11. Efficacy Research - Utah's Reading Intervention Program

AID 1833282 · View on Simbli

Agenda Item

g. Purchase of Amira Tutor, Supplemental High-Dosage Tutoring (Not to exceed $600,000)

Summary: Presented by: Ms. Stacy E. Stepney, Chief Academic Officer, Division of Curriculum and Instruction
Request: It is requested that the Board of Education approve the purchase of Amira Tutor, a high-dosage, research-based tutoring program to support foundational reading skills, fluency, and reading comprehension development in kindergarten through third grade, in an amount not to exceed $600,000.
Why: The DeKalb County School District (DCSD) administers the Northwest Evaluation Association Measure of Academic Progress (MAP) three times per year. The 2025 spring MAP Fluency data results in the areas of phonics and word recognition are as follows:




Phonics and Word Recognition


Grade
Below Expectation
Approaching Expectation
Meets Expectation
Exceeds Expectation


Kindergarten
27%
16%
25%
32%


First
52%
46%
2%
0%


Second
96%
0%
4%
0%


Third
95%
0%
5%
0%




Amira ISIP is the Georgia Department of Education’s approved reading and dyslexia screener provided to local school districts at no cost. The District is seeking to adopt and purchase Amira Tutor, a supplemental tutoring program that algins directly to the Amira Assess. Amira Tutor will provide high-dosage, evidence-based, AI-guided, 1:1 tutoring to support teaching and learning. Real-time micro interventions are individualized and use research-based techniques and explicit decoding strategies to enhance foundational reading skills.

Pursuant to Board Policy DJE (III.D.3.g.2), the purchase of Amira Tutor does not require a competitive bid because it meets the policy definition of supplemental resources needed for instruction. In addition, Amira Tutor was evaluated and selected by qualified professional personnel based on sound pedagogical judgment and in the school district’s best interest.
Details: Amira ISIP provides a comprehensive and culturally responsive assessment system that evaluates key early literacy components such as phonological awareness, decoding, fluency, and comprehension-many of which can be completed in 15-20 minutes. The voice-enabled technology listens as students read aloud, detects errors with high reliability and validity, and generates actionable data reports that support differentiated instruction and early identification of reading difficulties, including dyslexia risk.

Amira acts in three essential roles: a dyslexia screener and diagnostic tool, a skilled instructional assistant, and a personalized reading coach. These roles are particularly valuable for early learners who benefit from responsive, individualized practice grounded in the science of reading.

For educators, Amira Tutor enhances classroom efficiency by allowing small group, whole group, or individual support by delivering immediate feedback through an AI-powered 1:1 tutor. For students, Amira Tutor fosters independence and engagement through interactive read-aloud experiences and real-time positive reinforcement. This tool is especially effective for supporting diverse learners thanks to its universal design features and multilingual sensitivity.

Overall, Amira Tutor empowers K-3 teachers with the data and tools needed to deliver responsive instruction, monitor progress, and close early literacy gaps, making it an asset in building foundational reading proficiency across all student groups. The selection of the Amira Tutor followed a two-round review process:


Round 1 (March 17, 2025): The five GADOE approved vendors presented their universal reading and dyslexia screener in person. A total of 15 participants attended, including central office coordinators, directors, assistant superintendents, and chiefs. Participants provided structured feedback and posed questions regarding instructional design, alignment with science of reading, and support for differentiated instruction of each resource.



Round 2 (April 3, 2025): Vendors addressed specific questions that arose in Round 1.
Financial impact: The total contract amount will not exceed $600,000.




120.2100.553200.26021.7210.1613.8010.035.2025
GEER II Dyslexia
Communication-Web Subscription


100-1000-553200-00011-7580-9990-8010-035-000
Communication-Web Subscription


100-2220-530000-00011-7580-9990-8010-035-0000
Purchased Professional/Technical Services
Contact: Ms. Stacy E. Stepney, Chief Academic Officer, Division of Curriculum & Instruction, 678-676-0731
Dr. Sean R. Tartt, Deputy Chief Academic Officer, Division of Curriculum & Instruction, 678-676-0731
Dr. Penny Mosley, Assistant Superintendent K-5 Curriculum & Instruction, Division of Curriculum & Instruction, 678-676-0137
Dr. Lynn Angus Ramos, Director, Literacy, Division of Curriculum & Instruction, 678-676-0136
Mrs. Lummie Baker, Director, Educational Media and Instructional Materials, Division of Curriculum & Instruction, 678-676-2421
Effective: July 15, 2025- June 30, 2026
Status: Approved by Office of Legal Affairs
     2022
 Utah’s Reading Intervention Program
 Efficacy Research




The Science of Reading + AI = Growth
Efficacy Research

Evidence from Utah
                     Abstract
                     The Utah State Board of Educaton approves and funds
                     several intervention programs for use by Utah Districts.
                     Employing usage data at the student level, the State
                     then commissions independent research to analyze the
                     impact and effect size of these programs.

                     In 2022, for students using at the recommended
                     dosage, the State’s data showed that Amira’s effect
                     size was over .4, in Hattie’s Zone of Desired Effects.

                     The study shows gains on the Utah State Summative
                     Assessment for frequent users of Amira far in excess of
                     the average for Utah students employing other
                     interventions.



                                                                                1
What Does This Research Prove About Amira’s Efficacy?

From the “Evaluation & Training Institute” Report:
Mean score differences and effect sizes of matched program students who met Amira’s
recommendations across both average weekly
minutes and total weeks. As shown below, all treatment students in grades 1st - 3rd exhibited
higher predicted mean scores than their matched control counterparts. Second grade students
exhibited the highest mean score differences, with treatment students scoring 21 points higher
than their control counterparts, on average. Among this highest use analytic sample, first and
second graders had effect sizes within the large effect size range (g= 0.41 and g=0.45,
respectively), while third grade students had an effect size within the medium range (g=0.25).



