TALLY Claims: 10 - SUPPORTED 1 - STRETCHED 7 - NOT IN RESULTS 1 - CONTRADICTED 1 Most serious: S4 - families would believe their children were tutored every school day, when tutoring schools delivered 3.4 sessions a week on average against 5 planned. 1. THE STUDY, FROM ITS RESULTS Design - Cluster-randomised trial: 24 elementary schools randomly assigned, 12 to daily 30-minute small-group tutoring for grades 2 to 4 and 12 to usual instruction, stratified by prior school reading average. Who - 3,108 students analysed (1,528 tutoring, 1,580 comparison) of 3,412 enrolled, in grades 2 to 4 at 24 elementary schools (Brenmoor, going by the assessment's name), over the 2025-26 school year. Primary outcome - End-of-year Brenmoor Standardized Reading Assessment. Primary result - Adjusted mean difference 0.06 SD in favour of tutoring (95% CI -0.04 to 0.16), p = 0.24; the pre-specified primary hypothesis was not supported. Stated limits - "could only reliably detect effects of about 0.15 SD or larger"; "Teachers rating engagement knew whether their school had tutoring"; "Missing end-of-year scores were more common in tutoring schools"; "one year of partial implementation". 2. CLAIM LEDGER S1 - MEANING - "A Year of Growing Readers" - "Primary outcome, end-of-year Brenmoor Standardized Reading Assessment: adjusted mean difference 0.06 SD in favor of tutoring (95% CI -0.04 to 0.16), p = 0.24. The pre-specified primary hypothesis was not supported." - STRETCHED - NULL - The headline promises reading growth, but the main reading measure's interval runs from -0.04 to 0.16, including no difference. S2 - METHOD - "Last fall, Brenmoor County Schools launched Read Together Daily, giving students in grades 2 through 4 thirty minutes of small-group reading" - "24 elementary schools were randomly assigned, 12 to daily 30-minute small-group tutoring for grades 2 to 4 and 12 to usual instruction, for the 2025-26 school year" - SUPPORTED - Grades, 30-minute sessions, small groups and the school year all match the design. S3 - METHOD - "with a trained tutor" - nothing in the results - NOT IN RESULTS - The results never say who the tutors were or how they were trained. S4 - METHOD - "every school day" - "Implementation: tutoring schools delivered a mean of 3.4 sessions per week against 5 planned." - CONTRADICTED - Every school day is five sessions a week; tutoring schools delivered 3.4 on average. S5 - FINDING - "Second graders in Read Together schools are reading faster and more smoothly" - "Secondary outcomes, not adjusted for multiple comparisons: ... Grade 2 oral reading fluency: +4.1 words correct per minute (95% CI 0.6 to 7.6), p = 0.02." - STRETCHED - SWAP - A real gain, but on a secondary outcome not adjusted for multiple comparisons, presented as the result while the primary goes unreported. S6 - FINDING - "teachers tell us students are more engaged and excited about books than ever" - "Teacher-reported engagement, 5-point scale: difference 0.31 (95% CI 0.12 to 0.50), p = 0.002." - STRETCHED - SWAP - A secondary teacher rating, from teachers who knew their school had tutoring, becomes excitement "than ever", a comparison with past years nobody measured. S7 - FINDING - "attendance went up in our tutoring schools too" - "Attendance: 94.1% vs 93.6%, difference 0.5 percentage points (95% CI -0.3 to 1.3), p = 0.21." - STRETCHED - NULL - 94.1% vs 93.6% is 0.5 points, with an interval from -0.3 to 1.3 and p = 0.21: no established difference. S8 - MEANING - "This is what happens when we invest in our kids" - "Primary outcome, end-of-year..." - STRETCHED - NULL - Credits the program with the gains listed, but its primary reading difference and the attendance difference were not established. S9 - MEANING - "Read Together Daily works" - "The pre-specified primary hypothesis..." - STRETCHED - NULL - The results state the primary hypothesis was not supported; "works" writes that as an established effect. S10 - MEANING - "every child in Brenmoor deserves it" - "Primary outcome, end-of-year..." - STRETCHED - NULL - Recommends it for every child, beyond grades 2 to 4, on a primary reading effect the results did not establish. 3. WHAT THE SUMMARY LEAVES OUT The primary outcome is never reported: "adjusted mean difference 0.06 SD in favor of tutoring (95% CI -0.04 to 0.16), p = 0.24. The pre-specified primary hypothesis was not supported." The engagement gain rests on unblinded raters: "Teachers rating engagement knew whether their school had tutoring." Every gain the summary reports is among the "Secondary outcomes, not adjusted for multiple comparisons". Not everyone was counted: "3,412 students enrolled at baseline" but "3,108 had an end-of-year score and were analyzed", with "Missing scores: 10.2% in tutoring schools, 7.7% in comparison schools." The study's reach was limited: "with 24 schools the study could only reliably detect effects of about 0.15 SD or larger." 4. REWRITE Read Together Daily: First-Year Reading Results Mixed Last fall, Brenmoor County Schools launched Read Together Daily in 12 of 24 elementary schools, chosen at random: thirty minutes of small-group tutoring for grades 2 through 4, planned for every school day, while the other 12 kept usual instruction. Tutoring schools averaged 3.4 sessions a week. On the main measure, the end-of-year reading assessment, tutoring schools scored 0.06 standard deviations higher, a difference the study could not tell apart from none. Among other measures, second graders in tutoring schools read aloud about 4 more words correctly per minute than in comparison schools, but grades 3 and 4 showed no clear gain. Teachers, who knew whether their school had tutoring, rated students in tutoring schools 0.3 points more engaged on a 5-point scale. Attendance showed no clear difference: 94.1% vs 93.6%. Next fall, the district will bring Read Together Daily to every elementary school. Thank you to our tutors, teachers and families. Changes: kept S2; narrowed S1, S5, S6, S7; cut S3; corrected S4; flagged for the speaker S8, S9, S10. 5. NOT CHECKED Protocol or registration - the results call the primary hypothesis "pre-specified", but no protocol or registration was pasted. S3 - how tutors were chosen and trained could sit in a protocol or implementation report that was not pasted. Adjustment - what the "adjusted mean difference" was adjusted for is not stated. Blinding - whether those giving or scoring the reading assessment knew each school's assignment is not stated; only the engagement raters are described. Comparison - what "usual instruction" involved in the 12 comparison schools is not stated.