Key findings
- In the ELLNA for Grade 3, 30.52 percent of learners (430,457 of 1,410,408 assessed) were at least proficient, per the EDCOM 2 Final Report; the report labels the baseline SY 2023-2024 in its main table.
- In the 2024 NAT for Grade 6, 19.56 percent of learners were considered proficient, per the EDCOM 2 Final Report.
- In the NAT Grade 10 for SY 2022-2023, 0.74 percent of learners were Proficient nationally (Not Proficient 4.02 percent, Low Proficient 65.17 percent, Nearly Proficient 30.06 percent), per Figure 18 of the Final Report. EDCOM 2's press release of 16 January 2026 gives a different Grade 10 figure, 1.36 percent, without stating a school year. The two are shown side by side in this Drop and are not merged.
- In the NAT Grade 12 for SY 2023-2024, 0.40 percent of learners (6,518 of 1,611,809) were Proficient; about 74 percent (Not Proficient plus Low Proficient) did not meet the minimum level. The press release headline rounds this to 0.47 percent; its body and the report say 0.40.
- Among 210,350 Grade 12 learners in GIDA, Last Mile, IP Education, Madrasah, and hardship post schools (SY 2023-2024), 266 (0.13 percent) reached Proficient and none Highly Proficient; 81.59 percent were below the minimum threshold.
- PIDS (Santos et al., 2025), in a discussion paper commissioned by EDCOM 2, found that generally more students reach the proficient level under expert standard-setting cut-offs than under DepEd's 75 percent rule, which suggests the cut score might be too high. The effect is large in NAT Grade 6 English (2.6 percent versus 50.0 percent) but reverses in ELLNA Filipino (27.9 percent versus 14.5 percent). PIDS states the comparison does not imply either method is superior.
- EDCOM 2 states that the NAT is not a longitudinal dataset and the four assessments are not equated, so the figures cannot be read as one cohort's decline across grades.
Important numbers: the key-stage matrix
| Assessment | School year | Proficient or at least proficient (%) | Learners assessed (sample-based) | Where stated |
|---|---|---|---|---|
| ELLNA, Grade 3 | Labelled SY 2023-2024 in the report's main table (SY 2024-2025 in an adjacent heading; "2024" in the press release) | 30.52 | 1,410,408 Key Stage 1 learners; 430,457 at least proficient | Source 1 pp. 152, 447; Source 3 |
| NAT, Grade 6 | 2024 | 19.56 | Not stated in extracted pages | Source 1 p. 152; Source 3 |
| NAT, Grade 10 | SY 2022-2023 | 0.74 | Not stated | Source 1 p. 152 and Figure 18, p. 175 |
| NAT, Grade 10 (press release figure) | Not stated (assessments 2023 to 2025) | 1.36 | Not stated | Source 3, 16 January 2026 |
| NAT, Grade 12 | SY 2023-2024 | 0.40 | 1,611,809 learners; 6,518 proficient | Source 1 pp. 153-154, Figure 14 |
| NAT, Grade 12, disadvantaged school groups combined | SY 2023-2024 | 0.13 | 210,350 learners; 266 proficient | Source 1 p. 154 |
Which rows mean what: the ELLNA Grade 3, NAT Grade 6, and press-release Grade 10 figures are described by their sources as "at least proficient" (Proficient plus Highly Proficient). The NAT Grade 10 SY 2022-2023 and NAT Grade 12 SY 2023-2024 figures are the Proficient band as printed in the report's Figures 18 and 14, which print no Highly Proficient value. Learner counts are the numbers assessed as reported by EDCOM 2; DepEd Order No. 55, s. 2016 provides for these assessments to be administered to sampled schools using stratified random sampling, so the counts are not enrolment.
Read across, not down. Each row is a different group of learners in a different year on a different test. EDCOM 2 says the NAT captures different cohorts at different levels across different school years and is not designed to compare performance over time or across key stages. The rows show how much of each key stage meets its own standard; they do not show one cohort losing proficiency as it ages.
