Improving Persistence through Goal Support in i-Ready: A Classroom-Randomized Crossover Trial

Last registered on September 09, 2026
View Trial History

Pre-Trial

Trial Information

General Information

Title
Improving Persistence through Goal Support in i-Ready: A Classroom-Randomized Crossover Trial
RCT ID
AEARCTR-0019569
Initial registration date
September 01, 2026

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
September 09, 2026, 3:39 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country
United States of America
Region

Primary Investigator

Name
Conrad Borchers
Affiliation
Vanderbilt University
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Kenneth R. Koedinger
PI Affiliation
Carnegie Mellon University
Contact Investigator
PI Name
Vincent Aleven
PI Affiliation
Carnegie Mellon University
Contact Investigator
PI Name
Danielle R. Thomas
PI Affiliation
Carnegie Mellon University
Contact Investigator
PI Name
Erin E. Gatz
PI Affiliation
Carnegie Mellon University
Contact Investigator

Additional Trial Information

Status
In development
Start date
2026-09-08
End date
2027-08-31
Keywords
Education
Additional Keywords
Goal orchestration, goal support, persistence, K-12 math, field experiments, student persistence, AI in education, learning analytics, teachers, computer-assisted learning, human-AI tutoring, high-impact tutoring
JEL code(s)
C93, I21
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This classroom-randomized crossover trial, involving approximately 31 classrooms and 670 students, evaluates two approaches to supporting student persistence and mathematics learning in i-Ready Mathematics. The first is Personalized Learning Squared (PLUS), a remote tutoring program in which tutors support students during i-Ready practice and conduct weekly goal support. The second is teacher-led Goal Orchestration (GO), delivered through the GO Platform, which provides teachers with weekly i-Ready data, personalized goal recommendations, progress information, reward eligibility reports, and tools for individual and small-group support. Business-as-usual (BaU) classrooms continue their usual i-Ready routines without PLUS tutoring or the GO Platform. During Fall 2026, each participating classroom is assigned to two different conditions across two study periods. The six possible sequences are PLUS-BaU, PLUS-GO, BaU-PLUS, BaU-GO, GO-PLUS, and GO-BaU. Randomization occurs at the classroom level within school-by-grade strata. The primary learning outcome is change in i-Ready Diagnostic scale score from the baseline Diagnostic administered before Period 1 to the endline Diagnostic administered after Period 2. Weekly i-Ready Personalized Instruction practice time and lessons passed are additional primary outcomes used to estimate within-period effects of PLUS, GO, and BaU. For the weekly outcomes, the study will estimate pairwise effects of PLUS, GO, and BaU within each study period. Analyses of Diagnostic growth will account for the randomized two-period treatment sequences, with treatment-level contrasts specified prospectively in the pre-analysis plan. Eligible students are in grades 5-8 and enrolled in participating classrooms at schools that use i-Ready Mathematics as part of regular mathematics instruction. Students are eligible for inclusion unless a parent or legal guardian opts them out of study participation.
External Link(s)

Registration Citation

Citation
Aleven, Vincent et al. 2026. "Improving Persistence through Goal Support in i-Ready: A Classroom-Randomized Crossover Trial." AEA RCT Registry. September 09. https://doi.org/10.1257/rct.19569-1.0
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
The study includes three conditions. In the PLUS condition, students receive remote tutoring during scheduled i-Ready practice, planned at approximately two 45-minute sessions per week. PLUS tutors support students' i-Ready work and conduct weekly goal setting and progress review using student learning data. Teachers receive weekly goal-achievement reports by email and can distribute rewards to students who meet their goals. In the Goal Orchestration (GO) condition, students do not receive PLUS tutoring. Teachers use the GO Platform to review updated i-Ready data, set or revise weekly student goals, review goal recommendations and progress, recognize goal achievement, and identify students or groups who may benefit from additional support. Teachers can configure goal metrics, thresholds, reward requirements, and the rule used to establish students' starting goals. Teachers can also help students review their current goals and progress through a dedicated view. In the business-as-usual (BaU) condition, classrooms continue their usual i-Ready instructional and practice routines without PLUS tutoring or access to the GO Platform. Each classroom receives two different conditions across two consecutive Fall 2026 study periods.
Intervention Start Date
2026-09-14
Intervention End Date
2026-12-04

