LET Elementary Assessment of Learning — Principles of Assessment and the Table of SpecificationsRevision Notes

Exam Tips

• LET stem trigger: 'assigns numbers' or 'quantifies' → answer is MEASUREMENT.
• LET stem trigger: 'gathers evidence from multiple sources' or 'uses portfolios, tests, and observations' → answer is ASSESSMENT.
• LET stem trigger: 'renders a judgment,' 'decides pass or fail,' 'interprets scores against a standard,' or 'grades' → answer is EVALUATION.
• Remember the chain: Measurement → Assessment → Evaluation. Each step builds on the previous one.

Key Points

• Measurement assigns a number to a performance or trait. It answers the question 'How much?' Example: A Grade 5 pupil scores 38 out of 50 on a Science quiz.
• Assessment is the broader, systematic process of collecting evidence of learning from multiple sources — tests, observations, portfolios, interviews, and projects. It answers 'What is the evidence of learning?'
• Evaluation interprets that evidence and renders a value judgment or makes a decision against a standard. It answers 'How good is the performance? Does it pass?' Example: Deciding that 38/50 earns a grade of 'Satisfactory' based on the DepEd grading scale.
• The logical chain is: MEASURE first (get the numbers), then ASSESS (gather all evidence), then EVALUATE (judge and decide).
• Grading is evaluation, not just measurement, because it involves a judgment. Raw scores are measurement.
• Under the Code of Ethics for Professional Teachers, teachers must use evaluation results fairly, honestly, and solely for the benefit of learners — never to humiliate or harm them.

Definitions

Section Title

Measurement, Assessment, and Evaluation: Three Distinct Concepts

Common Mistakes

• Treating 'assessment' and 'evaluation' as synonyms. They are NOT. Assessment gathers evidence; evaluation judges it.
• Thinking that giving a test IS evaluation. Giving a test is measurement (or part of assessment). Only when you interpret the score against a standard does evaluation occur.
• Forgetting that assessment includes non-test tools like observations, portfolio review, and peer feedback — not just written exams.
• Confusing 'measurement' with 'assessment.' Measurement is just the quantifying step; assessment is the full process of evidence-gathering.

Exam Tips

• Key phrase 'to improve instruction' or 'give feedback while teaching' → Assessment FOR Learning.
• Key phrase 'to assign grades' or 'at the end of the unit/quarter' → Assessment OF Learning.
• Key phrase 'students monitor their own progress' or 'self-reflection journal' or 'metacognition' → Assessment AS Learning.
• DepEd DO 8, s. 2015 uses a 4-component grading system (Written Works 20-40%, Performance Tasks 40-60%, Quarterly Assessment 20-25%) — know this for context when LET items describe grading components.

Key Points

• Assessment FOR Learning is FORMATIVE in nature. It occurs DURING instruction. Its main purpose is to provide feedback that helps the teacher adjust instruction and helps learners improve while there is still time. The teacher is the primary user of the data.
• Assessment OF Learning is SUMMATIVE in nature. It occurs at the END of a unit, quarter, or school year. Its purpose is to certify and report what the learner has achieved. Results are used for grading and reporting to parents and the school.
• Assessment AS Learning develops METACOGNITION — the learner's ability to think about their own thinking and monitor their own progress. Tools include self-assessment checklists, learning journals, and reflection sheets. The STUDENT is the primary user of the data.
• DepEd's K-12 BEC and the DepEd Grading System (DO 8, s. 2015) embed all three purposes: Written Works, Performance Tasks, and Quarterly Assessments reflect different purposes.
• A quick decision rule: If the goal is FEEDBACK TO IMPROVE → FOR learning. If the goal is TO GRADE OR CERTIFY → OF learning. If the goal is STUDENT SELF-MONITORING → AS learning.
• All three purposes complement each other and should be present in effective classroom assessment practice.

Definitions

Section Title

Purposes of Assessment: FOR, OF, and AS Learning

Common Mistakes

• Equating 'formative' only with 'ungraded.' Some formative assessments carry small grades, but their PRIMARY purpose is still feedback and improvement.
• Forgetting that Assessment AS Learning is DIFFERENT from Assessment FOR Learning. Both involve feedback, but AS learning is student-driven; FOR learning is teacher-driven.
• Thinking that summative assessments cannot provide learning information. They can, but their PRIMARY purpose is certification and grading.
• Confusing peer assessment with Assessment AS Learning. Peer assessment can be used FOR or AS learning depending on how the teacher designs it.

