Where can I download the EDCOM II CRLA data?

Every file is listed in the downloads section of this page with its rows, columns, size and checksum. The files are served from data.nash.com.ph/edcom-data/v1, starting with the README. They are free to download for personal study; any other use, such as republishing them or using them in research, reports or apps, needs NASH Reviewer’s written permission first (email support@nash.com.ph).

What is a struggling reader in the CRLA?

A learner whose CRLA result is Low Emerging or High Emerging, the two lowest of the assessment’s five reading levels. The others are Developing, Transitioning and At Grade Level. In these data each learner is counted once: Grade 1 by the Mother Tongue result, Grade 2 by Filipino and Grade 3 by English.

Is this an official DepEd or EDCOM II dataset?

No. NASH Reviewer extracted and cleaned the data from the public EDCOM II dashboard and is not affiliated with EDCOM II or DepEd. The raw extract is published unchanged beside the cleaned files, so every change can be checked against the dashboard.

Why do these CRLA results differ from DepEd’s CRLA file?

DepEd’s file covers a selected set of schools in five regions and is dated before the mid-year assessment window, so it is a start-of-year round. The EDCOM II dashboard holds end-of-year results for public schools in all 18 regions. In the schools found in both, the end-of-year shares at grade level are far higher.

Can I use the school estimates to rank schools?

Not on their own. A small school’s share swings by chance, so each estimate comes with a 95% interval, and for many schools the data cannot tell whether the share is above or below the national share. Use the estimates to read a school’s result fairly, compare schools of similar size, and look at several years before drawing conclusions.

How should I cite the EDCOM data?

Llego, M. A. (2026). EDCOM II early-childhood dashboard data, cleaned, version 1. NASH Reviewer. https://nash.com.ph/edcom-data. Name the EDCOM II dashboard as the original source of the data.

EDCOM II CRLA data: reading results and downloads

Name: EDCOM II early-childhood dashboard data, cleaned
Creator: Mark Anthony Llego
Published: 2026-10-27
License: https://nash.com.ph/terms#website-content-and-data

Grade 1–3 learners with an end-of-year CRLA result in SY 2025-26, in 38,679 public schools: 5,127,670
Share who are struggling readers (Low or High Emerging): 511,753 learners: 9.98%
Gap between the regions with the highest and lowest shares: BARMM (23.4%) and NCR (5.3%): 4.4×
Share of the variation between schools that lies between schools in the same division: 65%

What the EDCOM II dashboard data cover

The dashboard holds 6 tables: CRLA results by school, school locations, a barangay reference, barangays without a child development center, stunting among children aged 0–2, and program records. They were recovered from the dashboard’s public workbook, 135,406 rows in 27 worksheets, and kept unchanged before any cleaning.

The six tables in the EDCOM II dashboardRows as published and after cleaning. Cleaning dropped header rows and duplicate barangay codes; it changed no CRLA count.

Dashboard tables, their rows before and after cleaning, sources and use
Dashboard table	Rows as published	Rows after cleaning	Source	Used for
Barangay matching	42,025	42,010	PSGC barangay reference with match keys	The PSGC barangay reference for the links
Barangays without a CDC	4,604	4,604	DILG CBMS, as of 17 Sep 2025	Day care center gaps by municipality
Programs	100	100	Intervention program records	Literacy, nutrition and childcare programs compared with need
Stunting 2025	1,743	1,742	DOH/NNC OPT Plus 2025, as of 2 Mar 2026	Stunting among children aged 0–2 by municipality
School locations	48,255	48,254	School coordinates and geography	Linking schools to barangays and municipalities
Reading assessments	38,679	38,679	DepEd CRLA, end of SY 2025-26, as of 10 May 2026	CRLA levels by school, grade and language

Source: EDCOM II early childhood care and development (ECCD) dashboard: CRLA results at the end of SY 2025-26, as of 10 May 2026; analysis script 45

Each learner is counted once: Grade 1 by the Mother Tongue result, Grade 2 by Filipino and Grade 3 by English. A struggling reader is a learner at Low Emerging or High Emerging, the two lowest of the CRLA’s five levels. The CRLA covers almost every public-school learner in these grades, so national and regional shares are population values and carry no sampling intervals.

The data follow the 18 regions of the Philippine Standard Geographic Code (PSGC), with Sulu in Region IX and the Negros Island Region (NIR) as its own region. School inputs such as teachers, classrooms and internet come from DepEd’s SY 2023-24 school files, two school years earlier, joined by school ID.

Of the 38,679 schools, 38,678 (99.997%) are linked to a barangay and its municipality. The dashboard’s own names linked most of them; 299 links were added from validated PSGC codes, research and normalized names. One school stays unlinked because its barangay has no PSGC code yet.

Mark Anthony Llego extracted, cleaned and analyzed the data and drew the charts for NASH Reviewer (https://nash.com.ph/). The report was published on 28 October 2026 from the dashboard as of 10 May 2026. The cleaning and every figure can be re-run from the raw extract.