Primary Outcomes

Primary Outcomes (end points)
Change in i-Ready Diagnostic scale score from the baseline Diagnostic administered before Period 1 to the endline Diagnostic administered after Period 2. Weekly i-Ready Personalized Instruction practice time and weekly lessons passed during each study period.
Primary Outcomes (explanation)
Diagnostic growth is the change in i-Ready Diagnostic scale score from the baseline administration before Period 1 to the endline administration after Period 2. No Diagnostic is administered at the crossover, so Diagnostic growth is observed only across the full study. Weekly practice time is measured as minutes of i-Ready Personalized Instruction activity, and weekly lessons passed is the number of i-Ready Personalized Instruction lessons passed. These weekly measures provide period-specific outcomes.

Secondary Outcomes

Secondary Outcomes (end points)
i-Ready lesson accuracy, attainment of i-Ready Typical Growth and Stretch Growth targets, weekly goal attainment, goal assignments and adjustments, GO Platform use, PLUS tutoring sessions, PLUS tutoring minutes, and PLUS tutoring attendance.
Secondary Outcomes (explanation)
i-Ready growth-target outcomes include attainment of Typical Growth and Stretch Growth targets across available Diagnostic administrations.

Lesson accuracy is available from i-Ready Personalized Instruction records.

Goal-support measures include whether weekly goals are achieved, the goals assigned to students, teacher or tutor goal adjustments, and available GO Platform interaction data.

For periods in which a classroom is assigned to PLUS, tutoring implementation measures include tutoring sessions, tutoring minutes, and tutoring attendance from PLUS records.

Implementation measures include teacher use of the GO Platform through log data, student access to goal information, and completion of the specified weekly and practice-day routines where recorded.

Experimental Design

Experimental Design
The study is a three-condition, classroom-randomized crossover trial conducted during Fall 2026. Eligible classrooms are assigned to one of six two-period sequences: PLUS-BaU, PLUS-GO, BaU-PLUS, BaU-GO, GO-PLUS, or GO-BaU. Each classroom therefore receives two different conditions and does not receive the same condition in both periods. Period 1 runs from September 14 through October 16, 2026. Period 2 runs from October 19 through December 4, 2026. Randomization is stratified by school and grade. Within each school-by-grade stratum, the numbers of classrooms assigned to PLUS, BaU, and GO in each period are made as even as mathematically possible. Across the full sample, the six crossover sequences are also kept as evenly represented as the total number of classrooms permits. The current planning rosters imply approximately 31 classroom sections across three participating schools: about 9 classrooms at School 1 (TN) and 10 at School 2 (TN), and 12 at School 3 (PA). Current planning rosters contain 669 participating students, including 414 students in TN and 255 students in PA. The final classroom and student counts will be based on finalized rosters. A brief scheduled pause in PLUS tutoring occurs around the crossover due to the university academic calendar, but the study does not include a formal washout period.
Experimental Design Details
Not available
Randomization Method
Random assignment is performed in Python using a prespecified random-number seed. Classrooms are randomized within school-by-grade strata to one of the six crossover sequences, with assignments balanced as evenly as possible across conditions in each period and across sequences overall. If baseline i-Ready Diagnostic scores are available before randomization, scores will be standardized within school-by-grade strata. For each candidate assignment, all students with available baseline Diagnostic scores across participating schools will be pooled, and pairwise standardized mean differences will be calculated between PLUS, GO, and BaU separately for Period 1 and Period 2. The first candidate assignment for which the maximum absolute standardized mean difference between any pair of conditions is no greater than 0.25 SD in both study periods will be accepted. Students or schools without baseline Diagnostic scores available at the time of randomization will not contribute to this balance criterion, and no separate school-specific balance threshold will be imposed. The school-by-grade and sequence-balance constraints will apply to all randomized classrooms regardless of baseline Diagnostic availability. The randomization seed and code will be retained, and classroom assignments will not be changed after randomization.
Randomization Unit
Classroom
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
Approx. 31
Sample size: planned number of observations
Approximately 670 students
Sample size (or number of clusters) by treatment arms
Approximately 31 classrooms across six crossover sequences; approximately 10 to 11 classrooms per condition in each period.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
Carnegie Mellon University Institutional Review Board
IRB Approval Date
2026-08-25
IRB Approval Number
STUDY2025_00000406 (MOD202600000619)
Analysis Plan