Exam Tips

• LET trigger words for PLACEMENT: 'entry behavior,' 'readiness,' 'where to start,' 'before the school year begins.'
• LET trigger words for DIAGNOSTIC: 'specific difficulties,' 'causes of errors,' 'misconceptions,' 'learning disability,' 'prescriptive.'
• LET trigger words for FORMATIVE: 'during instruction,' 'feedback,' 'monitor progress,' 'improve learning while it is happening.'
• LET trigger words for SUMMATIVE: 'end of unit/quarter/year,' 'certify,' 'grade,' 'judge overall achievement,' 'quarterly exam.'

Key Points

• PLACEMENT assessment is given BEFORE instruction begins to determine the learner's entry-level knowledge, skills, and readiness — and to decide WHERE to place them in the instructional sequence. Example: A pre-test given on the first day of school to see if pupils already know the Grade 4 topics.
• DIAGNOSTIC assessment is given BEFORE or DURING instruction to pinpoint SPECIFIC learning difficulties, their causes, and the nature of errors. It goes deeper than placement. Example: An error-analysis test to find out WHY a Grade 3 pupil consistently makes mistakes in subtraction with regrouping.
• FORMATIVE assessment is given DURING instruction continuously to MONITOR progress and provide feedback. It guides both teacher and learner. Usually ungraded or carries minimal weight.
• SUMMATIVE assessment is given AT THE END of an instructional period (unit, quarter, year) to JUDGE and CERTIFY overall achievement. It is always graded.
• The most commonly confused pair is PLACEMENT vs. DIAGNOSTIC. Remember: Placement = 'Where do we start?' Diagnostic = 'What exactly is the problem and why?'
• Diagnostic assessment identifies CAUSES of learning difficulties, making it more analytical than placement assessment.
• Under RA 7836 and the Code of Ethics, teachers are mandated to use diagnostic results constructively — to help, never to label or stigmatize learners.

Definitions

Section Title

Types of Assessment: Placement, Diagnostic, Formative, and Summative

Common Mistakes

• Mixing up placement and diagnostic. Placement = sorting/readiness; Diagnostic = causes of specific difficulties.
• Thinking that 'pre-test' is always diagnostic. A pre-test can be a placement tool. Only when it pinpoints causes of errors is it truly diagnostic.
• Assuming formative assessment is always ungraded. Its defining feature is PURPOSE (monitoring/feedback), not whether it is graded.
• Forgetting that diagnostic can occur DURING instruction (not just before). If a teacher gives an error-analysis exercise mid-unit, that is still diagnostic.

Exam Tips

• If a test result is reported as a RANK or PERCENTILE → Norm-Referenced.
• If a test result tells you whether you PASSED or MASTERED a set of objectives → Criterion-Referenced.
• The LET, NAT (National Achievement Test), and driving tests are CRITERION-REFERENCED.
• The UPCAT, school entrance exams, and standardized IQ tests are NORM-REFERENCED.
• DepEd classroom grading (DO 8, s. 2015) is CRITERION-REFERENCED because it measures mastery of competencies.

Key Points

• NORM-REFERENCED TESTS (NRT) compare a learner's score to the scores of a NORM GROUP (a representative sample of test-takers). The key output is RELATIVE STANDING: rank, percentile, standard score.
• CRITERION-REFERENCED TESTS (CRT) compare a learner's score to a fixed CRITERION or standard — not to other people. The key output is whether the learner has MASTERED specific objectives.
• NRT example: The UPCAT, where your percentile rank tells you how you performed compared to all other examinees. Entrance exams and IQ tests are norm-referenced.
• CRT example: A driving test (pass/fail based on a set standard), a mastery test with 80% cut-off, and the LET itself (passing = weighted average of at least 75%, no component below 50%). The LET's passing standard is fixed — it does NOT change based on how many people passed.
• NRT items are designed to SPREAD out scores (some very easy, some very hard) to rank learners. CRT items are tied to SPECIFIC OBJECTIVES regardless of how they spread scores.
• For elementary classroom teaching in DepEd schools, CRT is more commonly appropriate because the goal is mastery of competencies, not ranking pupils against each other.
• The LET is CRITERION-REFERENCED because passing is based on a fixed standard (75%), NOT on being in the top X% of all examinees.

Definitions

Section Title

Norm-Referenced vs. Criterion-Referenced Assessment

Common Mistakes

• Thinking the LET is norm-referenced because many people take it. The LET's passing standard (75%) is FIXED — that makes it criterion-referenced.
• Confusing 'criterion' with 'criteria' in general. In NRT vs. CRT, 'criterion' specifically means a fixed performance standard, not just any requirement.
• Assuming CRT has no difficulty levels. CRT items can be easy, moderate, or difficult — what matters is that they are tied to objectives, not designed to spread scores.
• Thinking that NRT is 'bad' and CRT is 'good.' Both have valid uses. NRT is appropriate for selection (scholarships, entrance), while CRT is appropriate for mastery certification.