How DepEd defines proficient
DepEd Order No. 55, s. 2016 (Source 6) provides that test scores are reported as percentage scores and that the proficiency level is at least 75 percent; it also provides for the assessments to be administered to sampled schools by stratified random sampling. The five reporting bands used by the Bureau of Education Assessment, as described by PIDS and EDCOM 2, are: 90 to 100 Highly Proficient, 75 to 89 Proficient, 50 to 74 Nearly Proficient, 25 to 49 Low Proficient, and 0 to 24 Not Proficient. "At least proficient" means a score of 75 or more. EDCOM 2 notes that DepEd treats Nearly Proficient as meeting the minimum skills for a key stage, which is why "below the minimum" in this Drop means Low Proficient plus Not Proficient.
| Band | Score range | Counted as |
|---|---|---|
| Highly Proficient | 90 to 100 | At least proficient |
| Proficient | 75 to 89 | At least proficient |
| Nearly Proficient | 50 to 74 | Meets minimum skills (DepEd) |
| Low Proficient | 25 to 49 | Below minimum |
| Not Proficient | 0 to 24 | Below minimum |
Band distributions where EDCOM 2 prints them
The Final Report prints full band percentages for two assessments. Bars are drawn from the same numbers as the table cells.
No Highly Proficient value is printed for these distributions (the report prints four bands for Grades 10 and 12, and states none for the disadvantaged group), so no Highly Proficient segment appears in the bars.
| Assessment | Not | Low | Nearly | Proficient | Highly | Bar |
|---|---|---|---|---|---|---|
| NAT Grade 10, SY 2022-2023, national | 4.02 | 65.17 | 30.06 | 0.74 | not printed | |
| NAT Grade 12, SY 2023-2024, national (1,611,809 learners) | 5.02 | 69.10 | 25.47 | 0.40 | not printed | |
| NAT Grade 12, SY 2023-2024, disadvantaged school groups combined (210,350 learners) | 6.13 | 75.46 | 18.28 | 0.13 | 0.00 |
Grade 12 counts: Not Proficient 80,938; Low Proficient 1,113,757; Nearly Proficient 410,596; Proficient 6,518. For the disadvantaged group, EDCOM 2 prints 12,884 Not Proficient (6.13 percent), 158,739 Low Proficient (75.46 percent), 266 Proficient (0.13 percent), and none Highly Proficient; the Nearly Proficient share of 18.28 percent is the remainder computed by Evidence Atlas and is not printed by EDCOM 2. The Final Report's per-subgroup chart (Figure 15) is not reproduced because its printed rows do not sum to 100 percent; see Limitations.
Cut-off comparison: DepEd's 75 percent rule versus expert standard-setting (2022 tests)
PIDS Discussion Paper 2025-25 (Santos et al., September 2025), commissioned by EDCOM 2, ran standard-setting workshops in which expert panels defined proficiency levels and cut scores for the 2022 ELLNA and 2022 NAT Grade 6 by subject, then compared the share of learners at each level under the DepEd Bureau of Education Assessment (BEA) bands and under the panels' cut-offs. PIDS found that generally more students reach the proficient level under the standard-setting cut-offs, suggesting the 75 percent cut score might be too high, while stating that the comparison does not imply one set of results is superior. The paper is labelled preliminary, unedited, and unreviewed.
Bars show the share of learners at least proficient. BEA = DepEd bands (Proficient plus Highly Proficient). SS = PIDS standard-setting levels at or above the panel's minimum proficiency level. The at-least-proficient sums and the gaps are computed by Evidence Atlas from the level percentages PIDS prints in Appendix 12 to 19. The NAT Grade 6 Filipino and Science subjects are not shown as bars; see the caveat below the table.
