TL;DR: A Table of Specification (TOS) is a two-way grid, topics against cognitive levels, that a teacher fills out before writing a test to make sure item counts actually match how much class time each topic got. After the test, item analysis checks whether individual items worked: the difficulty index (percent who got it right) flags items that are too easy or too hard, and the discrimination index (how much the item separates high scorers from low scorers) flags items that may be mis-keyed or confusing. Under a commonly used convention, items in the 51-84% difficulty range with a discrimination index above 0.30 are solid; anything below a 0.10 discrimination index, or negative, needs a second look before you finalize grades.
Two things come up constantly in Philippine classrooms and rarely get taught step by step: how to actually build a Table of Specification before a summative test, and how to run item analysis after it. Skipping the first means your test can quietly over-test whatever chapter you personally like; skipping the second means a badly worded or mis-keyed item can sit in your item bank for years without anyone catching it. This guide walks through both with a full worked example, using real formulas, not just definitions. It pairs well with our DepEd grading system guide and transmutation table guide if you also need to see how raw scores like these eventually become report card grades.
What Is a Table of Specification?
A Table of Specification, sometimes called a test blueprint, is a two-way chart a teacher completes before writing a single test item. One axis lists the content or learning competencies covered in the grading period; the other lists cognitive levels, most commonly the six levels of the revised Bloom's Taxonomy: remembering, understanding, applying, analyzing, evaluating, and creating. Each cell in the grid records how many items that topic gets at that cognitive level.
The point of building one is to prevent two common test-writing problems: over-representing whatever topic the teacher feels most confident writing questions about, and clustering all items at the "remembering" level because recall questions are the fastest to write. A properly weighted TOS forces item counts to track actual instructional time and to reach into higher-order thinking for grade levels where that's expected.
How Do You Build a Table of Specification? A Worked Example
Start with total teaching days per topic, not a guess. Suppose a Grade 8 Araling Panlipunan quarter covered four topics over 30 teaching days, and the teacher is building a 40-item quarterly test.
| Topic | Teaching Days | % of Total | Items (of 40) |
|---|---|---|---|
| Pre-colonial Philippine society | 6 days | 20% | 8 |
| Spanish colonization | 9 days | 30% | 12 |
| The Philippine Revolution | 10 days | 33% | 13 |
| American colonial period | 5 days | 17% | 7 |
| Total | 30 days | 100% | 40 |
The item count per topic (Teaching Days ÷ Total Days × Total Items) is the mechanical part. The next step is splitting each topic's items across cognitive levels based on what was actually taught. If "Spanish colonization" was mostly taught through recall of dates and reading comprehension, with one class period spent on analyzing primary sources, its 12 items might break down as 4 remembering, 4 understanding, 2 applying, and 2 analyzing, rather than an even split. There's no single DepEd-mandated ratio for this step; it should reflect your actual lesson delivery, which is exactly why the TOS has to be built before the test, not reverse-engineered after.
Disclaimer: DepEd does not publish one single, mandatory TOS template; schools and divisions vary in the exact format required. Confirm the specific format and item-count expectations for your grade level and subject with your school head or Schools Division Office.
What Does DepEd's Current Assessment Policy Say About Item Counts?
DepEd Order No. 015, s. 2026, the "Revised Guidelines on Classroom Assessment, Grading System, and Awards and Recognition for the K to 12 Basic Education Program," takes effect SY 2026-2027 and supersedes the older DO 8, s. 2015. It carries forward the requirement that summative tests be competency-aligned and built to generate valid evidence of learning, and continues to expect higher-order cognitive levels (applying, analyzing, evaluating, creating) to appear more as students move into junior and senior high school, with lower key stages leaning more toward remembering and understanding. Third-party tools built around the order have summarized indicative maximum item counts by key stage (for example, roughly 20-30 items for a single summative test depending on key stage), but treat any specific item-count ceiling you see online as a secondary summary rather than a verbatim quote until you've confirmed it against your division's own implementation guidance, since DepEd has not published the order's full text in an easily quotable public web page as of this writing.
What Is Item Analysis, and Why Run It After Every Test?
Item analysis is the process of reviewing how each individual test item performed once real students answered it, using two core numbers: the difficulty index and the discrimination index. A TOS controls what the test is supposed to measure before it's given; item analysis tells you whether the items you actually wrote did that job. A perfectly planned TOS can still include a poorly worded item, a wrong answer key, or a question that accidentally gives away the answer in its phrasing, and difficulty and discrimination indices are what catch those problems after the fact, ideally before the score becomes part of a student's grade.