CRLA reading levels by grade and language, SY 2025-26

At the end of SY 2025-26, 47.6% of learners read at grade level and 9.98% are struggling readers: 511,753 children, about one in 10.

Stacked bars of CRLA reading levels at the end of SY 2025-26 for five grade-and-language blocks and all 5,127,670 learners. At Grade Level is 45% in Grade 1 Mother Tongue, 47% in Grade 2 Filipino and 50% in Grade 3 English; Low Emerging is 9% in Grade 1 and 4% in Grade 3 Filipino. Across all learners 48% read at grade level and 10% are struggling. — Reading levels by grade and languageEach bar splits one grade-and-language block across the five CRLA levels. The last row counts each learner once.

Learners at each CRLA level by grade and language, end of SY 2025-26
Grade and language	Learners assessed	Low Emerging	High Emerging	Developing	Transitioning	At Grade Level	Struggling (Low + High Emerging)
Grade 1 Mother Tongue	1,594,608	9.4%	5.0%	13.0%	27.3%	45.4%	14.4%
Grade 2 Mother Tongue	1,634,270	5.8%	2.1%	10.0%	34.1%	47.9%	7.9%
Grade 2 Filipino	1,693,777	5.8%	2.1%	10.1%	35.0%	47.1%	7.9%
Grade 3 Filipino	1,826,546	4.2%	1.5%	6.8%	36.7%	50.8%	5.6%
Grade 3 English	1,839,285	5.7%	2.4%	8.5%	33.4%	50.1%	8.1%
All learners (one result each)	5,127,670	6.9%	3.1%	10.4%	32.0%	47.6%	10.0%

Grade 1 Mother Tongue has the largest share of struggling readers, 14.4%, and the smallest share at grade level, 45.4%. By Grade 3, 5.6% struggle in Filipino and 8.1% in English.

Most learners who are not yet at grade level are Developing or Transitioning, the two middle levels: they read, but below the level expected for their grade.

These are end-of-year results. DepEd’s selected-school CRLA file is a start-of-year round, so its shares at grade level are far lower; the section on the start and end of the school year compares the two in the 6,061 schools found in both.

Struggling readers by region

Shares of struggling readers range from NCR (5.3%) to BARMM (23.4%), a gap of 4.4 to one. Of the 18 regions, 9 are above the national share of 10.0%.

Share of struggling readers by regionRegions above the national share of 10.0% are highlighted.

BARMM23.4%

Region XII17.7%

Region IX14.7%

Region V12.1%

Region XI11.7%

MIMAROPA11.4%

NIR10.4%

Region VII10.4%

Caraga10.1%

Region I9.5%

Region X9.3%

Region II8.6%

Region VI8.2%

Region III7.9%

Region VIII7.2%

Region IV-A6.7%

CAR5.4%

NCR5.3%

Source: EDCOM II early childhood care and development (ECCD) dashboard: CRLA results at the end of SY 2025-26, as of 10 May 2026; analysis script 33

Show the data as a table

Schools, learners, struggling readers and shares by region, end of SY 2025-26
Region	Schools	Learners assessed	Struggling readers	Share struggling	Share of the country's struggling readers	At Grade Level
NCR	520	469,125	24,884	5.3%	4.9%	55.5%
CAR	1,500	75,844	4,112	5.4%	0.8%	55.8%
Region IV-A	2,754	724,571	48,777	6.7%	9.5%	50.0%
Region VIII	3,656	245,652	17,765	7.2%	3.5%	55.6%
Region III	3,024	520,994	41,175	7.9%	8.0%	54.7%
Region VI	2,433	199,208	16,243	8.2%	3.2%	53.4%
Region II	2,212	164,618	14,237	8.6%	2.8%	48.9%
Region X	2,176	273,425	25,563	9.3%	5.0%	49.3%
Region I	2,409	218,318	20,669	9.5%	4.0%	46.5%
Caraga	1,706	159,091	16,044	10.1%	3.1%	48.3%
NIR	1,742	254,077	26,444	10.4%	5.2%	44.4%
Region VII	2,189	320,467	33,390	10.4%	6.5%	44.6%
MIMAROPA	1,924	186,639	21,230	11.4%	4.1%	44.4%
Region XI	1,749	274,481	32,147	11.7%	6.3%	42.1%
Region V	3,162	346,762	41,812	12.1%	8.2%	41.0%
Region IX	2,310	250,036	36,716	14.7%	7.2%	43.3%
Region XII	1,744	236,985	41,972	17.7%	8.2%	37.2%
BARMM	1,469	207,377	48,573	23.4%	9.5%	30.7%

BARMM has the highest share, followed by Region XII and Region IX. NCR and CAR have the lowest, about half the national share.

By count, the three regions with the most struggling readers (Region IV-A, BARMM, Region XII) hold 27.2% of the country’s total. Region IV-A is among them because it is the largest region, although its share is among the lowest.

A region’s share averages very different divisions and schools, so a regional figure says little about any one school. The regional reports on the blog break each region down by division, language and municipality.