| Test and subject | Highly Proficient | Proficient | Nearly Proficient | Low Proficient | Not Proficient | At least proficient |
|---|---|---|---|---|---|---|
| ELLNA English | 5.2 | 21.2 | 31.6 | 33.5 | 8.5 | 26.4 |
| ELLNA Filipino (printed row sums to 95.5) | 5.2 | 22.7 | 33.2 | 27.9 | 6.5 | 27.9 |
| ELLNA Numeracy (Tagalog) | 5.0 | 10.0 | 20.4 | 40.1 | 24.6 | 15.0 |
| NAT Grade 6 English | 0.0001 | 2.6 | 33.8 | 50.2 | 13.4 | 2.6 |
| NAT Grade 6 Mathematics | 1.1 | 7.6 | 21.8 | 45.9 | 23.5 | 8.7 |
| NAT Grade 6 Araling Panlipunan | 0.0001 | 5.0 | 32.2 | 47.4 | 15.4 | 5.0 |
| NAT Grade 6 Filipino | 1.1 | 16.7 | 45.6 | 31.5 | 5.1 | 17.8 |
| NAT Grade 6 Science | 0.4 | 7.1 | 28.9 | 48.0 | 15.6 | 7.5 |
| Test and subject | Level 5 | Level 4 | Level 3 | Level 2 | Level 1 | At least proficient |
|---|---|---|---|---|---|---|
| ELLNA English | 23.6 | 14.0 | 13.2* | 15.2 | 34.0 | 50.8 |
| ELLNA Filipino | n/a | n/a | 14.5* | 23.8 | 61.7 | 14.5 |
| ELLNA Numeracy (Tagalog) | n/a | 17.7 | 12.5* | 6.5 | 63.4 | 30.2 |
| NAT Grade 6 English | n/a | n/a | 7.4 | 42.6* | 50.1 | 50.0 |
| NAT Grade 6 Mathematics | n/a | n/a | 17.6 | 6.5* | 75.8 | 24.1 |
| NAT Grade 6 Araling Panlipunan | n/a | n/a | 10.9* | 24.8 | 64.4 | 10.9 |
| NAT Grade 6 Filipino (see caveat) | n/a | n/a | 26.0 | 16.9* | 57.1 | not computed (see caveat) |
| NAT Grade 6 Science (see caveat) | n/a | n/a | 26.0 | 16.9* | 57.1 | not computed (see caveat) |
"n/a" means the panel defined fewer than five levels for that test, so the level does not exist; it is not a missing value.
Caveat on the Filipino and Science rows. PIDS Appendix 16 (NAT Grade 6 Filipino) and Appendix 18 (NAT Grade 6 Science) print identical standard-setting percentages (26.0, 16.9, 57.1) despite different cut-score ranges. This is likely a transcription error in the discussion paper. Those two SS rows are shown for completeness only, are excluded from the bars above, and should not be read as evidence that the two subjects perform identically.
What standard-setting means, in plain terms
DepEd's rule fixes proficiency at a score of 75 percent on every test, regardless of how hard the items are. Standard-setting instead asks a panel of subject experts to place the test's items on a difficulty continuum, describe what learners at each level can do, and choose the level that represents the minimum a learner at that grade should reach. Because the cut then depends on the actual test content, the share of learners counted as proficient can move up or down. PIDS reports that in most subjects it moved up, and in ELLNA Filipino it moved down. The panels' level descriptions (for example, what a Grade 6 learner at Level 2 in Mathematics can likely do) are in the PIDS paper's Appendix C.
Where the figures disagree, and what this Drop does about it
| Item | Source A | Source B | Treatment here |
|---|---|---|---|
| NAT Grade 10 at least proficient | EDCOM 2 press release, 16 January 2026: 1.36 percent, school year not stated | EDCOM 2 Final Report p. 152 and Figure 18: 0.74 percent, SY 2022-2023 | Both shown with attribution; not merged or averaged. The difference may reflect different school years within the 2023 to 2025 administrations, but the sources do not say. |
| NAT Grade 12 proficient | Press release headline: 0.47 percent | Press release body and Final Report: 0.40 percent (6,518 of 1,611,809) | 0.40 percent used; headline variant noted. |
| ELLNA baseline year | Final Report Outcome 2.1.D table: SY 2023-2024 | Adjacent equity-targets table: SY 2024-2025; press release: "in 2024" | SY 2023-2024 used, with the label conflict noted. |
| Grade 12 subgroup bands (Figure 15) | Printed subgroup rows repeat values from the Grade 10 chart (Figure 18) and sum to 103.8 to 112.2 percent | Page 154 prose gives consistent combined-group counts | Figure 15 rows excluded; combined-group prose figures used. |
| PIDS NAT Grade 6 Filipino and Science SS rows | Appendix 16: 26.0 / 16.9 / 57.1 | Appendix 18: 26.0 / 16.9 / 57.1 with different cut ranges | Shown with caveat; excluded from bars. |
Targets and international context
EDCOM 2's National Education Plan sets ELLNA at-least-proficient targets of 50 percent by 2028, 70 percent by 2031, and 90 percent by 2035 from the 30.52 percent baseline. Subgroup baselines in the same table are GIDA 24.65 percent, Last Mile Schools 29.26 percent, Hardship Post 22.17 percent, IPEd 24.91 percent, and MEP 25.23 percent. These are policy targets, not forecasts.