Formulas

Exam Tips

• If an LET item asks 'What tool does a teacher use to ensure content validity?' → Answer: TABLE OF SPECIFICATIONS (TOS).
• If asked to compute the number of items for a topic: use the formula (topic hours ÷ total hours) × total items.
• The TOS has TWO dimensions: CONTENT (rows) and COGNITIVE LEVEL (columns). Items go at the intersection.
• Remember the LET's own difficulty distribution: 30% easy, 50% moderate, 20% difficult — this is a TOS principle applied to the LET itself.
• If an LET item gives you hours and total items, ALWAYS check that your computed items add up to the given total.

Key Points

• The TABLE OF SPECIFICATIONS (TOS) is a two-way chart (matrix) that maps CONTENT TOPICS against COGNITIVE LEVELS (Bloom's Revised Taxonomy) and specifies the NUMBER OF ITEMS for each cell.
• The TOS is the primary tool for ensuring CONTENT VALIDITY — evidence that a test adequately and proportionately samples the intended content domain.
• A TOS prevents over-testing easy topics and under-testing important but difficult ones. It ensures the test reflects the EMPHASIS given during instruction.
• The TOS aligns test items with LEARNING OBJECTIVES and their intended cognitive level, preventing a test from becoming entirely recall-based when higher-order thinking was targeted.
• The LET's own Enhanced TOS follows a difficulty mix of approximately 30% easy, 50% moderate, and 20% difficult items.
• Bloom's Revised Taxonomy cognitive levels (lowest to highest): Remember → Understand → Apply → Analyze → Evaluate → Create.
• Steps to construct a TOS: (1) List content topics, (2) Identify objectives and their cognitive levels, (3) Determine total number of items, (4) Allocate items proportionally by instructional time or emphasis, (5) Distribute items across cognitive levels per topic.
• Formula for allocating items: Items per topic = (Hours spent on topic ÷ Total instructional hours) × Total number of items.
• The TOS is a professional and ethical tool — it protects learners from unfair testing by ensuring the test covers what was actually taught.

Definitions

Section Title

The Table of Specifications (TOS): The Test Blueprint

Common Mistakes

• Forgetting that the TOS establishes CONTENT validity, not criterion or construct validity.
• Allocating items based on topic importance alone, ignoring instructional time. The standard practice is to weight items by TEACHING TIME or emphasis given.
• Constructing a TOS only at the REMEMBER level, producing a low-quality test that does not measure higher-order thinking.
• Confusing the TOS with a lesson plan. The TOS is a TEST blueprint, not an instructional plan.
• Not making the TOS items sum to the total number of test items. Always check that all cells add up correctly.

Exam Tips

• LET stem: 'What does the TOS ensure?' → CONTENT VALIDITY.
• LET stem: 'A test that looks like it measures what it should, based on inspection' → FACE VALIDITY (weakest).
• LET stem: 'Scores on a new test correlate with scores on an established test given at the same time' → CONCURRENT VALIDITY.
• LET stem: 'Scores on a test are used to forecast success in college' → PREDICTIVE VALIDITY.
• LET stem: 'Measures a theoretical trait like reading ability or creativity' → CONSTRUCT VALIDITY.
• Memory shortcut: Validity types by strength (strongest to weakest): Construct ↔ Criterion-related (Concurrent/Predictive) ↔ Content → Face (weakest).

Key Points

• VALIDITY is the most important quality of any test. It is the degree to which a test MEASURES WHAT IT IS INTENDED TO MEASURE and the appropriateness of the interpretations made from the scores.
• CONTENT VALIDITY: The test adequately samples the content domain. Checked through EXPERT JUDGMENT against a TOS. This is the most relevant type for teacher-made classroom tests.
• CRITERION-RELATED VALIDITY: The test scores correlate with an external criterion. It has TWO subtypes: (a) CONCURRENT validity (scores relate to a criterion measured at the SAME TIME) and (b) PREDICTIVE validity (scores forecast FUTURE performance).
• CONSTRUCT VALIDITY: The test measures the theoretical construct or trait it claims to measure (e.g., intelligence, reading comprehension, anxiety). Verified through statistical methods like factor analysis.
• FACE VALIDITY: The test APPEARS to measure what it should, based on superficial inspection. It is the WEAKEST form — a test can look valid but not actually be valid. It is not considered a rigorous type of validity.
• Concurrent validity example: A new reading test score correlates highly with scores on a well-established reading test given at the same time.
• Predictive validity example: UPCAT scores predict (correlate with) college GPA — this is why entrance tests are used for selection.
• Key principle: Validity refers to the APPROPRIATENESS OF INTERPRETATIONS, not just the test itself. A test is not 'valid' in isolation — it is valid FOR a specific purpose and group.