Who Conducted and Published the Research?
The research was sponsored by the Utah State Board of Education. The study
was conducted by The Evaluation and Training Institute, a non-profit, nonpartisan research
organization, is dedicated to improving academic excellence and social-emotional well being
from preschool to adulthood.




                                                                                                 2
Utah students who worked at recommended dosage saw remarkable gains. As shown from the
research below, student gains exceed .40 effect size.



                                                                   MRU Sample Predicted End-of-Year Acadience Reading Composite
                                                                                           Mean Scores

                                                                          Grade               Ctrl                      Tr              Dif.                ES
                                                                                    Mean               SE     Mean              SE
                                                            Kindergarten             ISS

Background                                                  First Grade
                                                            Second Grade
                                                                                     76.37
                                                                                    284.92
                                                                                                      0.44
                                                                                                      0.71
                                                                                                              87.16
                                                                                                              305.89
                                                                                                                               1.94
                                                                                                                               3.27
                                                                                                                                       10.79
                                                                                                                                       20.97
                                                                                                                                                          0.41
                                                                                                                                                          0.45

(From Utah’s Report)                                        Third Grade             384.06            1.03    403.87           4.77    19.81              0.25




“The Early Intervention Software Program (EISP) was
                                                                        Even students who worked at only 80% of the
designed to increase the literacy skills of all students
in K-3 through adaptive computer-based literacy                     recommended dosage saw meaningful improvements.
software.                                                       MRU80 Sample Predicted End-of-Year Acadience Reading Composite
                                                                                        Mean Scores
The program provided Utah’s Local Education                              Grade               Ctrl                      Tr              Dif.              ES

Agencies (LEAs) with an option to select among four                                Mean               SE     Mean             SE
                                                           Kindergarten             ISS
adaptive computer-based programs. State-wide                                       75.21             0.40    79.42            0.99    4.21               0.15
                                                           First Grade
program implementation provided the opportunity            Second Grade            270.00            0.69    287.05           1.74    17.05              0.30

for large numbers of students to receive program           Third Grade             380.88            0.90    403.32           2.29    22.44              0.34

benefits. ”

                                                                                   2022-2023 Program Enrollment by Grade
                                                                    Kindergarten              First Grade             Second Grade            Third Grade

                                                                           179                      7,922                   8,010                8,016
                                                                                                                                                                 3
How Was the Study Conducted?
Statistical Modeling of Program Impacts on Acadience Test Scores. Ordinary least squares (OLS)
regression models were computed for each analytic sample. The OLS models predicted the
differences in treatment and control groups’ end-of-year group mean scores, while controlling for
students’ beginning-of-year (BOY) reading scores and key demographics, gender, race, ELL                 1.5 Years Of Growth In 22/23 SY.
status, SPED designation, and poverty status. We examined treatment effects for each analytic
sample based on their usage and grade.
1 Students in kindergarten, 2nd and 3rd grade were matched on reading composite scores (BOY
Comp) and students in 1st grade were matched on nonsense word fluency, correct letter sounds
(NWF-CLS) scores.

Amira showed significant treatment effects for students in first through third
grade among students who met Amira’s usage requirements, and even demonstrated
benefits for those meeting a more lenient usage criteria.

The following results are broken up into two different usage groups of K-3rd grade students and
their matched control counterparts, (1) students who met Amira’s recommended weeks and
average minutes, and (2) students who met 80% of recommended weeks and average minutes.
This section is focused on participants who engaged with the Amira program most closely aligned
to the recommendation. Results for the third usage group (ITT), which included the students
whose time with the program fell far below the recommend levels, can be found in Appendix B.
                                                                                                                   Other Edtech
To determine if the mean score differences could be interpreted as meaningful, we examined                                        Amira in 2023
                                                                                                                     in 2022
their. effect sizes. Effect sizes show the magnitude of the difference between two groups on an
                                                                                                    1st Grade          .09             .41
outcome and are often interpreted as meaningful if they reach a certain minimum threshold. We
adapted a. set of effect size benchmarks based on categories from Kraft (2020) that were            2ndGrade           -.03            .45
adjusted for early. literacy outcome measures: less than 0.10 is small, 0.10 to less than .30 is
medium, and .30 or. greater is large (M. Kraft, personal communication, October 13, 2023).          3rd Grade          .09             .25



                                                                                                                                                  4
Where Can One Access the Study?

    Background is available on the
    Amira Learning website.

                                                                                                  Summary &
  Related Studies                                                                                 Conclusion
  Several other studies support Amira’s impact on literacy growth:

             Evaluation and Training Institute (2023): Found Amira        The research shows that Amira delivers effect size in
             significantly outperformed a control group in Utah’s Early   Hattie’s Zone of Desired Effect when a student works
             Intervention Reading Software Program (Read Report).
                                                                          with the program more than 30 minutes a week.
                                                                          The researchers concluded:
             Mostow et al. (CMU): Demonstrated that AI-                    ”Students served by Amira outperformed the
             powered reading tutors can be as effective as                students who were not. Further, the students who
             human tutors (Read Study).                                   were able to use with the software as it was intended
                                                                          by Amira also showed. greater end-of-year literacy
                                                                          scores relative to those participating below the
             Poulsen et al. (UBC): Showed Amira’s effectiveness for
                                                                          recommended usage levels in the program.”
             bilingual students (Read Study)




                                                                                                                                  5