EDCOM 2's press release also cites international assessments: PISA results in which 76 percent of 15-year-old Filipino students scored below Level 2 in reading; SEA-PLM data in which 14 percent of Grade 5 learners reached the highest reading bands; and a UNICEF and World Bank estimate that 91 percent of children at late primary age cannot read and understand a simple story. The press release does not state the year or cycle of these studies, and the original documents were not opened for this Drop; they are reported here only as EDCOM 2 cites them.
Sources and evidence
- Second Congressional Commission on Education (EDCOM 2), "Turning Point: A Decade of Necessary Reforms", EDCOM II Final Report (Year 3), January 2026 (file dated 26 January 2026; ceremonial submission 29 January 2026); retrieved 3 September 2026. PDF, 634 pages. Pages cited by printed page number: 138 to 142 (assessment system), 152 to 155 (proficiency results, Figures 12 to 15), 175 (Figure 18), 447 (Outcome 2.1.D targets). Data attributed by EDCOM 2 to DepEd Bureau of Education Assessment as of September 2025. Supports CL-01 to CL-07, CL-15.
- Santos, K. C. P., Cagasan, L. P. Jr., Loberiza, K. B., Tenorio, A. D., Angeles, A. J. G., and Lapitan, M. A. B., "Examining the DepEd's National Assessments: A Review of Framework, Design, Development, Psychometric Properties, and Utilization", PIDS Discussion Paper Series No. 2025-25, September 2025; retrieved 3 September 2026. PDF, 89 pages. Pages 56 to 57 (standard-setting findings), 61 to 62 (cut scores), 86 to 88 (Appendix 12 to 19). Labelled preliminary, unedited, and unreviewed. Supports CL-05, CL-08 to CL-14.
- EDCOM 2 Communications, "Student proficiency rates plunge from 30% in Grade 3 to 0.47% in Grade 12", press release, 16 January 2026 (modified 11 February 2026); retrieved 3 September 2026. link. Supports CL-01, CL-02, CL-05, CL-07, CL-16; source of the 1.36 percent Grade 10 figure.
- Manila Bulletin, "Filipino students' proficiency plummets to near zero by Senior High School, study finds", 16 January 2026; retrieved 2 September 2026. link. Secondary reporting.
- PIDS, "PIDS study seeks review of DepEd's 75% proficiency standard", press release, December 2025. link. Seen in search results only; not opened.
- Department of Education, DepEd Order No. 55, s. 2016, "Policy Guidelines on the National Assessment of Student Learning for the K to 12 Basic Education Program", 30 June 2016; retrieved 4 September 2026. PDF, 37 pages. Enclosure pages 6, 7, 9, and 11: stratified random sampling of schools; scores reported as percentage scores; proficiency level at least 75 percent. The order does not itself print the five band ranges, which are as described by Sources 2 and 3. Supports CL-05 and the sampling disclosure.
Claim IDs (CL-…) refer to the Evidence Atlas verified claim register CR-20260903-01 for this Drop.
Methodology
All assessment figures were read from the machine-readable text of the EDCOM 2 Final Report PDF and the PIDS discussion paper PDF, retrieved on 3 September 2026, and cross-checked against the EDCOM 2 press release. DepEd Order No. 55, s. 2016 was retrieved on 4 September 2026 to confirm the 75 percent proficiency threshold, the percentage-score reporting, and the sample-based administration (stratified random sampling of schools); it does not print the five band ranges, which remain as described by PIDS and EDCOM 2. Figures were used only where the source prints them as text or as chart data labels (Final Report Figures 14 and 18; PIDS Appendix 12 to 19). Charts without printed labels (Final Report Figures 12 and 13) were not used. Every claim was independently re-verified in a separate pass against the same extracts before this Drop was built.
Evidence Atlas computations are limited to: (a) "at least proficient" sums under BEA bands (Proficient plus Highly Proficient) and under PIDS standard-setting levels (the minimum proficient level and above), from the level percentages PIDS prints; (b) the Nearly Proficient remainder (18.28 percent) for the disadvantaged Grade 12 group, from EDCOM 2's printed counts; and (c) the checks that printed rows sum to 100 percent, which surfaced the Figure 15 and Appendix 16/18 anomalies. No trend across grades is computed, because the sources say the tests are not equated and the cohorts differ.
