manual for first graders is 0.98. For this sample, internal consistency at pretest was 0.88 (0.85 for LMs), and at posttest was 0.95 (0.93 for LMs).

The subtests measure word reading.

Mean words correctly read in one min on two grade-level story passages

Both stories are from the Primary Phonics series. One passage was considered more decodable (“Mac and Tab”) and the other was considered less decodable (“Ben Bug”). Students read each passage aloud for one min. Words omitted, substituted, and hesitations of more than three sec are scored as errors. Words self-corrected within three s were scored as accurate. The score is calculated as the number of words correctly read in one min.

For our kindergarten sample, internal consistency was 0.97 for both the more and less decodable passages (for all students as well as LMs only). The correlation between the two passages (which may be considered as a form of alternate form reliability) across all students was 0.91 (for LMs, 0.90).

This measures passage reading fluency.

Broader Measure Score type & range of measure Reliability statistics Relevance to program instructional content

WRMT-R/NU Passage Comprehension subtest

The student is asked to silently read a short passage, and then orally provide a missing key word. (For each blank, the student is asked to supply a word appropriate in the context of the passage.) Acceptable responses are listed on the examiner’s easel page, and testing is discontinued after 6 consecutive incorrect responses.

Internal consistencies reported by test developers range from 0.94 to 0.97. For this sample, internal consistency was 0.86 (for LMs, 0.70).

This measures comprehension.

Composite standard score of three subtests from the Comprehensive Test of Phonological Processing (CTOPP) Blending Words (pretest), Elision (pretest), and Sound Matching (screening).

During the Blending Words subtest, the student is asked to listen to parts of words and blend them together to make a whole word. This subtest has 20 items, and testing is discontinued after the student misses 3 items in a row or when the student completes all items.

During the Elision subtest, the student is asked to listen to the sounds in a spoken word and is then asked to say the word without one or more of its sounds, creating a new word (e.g., the student is asked, “Say the word spider without saying der.”). This subtest has 20 items, and testing is discontinued after the student misses 3 items in a row or when all items are completed.

The Sound Matching subtest has two parts: in Part I, the tester says a word and asks the student to say, out of three word choices, the word that starts with the same sound as the initial word (e.g., the student is asked, “Which word starts with the same sound as sock? Sun, cake, or bear?”). Part II of this subtest asks the student to say, out of three word choices, the word that ends with the same sound as the initial word. This subtest has 20 items, and testing is discontinued after 4 out of 7 items are incorrect.

Blending Words subtest: Test-retest reliability reported in the manual for 5- to 7-year-olds is 0.88. For this sample, internal consistency at pretest was 0.78, and at posttest was 0.83 (correspondingly, 0.74 and 0.82 for LMs).

Elision subtest: Test-retest reliability reported in the test manual for 5- to 7-year-olds is 0.88. For this sample, internal consistency at pretest was 0.79, and at posttest was 0.83 (for LMs, we found 0.65 and 0.80 at pretest and posttest, respectively).

Sound Matching subtest: Test-retest reliability reported in the manual for 5- to 7-year-olds is 0.83. For our sample, internal consistency was 0.71 at screening (0.66 for LMs), and at posttest was 0.90 (0.88 for LMs).

CTOPP Blending Words, Elision, and Sound Matching measure phonological awareness.

 

Number of Outcome Measures: 2 Prereading, 3 Reading, 1 Writing

Mean ES - Targeted: 0.76*

Mean ES - Broader: 0.40*

Effect Size:

Targeted Measures

Construct Measure Effect Size
Prereading Mean Letter Names/Sounds 0.75***
Reading Word Reading Mean WRMT 0.73***
Reading Passage Reading Fluency 0.97***
Writing WRAT-R Spelling 0.60***

Broader Measures

Construct Measure Effect Size
Prereading CTOPP Phonological Awareness 0.20
Reading Comprehension: WRMT-R/NU Passage Comprehension 0.60***

 