Division results with 95% intervals

Division shares of struggling readers range from 0.1% in City of San Juan (NCR) to 46.3% in Tawi-Tawi (BARMM).

Funnel plot of the share of learners struggling in each of the 232 divisions against the number of learners assessed, with 95% and 99.8% limits around the national 10.0%. 1 divisions lie above the 99.8% limit and 2 below it; the limits stay wide even for the largest divisions because divisions differ by more than chance. — Division funnel plotEach dot is a division’s share against the learners it assessed. The lines mark how far a division of that size usually strays from the national share.

Divisions outside the 95% funnel limits, with estimates, rank intervals and flags
Division	Region	Learners	Share struggling	Estimate (95% interval)	Rank of 232 (95% interval)	Funnel	Flagged at FDR 5%
Tawi-Tawi	BARMM	11,664	46.3%	46.2% (45.3%–47.2%)	232 (232–232)	above 99.8% limit	yes
Maguindanao del Norte	BARMM	40,320	34.1%	34.1% (33.7%–34.6%)	231 (230–231)	above 95% limit	no
Sulu	Region IX	22,946	34.1%	34.1% (33.5%–34.7%)	230 (230–231)	above 95% limit	no
Special Geographic Area	BARMM	2,684	29.5%	29.4% (27.6%–31.1%)	229 (229–229)	above 95% limit	no
Lamitan City	BARMM	6,909	27.3%	27.3% (26.2%–28.3%)	228 (227–228)	above 95% limit	no
Cotabato City	BARMM	13,644	26.6%	26.5% (25.8%–27.3%)	227 (227–228)	above 95% limit	no
Sarangani	Region XII	35,143	24.9%	24.9% (24.5%–25.4%)	226 (226–226)	above 95% limit	no
Mt. Province	CAR	6,965	1.8%	1.9% (1.6%–2.2%)	5 (4–7)	below 95% limit	no
Sto. Tomas City	Region IV-A	9,194	1.7%	1.8% (1.5%–2.0%)	4 (4–6)	below 95% limit	no
Caloocan City	NCR	65,928	1.0%	1.0% (1.0%–1.1%)	3 (3–3)	below 95% limit	no
Pasay City	NCR	13,354	0.5%	0.5% (0.4%–0.6%)	2 (2–2)	below 99.8% limit	no
City of San Juan	NCR	2,906	0.1%	0.2% (0.1%–0.4%)	1 (1–1)	below 99.8% limit	yes

Of the 232 divisions, 12 lie outside the funnel’s 95% limits, listed in the table, and 2 remain flagged after the Benjamini–Hochberg adjustment at a 5% false discovery rate. The division estimates file has all 232.

Each division also has a rank with a 95% interval. The median interval spans 14 places, so most divisions cannot be placed in a precise league table.

Division estimates use empirical Bayes: each division’s share is pulled toward its region’s, more for small divisions. Every division assessed several hundred learners or more, so the pull is small.

School, division or region: where the variation sits

Of the variation between schools in the share of struggling readers, 65% lies between schools in the same division, 22% between divisions in a region and 13% between regions.

Stacked bars of how the variation between schools in the share of struggling readers splits across regions, divisions and schools, for all learners and five grade-and-language blocks. For all learners, 13% lies between regions, 22% between divisions in the same region and 65% between schools in the same division; the school share ranges from 69% to 70% across the five blocks. — How the variation between schools splits across levelsFor all learners and for each grade-and-language block, from a three-level Bayesian model.

Share of the variation between schools at each level, by grade and language, with 95% intervals
Grade and language	Schools	Region: share of between-school variation (95% interval)	Division: share of between-school variation (95% interval)	School: share of between-school variation (95% interval)	Between-school share of all variation (latent scale)
All learners (one result each)	38,679	13% (6%–25%)	22% (17%–26%)	65% (57%–71%)	41%
Grade 1 Mother Tongue	38,234	11% (5%–21%)	19% (16%–23%)	70% (62%–76%)	45%
Grade 2 Mother Tongue	36,152	11% (5%–21%)	19% (15%–23%)	70% (62%–75%)	46%
Grade 2 Filipino	37,227	11% (5%–23%)	19% (15%–23%)	70% (61%–75%)	46%
Grade 3 Filipino	37,589	12% (5%–22%)	18% (15%–22%)	70% (61%–76%)	46%
Grade 3 English	37,860	12% (5%–22%)	19% (15%–22%)	69% (61%–75%)	47%

Two schools picked at random from the same division differ in the odds that a learner is struggling by a median factor of 3.19, against 1.95 for two divisions in a region and 1.68 for two regions.

The division and region intervals overlap (17% to 26% and 6% to 25%), so the data cannot firmly rank those two levels; the probability that divisions vary more than regions is 0.90.

A division’s average is therefore a poor guide to its schools. Support has to be aimed at schools, using their estimates and intervals, rather than spread evenly across a division.