Source hierarchy: the EDCOM 2 Final Report and the PIDS paper (which is original research commissioned by EDCOM 2) first; the EDCOM 2 press release as a second primary statement from the same commission; news reporting only to confirm what has been publicly quoted.
Limitations
- The ELLNA and NAT are administered to sampled schools using stratified random sampling under DepEd Order No. 55, s. 2016, with all learners in a sampled school tested. The learner counts shown (1,410,408 and 1,611,809) are the numbers assessed as reported by EDCOM 2, not enrolment, and the rates describe the sampled population as DepEd reports it.
- The four key-stage figures come from different cohorts, school years, and non-equated instruments; EDCOM 2 states the NAT is not longitudinal and is not designed for comparison over time or across key stages. Nothing here shows one cohort's decline.
- Grade 10: the Final Report's 0.74 percent (SY 2022-2023) and the press release's 1.36 percent (year not stated) cannot be reconciled from the sources; both are shown, neither is presented as the settled rate.
- Grade 12: the press release headline says 0.47 percent; its body and the Final Report say 0.40 percent. The 0.40 percent figure is used.
- The ELLNA baseline is labelled SY 2023-2024 in one Final Report table and SY 2024-2025 in an adjacent heading; the press release calls it the 2024 ELLNA.
- The Final Report's Figure 15 (Grade 12 by disadvantaged subgroup) prints rows that repeat Grade 10 values and sum to between 103.8 and 112.2 percent; those rows are excluded, and only the combined-group figures stated in the report's prose are used.
- Cohort sizes for the 2024 NAT Grade 6 and the SY 2022-2023 NAT Grade 10 were not found in the pages extracted.
- DepEd Order No. 55, s. 2016 states the 75 percent proficiency threshold but does not print the five band ranges (90 to 100, 75 to 89, and so on); those ranges are as described by PIDS and EDCOM 2.
- The PIDS standard-setting results apply to the 2022 tests only and come from a discussion paper that PIDS labels preliminary, unedited, and unreviewed. PIDS states its comparison does not imply either method is superior; "generally" more learners are proficient under standard-setting, and ELLNA Filipino is a counter-example. The printed BEA row for ELLNA Filipino sums to 95.5 percent.
- PIDS Appendix 16 and 18 print identical standard-setting rows for NAT Grade 6 Filipino and Science; these are shown with a caveat and excluded from the comparison bars.
- The international figures (PISA, SEA-PLM, UNICEF and World Bank) are reported as EDCOM 2 cites them, without years, and the original studies were not opened.
- Regional breakdowns were not extracted and do not appear here.
- The press release page was modified on 11 February 2026; what changed is not recorded by EDCOM 2.
Freshness
Freshness class F4 (research). Last verified 3 September 2026. Monitored sources: the EDCOM 2 website for corrections to the Final Report or press release and for any published reconciliation of the Grade 10 figures; PIDS for a revised or peer-reviewed version of Discussion Paper 2025-25; DepEd's Project BUKAS for direct Bureau of Education Assessment tables, which would replace figures currently taken via EDCOM 2. Next expected release: DepEd NAT and ELLNA results for SY 2024-2025 and SY 2025-2026, dates not announced. "Updated" for this Drop means a change to a published number, definition, or conclusion, not a routine source check.
Version
v1.1, 4 September 2026. Changelog: v1 initial build (3 September 2026). v1.1 added the sample-based administration disclosure and DepEd Order No. 55, s. 2016 as a retrieved source; labelled the ELLNA school year as a disputed label; removed the derived at-least-proficient figure for the anomalous NAT Grade 6 Filipino and Science rows; clarified which matrix rows are "at least proficient" and which are the Proficient band; corrected the band-distribution caption; replaced the hide-one-series toggle with a sort-by-gap control; split the PIDS comparison into two fully columned tables; raised the provenance notes to body size; shortened the title and split the abstract; removed the unused Highly Proficient legend swatch. No published figure changed. Drop ID D-20260903-3e8f. Evidence package EP-20260903-01; verified claim register CR-20260903-01.