How Do You Compute the Difficulty Index?
The difficulty index (sometimes written as p, for proportion) is the number of students who answered an item correctly, divided by the total number of students who took the test. Say 40 students took the quarterly test and 30 got item #7 right:
Difficulty Index = 30 ÷ 40 = 0.75, or 75%
Despite its name, a higher difficulty index means an easier item, since it reflects the percentage who got it right, not how "hard" it felt. Under a convention published by the University of Washington's Office of Educational Assessment, items around 85% correct or higher are classified as easy, 51-84% as moderate, and 50% or below as hard. That same office notes that for a typical multiple-choice item, an ideal target difficulty near 70% correct tends to maximize how well the item separates students who know the material from those who don't, since items that are too easy or too hard stop discriminating between the two groups at all.
How Do You Compute the Discrimination Index?
The discrimination index measures whether an item is actually separating your stronger students from your weaker ones, the way a good test item should. Split your class into a high-scoring group and a low-scoring group based on total test score, commonly the top 27% and bottom 27% for large groups, or a simple top-half/bottom-half split for a typical class-sized group of 20-40 students. Then apply this formula:
Discrimination Index = (Correct in High Group − Correct in Low Group) ÷ Number of Students in One Group
Worked example: with 40 students split into a high group of 20 and a low group of 20, if 18 students in the high group got item #7 correct and 10 in the low group did:
Discrimination Index = (18 − 10) ÷ 20 = 8 ÷ 20 = 0.40
| Discrimination Index | Interpretation (UW Office of Educational Assessment) |
|---|---|
| Above 0.30 | Good — item clearly separates strong and weak students |
| 0.10 to 0.30 | Fair — usable, but worth reviewing for improvement |
| Below 0.10 | Poor — item is not doing its job |
| Negative | Red flag — more weak students answered correctly than strong students |
A negative discrimination index is the one result that needs immediate attention: it usually means the item was mis-keyed (the "correct" answer marked in your key is actually wrong), or the wording is genuinely confusing in a way that trips up students who otherwise know the material.
What Do You Do With a Bad Item After Item Analysis?
Once you've flagged items by difficulty and discrimination, decide item by item rather than throwing out anything imperfect:
- Very easy item (95%+ correct) with poor discrimination: Consider replacing it or moving it later in the test; almost everyone gets it right regardless of mastery, so it's not adding information.
- Very difficult item (below 20% correct) with poor or negative discrimination: Check the answer key first. If the key is correct, the item may be poorly worded, testing something not actually taught, or relying on a distractor that's arguably also correct.
- Good difficulty (51-84%) with good discrimination (above 0.30): Keep the item as-is; this is your target zone.
- Any item with negative discrimination: Re-check the key immediately, even if you've already released grades. If the key was wrong, that's a fairness issue that needs correcting, not just a note for next time.
- Items you plan to reuse: Keep a running item bank with difficulty and discrimination noted per item, so future tests can be built from items with a known track record instead of starting from zero each quarter.
Quick Reference for Building Your Next Test
- Build the TOS before writing items, using actual teaching days per topic, not a rough guess made after the fact.
- Don't cluster items at the "remembering" level just because those questions are faster to write; match cognitive-level distribution to how the topic was actually taught.
- Run difficulty and discrimination index calculations on every summative test, not just ones that "felt off"; problems in well-planned tests are often invisible until the numbers are computed.
- Treat negative discrimination as an error to investigate, not a quirky result to ignore.
- Keep a reusable item bank annotated with difficulty and discrimination values so item quality improves over time instead of resetting every quarter.
Ready to find the right school? Browse schools or compare tuition and reviews across hundreds of Philippine schools on SchoolFinderPH.
Sources
- University of Washington Office of Educational Assessment — Understanding Item Analyses
- DepEd Order No. 015, s. 2026 — Revised Guidelines on Classroom Assessment, Grading System, and Awards and Recognition (official)
- DepEd Order No. 8, s. 2015 — Policy Guidelines on Classroom Assessment (official, repealed by DO 015 s. 2026)
- PMC / NCBI — Item analysis: the impact of distractor efficiency on the difficulty index and discrimination power of multiple-choice items
- ILAW Lesson Plan — Table of Specifications (TOS) and Summative Test under DO 015, s. 2026
- SchoolFinderPH — DepEd Grading System Guide
- SchoolFinderPH — DepEd Transmutation Table Explained
- SchoolFinderPH — RPMS and IPCRF Guide