The split is the same in every grade and language. It describes where differences sit, not what causes them: the school level includes learners’ backgrounds and how the screen was given, as well as teaching.

Where struggling readers are concentrated

Struggling readers cluster in place. Among 1,639 municipalities and cities, 64 form high-high clusters, places with high shares surrounded by places with high shares. They hold 9.6% of the country’s struggling readers but only 3.5% of its learners.

Map of local Moran clusters after false discovery rate control: 64 high-high municipalities (high shares of struggling readers next to high), 151 low-low, 2 high-low and 3 low-high; the rest are not significant. High-high clusters are concentrated in BARMM, Region XII, Region IX. — Clusters of high and low shares by municipalityLocal Moran’s I with 9,999 permutations and a 5% false discovery rate; gray places are not part of a cluster.

Municipalities and cities in each cluster type, by region
Region	Municipalities and cities	High-high	Low-low	High-low	Low-high	Gi* hot spots
BARMM	107	38	0	0	2	40
Region XII	49	15	0	0	0	15
Region IX	89	10	0	0	1	10
Region XI	49	1	0	0	0	1
CAR	77	0	26	0	0	0
MIMAROPA	73	0	4	0	0	0
Region I	125	0	5	0	0	0
Region II	93	0	3	0	0	0
NIR	63	0	0	0	0	0
NCR	17	0	1	0	0	0
Region IV-A	142	0	20	0	0	0
Region III	130	0	1	1	0	0
Region V	114	0	0	0	0	0
Region VI	101	0	9	1	0	0
Region VIII	143	0	40	0	0	0
Region VII	101	0	41	0	0	0
Region X	93	0	1	0	0	0
Caraga	73	0	0	0	0	0

The high-high clusters are in BARMM (38), Region XII (15), Region IX (10), Region XI (1). Their pooled share of struggling readers is 27.3%, against 4.3% in the 151 low-low clusters.

Global Moran’s I is 0.65: neighboring municipalities have similar shares far more often than chance would produce.

Need is concentrated among schools too. The tenth of schools with the most struggling readers hold 52.7% of them, with 33.7% of the learners assessed, and 3,475 schools (9.0%) hold half of all struggling readers.

These results describe places, not children. A high-high municipality also has many learners who read well, and a low-low one has learners who struggle.

School inputs, school heads and reading

Within the same division, schools with internet, electricity and fewer learners per teacher have fewer struggling readers. The inputs come from SY 2023-24 school files, two years before the assessment, and the results are associations, not measured effects.

Forest plot of odds ratios for struggling readers for 15 school inputs, within divisions, with 95% intervals, for all 37,578 schools with complete data and for schools with at least 30 learners. 6 inputs go with fewer struggling readers and 3 with more; the largest association is for annex school (odds ratio 1.25). — School inputs and the odds of strugglingOdds ratios with 95% intervals from a Bayesian model with division effects. Below 1 means lower odds of struggling.

Odds ratios for each school input, with a check without small schools and a frequentist check
School input	Odds ratio, Bayesian model (95% interval)	Excluding schools under 30 learners	Frequentist check, division fixed effects	VIF
Learners per teacher (log, per SD)	1.14 (1.12–1.16)	1.13 (1.10–1.15)	1.20 (1.15–1.25)	2.8
Learners per classroom (log, per SD)	1.05 (1.04–1.07)	1.04 (1.03–1.06)	1.05 (1.03–1.07)	2
Share of buildings needing major repair or worse (per SD)	0.99 (0.98–1.00)	0.99 (0.98–1.01)	0.98 (0.97–1.00)	1
School size, elementary enrollment (log, per SD)	0.90 (0.88–0.92)	0.88 (0.87–0.90)	0.78 (0.75–0.81)	3.5
Has electricity	0.84 (0.78–0.90)	0.87 (0.80–0.94)	0.88 (0.79–0.98)	1.1
Has internet	0.89 (0.86–0.91)	0.89 (0.86–0.91)	0.86 (0.82–0.91)	1.1
Has a water supply	1.05 (0.97–1.13)	1.05 (0.97–1.14)	1.06 (0.95–1.17)	1.1
No functional toilet	0.99 (0.91–1.07)	0.94 (0.86–1.02)	0.95 (0.80–1.12)	1
Offers multigrade classes	1.02 (0.98–1.06)	1.12 (1.06–1.17)	1.02 (0.94–1.11)	1.3
Annex school	1.25 (1.09–1.44)	1.26 (1.08–1.46)	1.15 (0.95–1.39)	1
Urban barangay	1.02 (0.99–1.05)	1.02 (0.99–1.05)	1.02 (0.97–1.06)	1.4
Municipal income class 2nd (vs 1st)	0.95 (0.92–0.99)	0.96 (0.93–1.00)	0.94 (0.85–1.04)	1.1
Income class 3rd (vs 1st)	0.97 (0.93–1.01)	0.99 (0.95–1.03)	0.94 (0.87–1.03)	1.1
Income class 4th (vs 1st)	0.88 (0.84–0.92)	0.87 (0.82–0.92)	0.86 (0.76–0.97)	1.1
Income class 5th (vs 1st)	0.80 (0.72–0.90)	0.88 (0.77–0.99)	0.89 (0.69–1.15)	1

Schools with and without a principalThe share of struggling readers in schools led by a School Principal and in other schools, unadjusted and after weighting for size, location and resources.