Key
*        p ≤ 0.05
**      p ≤ 0.01
***    p ≤ 0.001
–      Developer was unable to provide necessary data for NCII to calculate effect sizes
u      Effect size is based on unadjusted means
†      Effect size based on unadjusted means not reported due to lack of pretest group equivalency, and effect size based on adjusted means is not available

 

Visual Analysis (Single Subject Design): N/A

Disaggregated Data for Demographic Subgroups: No

Disaggregated Data for <20th Percentile: Yes

 

Sample Disaggregated for Students with Alphabetics < 15

Targeted Measures

Construct Measure Effect Size
Prereading Mean Letter Names/Sounds 0.79***
Reading Word Reading Mean WRMT 0.80***
Reading Passage Reading Fluency 1.31***
Writing WRAT-R Spelling 0.47*

Broader Measures

Construct Measure Effect Size
Prereading CTOPP Phonological Awareness 0.25
Reading Comprehension: WRMT-R/NU Passage Comprehension 0.68***

 

Key
*        p ≤ 0.05
**      p ≤ 0.01
***    p ≤ 0.001
–      Developer was unable to provide necessary data for NCII to calculate effect sizes
u      Effect size is based on unadjusted means
†      Effect size based on unadjusted means not reported due to lack of pretest group equivalency, and effect size based on adjusted means is not available

 

Administration Group Size: Individual

Duration of Intervention: 30 minutes, 4 times a week, 18 weeks

Minimum Interventionist Requirements: Paraprofessional, 4 hours of training

Reviewed by WWC or E-ESSA: WWC & E-ESSA

What Works Clearinghouse Review

Beginning Readers Protocol

Effectiveness: Sound Partners was found to have positive effects on alphabetics, fluency, and comprehension and no discernible effects on general reading achievement on beginning readers.

Studies Reviewed: 7 studies meet standards out of 11 studies total

Full Report

Evidence for ESSA

English Learners

Program Outcomes: Two studies evaluated Sound Partners with English learners. One involved kindergartners and one first graders. Effect sizes were significantly positive compared to controls at both grade levels. The effect size across Woodcock and CTOPP measures was +0.60 for kindergartners, +0.15 for first graders. Follow-up studies found that these outcomes were still seen two years later, on Word Reading and Comprehension. The positive outcomes qualify Sound Partners for the ESSA “Strong” category, and for the “Solid Outcomes” rating (effect size of at least +0.20 over at least two studies).

Number of Studies: 2

Average Effect Size: 0.36

Full Report

Struggling Readers

Program Outcomes: Two studies, one at the kindergarten level and the other at the first grade level, qualified for the review. The average effect size was +0.58 on measures from the Woodcock Reading Mastery Test and CTOPP. These met the criteria for the ESSA “Strong” category, and for “Solid Outcomes” (two studies with effect sizes of at least +0.20). Follow-up studies of the kindergartners and first graders both found positive effects maintained two years later on word reading and comprehension.

Number of Studies: 4

Average Effect Size: 0.58

Full Report

 

Other Research: Potentially Eligible for NCII Review: 4 studies

Jenkins, J. R., Peyton, J. A., Sanders, E. A., & Vadasy, P. F. (2004). Effects of reading decodable texts in supplemental first-grade tutoring. Scientific Studies of Reading, 8(1), 53–86.
 

Mooney, P. J. (2003). An investigation of the effects of a comprehensive reading intervention on the beginning reading skills of first graders at risk for emotional and behavioral disorders (Doctoral dissertation, University of Nebraska–Lincoln, 2003). Dissertation Abstracts International, 64(05A), 85–1599.
 

Vadasy, P. F., Jenkins, J. R., & Pool, K. (2000). Effects of tutoring in phonological and early reading skills on students at risk for reading disabilities. Journal of Learning Disabilities, 33(6), 579–590.
 

Vadasy, P. F., Jenkins, J. R., Antil, L. R., Wayne, S. K., & O’Connor, R. E. (1997a). The effectiveness of one-to-one tutoring by community tutors for at-risk beginning readers. Learning Disability Quarterly, 20(1), 126–139.