Struggling readers by school head, with the difference and its 95% interval
Comparison	Schools	Struggling, with principal	Struggling, without	Difference (points, 95% interval)	Risk ratio
Unadjusted	37,578	9.2%	11.3%	-2.1 (-2.8 to -1.5)	0.81
Overlap weights, division indicators (main)	37,578	9.7%	10.9%	-1.2 (-1.6 to -0.8)	0.89
Region instead of division indicators	37,578	9.6%	11.2%	-1.6 (-2.0 to -1.0)	0.86
Schools with 30 or more learners	31,655	9.8%	11.1%	-1.3 (-1.6 to -0.8)	0.89
Principal vs no head position only	26,034	10.2%	12.2%	-2.1 (-2.5 to -1.5)	0.83
Learner-weighted	37,578	10.3%	11.9%	-1.7 (-2.3 to -1.1)	0.86

Source: EDCOM II early childhood care and development (ECCD) dashboard: CRLA results at the end of SY 2025-26, as of 10 May 2026 joined with DepEd school data, SY 2023-24: personnel, facilities, ICT and WASH (BEIS and NSBI); analysis script 40

Internet goes with 0.89 times the odds of struggling (95% interval 0.86 to 0.91) and electricity with 0.84 (0.78 to 0.90). One standard deviation more learners per teacher goes with 1.14 times the odds (1.12 to 1.16), and larger schools have lower odds (0.90, 0.88 to 0.92).

Annex schools have higher odds of struggling readers than other schools in the same division: 1.25 (1.09 to 1.44).

Of the 37,578 schools with input data, 41.1% are led by a School Principal. Their share of struggling readers is 9.2%, against 11.3% elsewhere. Compared with similar schools, the difference is -1.2 percentage points (95% interval -1.6 to -0.8).

An unmeasured difference between the two groups would have to go with both having a principal and struggling readers by a risk ratio of 1.50 to explain that difference away (the E-value).

Stunting, day care centers and programs

In 2025, 4.3% of the 3,007,695 children aged 0–2 measured in Operation Timbang Plus were stunted: 129,690 children. Of the country’s barangays, 10.9% have no child development center (CDC).

Bar chart of the pooled share of learners struggling by the share of a municipality's barangays with no child development center: 10.4% where every barangay has one (684 places) and 10.1% where more than half have none (27 places). — Struggling readers by day care center gapMunicipalities and cities grouped by the share of their barangays with no child development center.

Places, learners and the share of struggling readers by the share of barangays without a CDC
Barangays without a CDC	Municipalities and cities	Learners assessed	Share struggling
None	684	1,706,534	10.4%
Up to 10%	529	1,830,691	9.6%
10–25%	279	928,474	9.3%
25–50%	120	518,559	11.1%
Over 50%	27	143,397	10.1%

Stunting varies widely between places: a tenth of municipalities have shares above 14.5%. In 684 municipalities and cities (41.7%), every barangay has a CDC.

Across municipalities, stunting and the share of struggling readers are barely correlated: Spearman’s rho is 0.05 (95% interval -0.06 to 0.17), and 0.01 (-0.10 to 0.12) for the share of barangays without a CDC. The share of struggling readers moves little across the CDC bands.

Stunting describes children aged 0–2 in 2025 and the CRLA describes Grades 1–3 in SY 2025-26: different children. Comparisons between them are about places, not individual children.

The dashboard also lists 100 literacy, nutrition and childcare program records from 24 organizations, reporting 12,210 beneficiaries. The three regions with the most struggling readers (Region IV-A, BARMM, Region XII) hold 27.2% of the need but 5.0% of the recorded programs. The list is not a census of programs.

Start and end of the school year in the same schools

DepEd’s own CRLA file is dated 28 September 2025, before the mid-year assessment window. In the 6,061 schools found in both files, the share of struggling readers fell from 40.5% at the start of the year to 8.6% at the end, and the share at grade level rose from 15.6% to 51.3%.

Stacked bars of CRLA reading levels for five grade-and-language blocks at the start and end of SY 2025-26 in the same 6,061 schools. At grade level rises from 5% to 49% in Grade 1 Mother Tongue, from 17% to 51% in Grade 2 Filipino and from 23% to 53% in Grade 3 English, while the Low Emerging share in Grade 1 falls from 55% to 8%. — Reading levels at the start and end of SY 2025-26The same schools in DepEd’s start-of-year file and the EDCOM II end-of-year results.

Learners at each CRLA level at the start and end of SY 2025-26 in the same schools, by grade and language
Grade and language	Round	Learners	Low Emerging	High Emerging	Developing	Transitioning	At Grade Level	Struggling (Low + High Emerging)
Grade 1 Mother Tongue	Start of year	197,476	55.1%	17.9%	10.1%	11.7%	5.3%	72.9%
Grade 1 Mother Tongue	End of year	191,363	7.9%	4.7%	11.7%	26.5%	49.2%	12.6%
Grade 2 Mother Tongue	Start of year	218,841	26.8%	5.5%	15.0%	35.4%	17.2%	32.4%
Grade 2 Mother Tongue	End of year	196,223	4.6%	2.1%	9.4%	32.3%	51.6%	6.7%
Grade 2 Filipino	Start of year	217,418	25.5%	5.2%	15.2%	36.9%	17.1%	30.8%
Grade 2 Filipino	End of year	196,905	4.4%	2.1%	9.2%	33.2%	51.2%	6.5%
Grade 3 Filipino	Start of year	229,468	15.6%	3.2%	10.3%	48.2%	22.7%	18.8%
Grade 3 Filipino	End of year	214,961	3.3%	1.4%	6.2%	34.3%	54.7%	4.8%
Grade 3 English	Start of year	224,986	13.4%	7.9%	13.4%	42.0%	23.2%	21.3%
Grade 3 English	End of year	219,074	4.6%	2.4%	7.9%	31.9%	53.3%	7.0%
All learners (G1 MT + G2 Fil + G3 Eng)	Start of year	639,880	30.4%	10.1%	13.0%	30.9%	15.6%	40.5%
All learners (G1 MT + G2 Fil + G3 Eng)	End of year	607,342	5.6%	3.0%	9.5%	30.6%	51.3%	8.6%

The comparison was set up before the results were seen. DepEd’s file counts as an earlier round only if its counts are within ±15% of EDCOM II’s for most schools, its shares are lower in every block, and its date fits the start of the year. All three hold; the counts agree within ±15% in 89.5% of schools.

The median school’s share at grade level rose by +36 percentage points, and 96.7% of schools improved.

The two files are two rounds of a screen given by teachers, not a controlled test. The gain combines a year of teaching with familiarity with the format and differences in how the rounds were given.

How the dashboard data were cleaned

Every change is in a cleaning log of 1,486 rows, each with the record, the rule, the old and new values and the reason. No CRLA count was changed.

Cleaning rules and the records each changed

Cleaning rules with the number of records each changed
Change	Records
Sulu moved to Region IX	547
Garbled text repaired	226
Barangay link confirmed by research	149
Coordinates from research	114
Barangay link by validated PSGC code	110
Location from a predecessor school	98
Region relabeled from the PSGC code (NIR and Sulu)	86
Region filled from the province	68
Region filled from the municipality	30
Barangay link by normalized name	26
Duplicate barangay code dropped	15
Two-way name match resolved by code	14
Header row dropped	2
Region filled from the barangay	1

Source: EDCOM II early childhood care and development (ECCD) dashboard: CRLA results at the end of SY 2025-26, as of 10 May 2026; analysis script 45

How schools were linked to their barangay

Schools by the method that linked them to a barangay
How the school was linked to its barangay	Schools	Share
Dashboard's exact name match	38,379	99.22%
Confirmed by research	149	0.39%
Validated PSGC code	110	0.28%
Normalized name	26	0.07%
Name match resolved by code	14	0.04%
Not linked	1	0.00%

Source: EDCOM II early childhood care and development (ECCD) dashboard: CRLA results at the end of SY 2025-26, as of 10 May 2026; PSA Philippine Standard Geographic Code, and PSA and NAMRIA municipal boundaries via OCHA HDX (CC BY-IGO); analysis script 45

Garbled text from a character-encoding error, such as ñ shown as two symbols, was repaired in 226 records. Region labels were filled or corrected from the PSGC codes, and Sulu schools were moved to Region IX.

Coordinates for 114 schools come from research, not DepEd. They are less precise: a median of 5.2 km from the barangay center, against 0.73 km for DepEd’s coordinates, and 26 lie more than 10 km away.

A school with no location of its own took its predecessor school’s location in 98 cases; 2 proposed predecessors were rejected and 1 awaits confirmation.

Among the 38,679 assessed schools, 769 have coordinates flagged as suspect, 368 lie far from their barangay and 324 have none. The flags are in the school locations file.

What the data mean for a school head

Six points from the data for a school head, or for someone preparing to become one.

Read your school’s estimate and its interval
A small school’s share swings by chance. In schools with fewer than 30 learners, the 95% interval of the estimated share is typically 17 percentage points wide; in the largest schools it is about 4. The school estimates file gives every school’s estimate and interval.
Compare your school within its division
Most of the variation between schools (65%) lies inside divisions, so a division’s average says little about any one school. Compare with similar schools in your division and with your own earlier results.
Start with Grade 1
Grade 1 Mother Tongue has the most struggling readers, 14.4% of learners. Those still at Low or High Emerging at the end of Grade 1 need support before Grade 2 begins.
Know where your municipality stands
Struggling readers cluster in place. If your school sits in a high-high cluster, its results reflect the area as well as the school, so plan with the municipality, the barangay and neighboring schools.
Treat school inputs as clues, not causes
Internet, electricity and fewer learners per teacher go with fewer struggling readers within a division. They are reasons to look closer at your own school, not proof that one purchase will change its results.
Check your data before you report it
22.1% of schools report no struggling reader, more than chance predicts, and many report identical Grade 2 Mother Tongue and Filipino results. Check that your school’s CRLA entries are complete and consistent before they are used for planning.

How the figures were computed

The cleaning and the analysis can be re-run from the raw extract. These are the choices that determine what the charts show.

The dashboard’s workbook was extracted table by table. The raw extract is kept unchanged and published, and 11 validation checks (row counts, totals and reconciliation with the dashboard’s national and regional figures) pass on the cleaned tables.
Each learner is counted once (Grade 1 Mother Tongue, Grade 2 Filipino, Grade 3 English), and struggling readers are Low Emerging plus High Emerging. National and regional shares are population values.
Regions follow the PSGC, with Sulu in Region IX and the Negros Island Region as its own region. Schools are linked to barangays by the dashboard’s names, then validated PSGC codes, research and normalized names, and every added link is logged.
School and division estimates use a beta-binomial empirical Bayes model with a concentration fitted for each region, and each comes with a 95% interval. Funnel plots mark the 95% and 99.8% limits, and the Benjamini–Hochberg procedure holds the false discovery rate at 5%.
The variance split, the language model and the school-input model are Bayesian multilevel models fitted in PyMC with the No-U-Turn Sampler. Every fit passed the convergence checks: R-hat at most 1.01 and an effective sample size of at least 400.
Clusters use queen contiguity between municipal boundaries, local Moran’s I and Getis–Ord Gi* with 9,999 permutations, and a 5% false discovery rate.
The principal comparison weights schools by overlap propensity scores built from size, location and resources within divisions, and is repeated under five other specifications.
Every number on this page and in the blog series comes from the same pipeline outputs, so the two agree. All results are associations.

Limits of the data

What these files cannot say, and the patterns in them worth checking before use.

The data are one snapshot of the dashboard, as of 10 May 2026, for one school year. The dashboard may be updated; the files of this release will not change.
The CRLA is a screen given by teachers, not a test under identical conditions. Differences in how and when it was given are part of every school’s result.
Only public schools are covered.
School inputs are from SY 2023-24, two school years before the assessment, and some schools have no input record.
Stunting (ages 0–2, 2025) and reading (Grades 1–3, SY 2025-26) describe different children, so comparisons between them are about places.
The program records are a list compiled for the dashboard, not a census of programs.
Associations between school inputs or school heads and reading do not show what causes what.

About these numbers

8,543 schools (22.1%) report no struggling reader, against about 7,490 expected by chance under a model that lets schools differ, and 29 report only struggling readers, against about 4 expected. These are data checks, not findings about the schools.
In 14,765 schools (49.8% of those with both Grade 2 results), the Mother Tongue and Filipino results are identical level by level; 64.9% of them teach in Tagalog, where the two languages are the same.
The dashboard’s region labels for NIR and Sulu schools differed from their PSGC codes; they were relabeled from the codes, as the cleaning rules table shows.

Download the EDCOM II CRLA data

The cleaned dashboard tables, the analysis tables and the raw extract, free to download with a README and checksums. Reusing them needs our written permission: email support@nash.com.ph.

Cleaned dashboard tables

The dashboard's tables after cleaning, with the log of every change and the validation summary.

Schools and CRLA results38,679 rows · 77 columns · 4.2 MBOne row per school: CRLA results by grade and language, barangay link, coordinates, flagsschools_literacy_clean.csv.gz · CSV (gzip)
School locations48,254 rows · 54 columns · 3.4 MBAll school location records with geography and coordinate qualityschool_locations_clean.csv.gz · CSV (gzip)
Barangay reference42,010 rows · 20 columns · 1.6 MBPSGC barangay reference with match keys and centroidsbarangays_clean.csv.gz · CSV (gzip)
Barangays without a child development center4,604 rows · 8 columns · 35 KBBarangays with no child development center (DILG CBMS, 17 Sep 2025)barangays_without_cdc_clean.csv.gz · CSV (gzip)
Stunting, ages 0–2, 20251,742 rows · 42 columns · 126 KBOPT Plus 2025 stunting, ages 0-2, by region, province and municipality (DOH/NNC, 2 Mar 2026)stunting_2025_clean.csv.gz · CSV (gzip)
Literacy, nutrition and childcare programs100 rows · 22 columns · 6 KBLiteracy, nutrition and childcare program records (100)barangay_programs_clean.csv.gz · CSV (gzip)
Regional overview19 rows · 13 columns · 1 KBThe dashboard's regional overviewoverview_by_region_clean.csv.gz · CSV (gzip)
Region crosswalk18 rows · 4 columns · 1 KBRegion labels to PSGC regionsregion_crosswalk.csv.gz · CSV (gzip)
Regional reconciliation19 rows · 19 columns · 1 KBRegional totals reconciled with the national totalsregion_reconciliation.csv.gz · CSV (gzip)
Cleaning log1,486 rows · 7 columns · 216 KBEvery change the cleaning made: table, record, rule, old and new value, reasoncleaning_log.csv · CSV
Validation summary2 KBRow counts, reconciliation checks and link summaryvalidation.json · JSON

Analysis tables

The tables behind the series: one row per school with its CRLA results and SY 2023-24 school inputs, and the school, division, municipal and regional estimates.

School analysis table38,679 rows · 262 columns · 8.4 MBAnalysis table: one row per school, CRLA blocks and totals, barangay and municipality codes, SY 2023-24 school inputs (in_ columns), head typeedcom_crla_2026.parquet · Parquet
School estimates38,679 rows · 16 columns · 2.9 MBEvery school: share struggling, empirical Bayes and hierarchical Bayes estimates with 95% intervals, funnel flag (read the guide before ranking schools)school_estimates.csv.gz · CSV (gzip)
Division estimates232 rows · 16 columns · 46 KB232 divisions: share struggling, empirical Bayes estimate and 95% interval, rank interval, funnel and FDR flagsdivision_estimates.csv · CSV
School inputs by region230 rows · 7 columns · 27 KBSchool inputs and struggling readers: odds ratios with 95% intervals from a separate model for each regioninputs_by_region.csv · CSV
Municipalities and cities1,639 rows · 23 columns · 437 KBMunicipality or city: learners, struggling readers, estimate, LISA and Gi* results, stunting and CDC gapsmunicipalities.csv · CSV

Raw extract

The six tables as recovered from the dashboard's workbook, unchanged, with the field dictionary and metadata.

Raw dashboard extract8.9 MBThe six tables as recovered from the dashboard workbook, unchanged, with the field dictionary and metadataedcom_raw_extract_csv.zip · ZIP

Version 1, released October 27, 2026: 19 files, 30.2 MB in total. The folder name carries the version number, so the file at a given URL never changes. The blog post on the cleaning documents every table and rule.

Documentation: README.md, LICENSE.txt, CHECKSUMS.sha256. Every table and cleaning rule: How the EDCOM Dashboard Data Was Cleaned.

Common questions about the EDCOM data

Short answers about the files, the definitions and fair use of the results.

The EDCOM CRLA series on the blog

The analysis continues on the blog in 14 articles and 18 regional reports. The articles read each question in depth, and each regional report covers one region’s divisions, languages, municipalities and school resources.

How to cite this page

Use APA style (7th edition). The date is the day the page was last updated. Where you use a figure or a rule, also cite the law, issuance, or report the page names for it.

Llego, M. A. (2026, October 30). EDCOM II CRLA data: reading results for SY 2025-26. NASH Reviewer. https://nash.com.ph/edcom-data

In-text citation: (Llego, 2026). To cite the cleaned data files rather than this page, use the data citation under Common questions about the EDCOM data.

To cite DepEd orders and memoranda themselves, see How to Cite DepEd Orders in APA 7 and Write the Abstract.

Sources

The EDCOM II dashboard compiles data from DepEd, DOH and the National Nutrition Council, and DILG. Each table was checked against the dashboard before any analysis.

EDCOM II early childhood care and development (ECCD) dashboard: CRLA results at the end of SY 2025-26, as of 10 May 2026CRLA reading levels by school, grade and language for public schools, with school locations, a barangay reference and the dashboard’s regional overview.
DOH and National Nutrition Council, Operation Timbang Plus 2025, as of 2 March 2026 (compiled in the EDCOM II dashboard)Children aged 0–2 weighed and measured, and those stunted, by region, province and municipality.
DILG Community-Based Monitoring System: barangays without a child development center, as of 17 September 2025 (compiled in the EDCOM II dashboard)Barangays that report no child development center.
DepEd school data, SY 2023-24: personnel, facilities, ICT and WASH (BEIS and NSBI)Teachers, classrooms, buildings, electricity, water, toilets, internet and school head positions, joined to the CRLA results by school ID.
PSA Philippine Standard Geographic Code, and PSA and NAMRIA municipal boundaries via OCHA HDX (CC BY-IGO)Barangay, municipal and regional codes for the links, and municipal boundaries for the maps and cluster tests.
Analysis pipelineNumbered Python scripts (32–48 in the deped_data repository) that clean the extract, fit the models, draw the charts, build the release and write this page’s tables.

The data were published by EDCOM II’s ECCD dashboard and the Department of Education. NASH Reviewer is not affiliated with EDCOM II, the Department of Education or the agencies whose data the dashboard compiles.

For aspiring school heads

Lead with the reading data in hand

The NASH asks how you would run a school: where to aim reading support, how to read a result fairly and what to do with too few teachers. Practice the PPSSH strands with situational questions that ask you to decide as a school head would.

Start practicing free Compare plans