Definitions

Section Title

Validity: Measuring the Right Thing

Common Mistakes

• Thinking face validity is sufficient. It is the weakest type and is not considered genuine scientific evidence of validity.
• Confusing concurrent and predictive validity. Remember: CONCURRENT = same time; PREDICTIVE = future.
• Saying 'a test is valid' without specifying what it is valid FOR. Validity is always specific to a purpose and a group.
• Mixing up content validity with face validity. Content validity uses EXPERT JUDGMENT and the TOS; face validity is just surface appearance.
• Forgetting that construct validity is for ABSTRACT TRAITS like intelligence or attitude — not for subject-matter content.

Formulas

Exam Tips

• LET trigger: 'same test, two occasions, same group' → TEST-RETEST reliability.
• LET trigger: 'two equivalent forms, same group' → PARALLEL FORMS reliability.
• LET trigger: 'odd and even items, one test' → SPLIT-HALF reliability (remember Spearman-Brown).
• LET trigger: 'right or wrong scoring, internal consistency, one administration' → KR-20 or KR-21.
• LET trigger: 'Likert scale, attitude measure, internal consistency' → CRONBACH'S ALPHA.
• Reliability coefficient of 0.80 or above is generally acceptable. Values close to 1.0 indicate high reliability.

Key Points

• RELIABILITY is the CONSISTENCY, STABILITY, or DEPENDABILITY of the scores a test produces across repeated measurements, different forms, or different parts of the test.
• A reliable test gives similar results under similar conditions. If a pupil takes the same test twice and gets very different scores for no apparent reason, the test is unreliable.
• TEST-RETEST RELIABILITY: Give the SAME test to the SAME group at TWO DIFFERENT TIMES and correlate the two sets of scores. This measures STABILITY over time. Weakness: memory and practice effects on the second sitting.
• PARALLEL (EQUIVALENT) FORMS RELIABILITY: Give TWO DIFFERENT but EQUIVALENT forms of the test to the SAME group and correlate. This measures EQUIVALENCE of forms. Weakness: requires two truly equivalent tests.
• SPLIT-HALF RELIABILITY: Give ONE test, split it into TWO halves (typically odd vs. even items), score each half separately, and correlate. Measures INTERNAL CONSISTENCY. Requires the SPEARMAN-BROWN CORRECTION FORMULA to compensate for the shorter length of each half.
• INTERNAL CONSISTENCY (KR-20, KR-21, Cronbach's Alpha): Analyzes how all items within ONE test relate to each other in a single administration. KR-20 and KR-21 are used for DICHOTOMOUS items (right/wrong). Cronbach's Alpha is used for SCALED items (e.g., Likert-type attitude scales).
• The RELIABILITY COEFFICIENT ranges from 0 to 1. The closer to 1, the more consistent the test. A coefficient of 0.80 or above is generally considered acceptable for educational tests.
• Factors that INCREASE reliability: more items, greater score variability, clear directions, appropriate difficulty level, good test conditions.
• Factors that DECREASE reliability: too few items, ambiguous questions, inconsistent scoring, guessing, fatigue, noisy testing environment.

Definitions

Section Title

Reliability: Measuring Consistently

Common Mistakes

• Forgetting to apply the Spearman-Brown formula after computing split-half correlation. The raw half-test correlation underestimates the full-test reliability.
• Confusing KR-20 with Cronbach's Alpha. KR-20 is for RIGHT/WRONG items; Cronbach's Alpha is for SCALED items.
• Thinking a higher reliability coefficient always means a better test. Reliability must be paired with validity — a consistent but invalid test is useless.
• Assuming test-retest reliability is free from error. Memory effects, practice, and changes in the learner between sittings all affect the coefficient.
• Forgetting that reliability coefficients range from 0 to 1 (not -1 to +1 like Pearson r for correlations in general).

Exam Tips

• Memorize this exactly: 'A test can be reliable without being valid, but a valid test is always reliable.'
• If an LET option says 'reliability is necessary but not sufficient for validity' → this is CORRECT.
• If an LET option says 'a valid test may or may not be reliable' → this is WRONG.
• Dartboard analogy for the LET: Bullseye + tight group = Valid + Reliable. Off-center + tight group = Reliable only. Scattered = Neither.
• The most important quality of a test is VALIDITY; reliability SUPPORTS validity but does not replace it.

Key Points

• RELIABILITY is NECESSARY but NOT SUFFICIENT for validity. You need consistency BEFORE validity is even possible, but consistency alone does NOT guarantee you are measuring the right thing.
• A test can be RELIABLE BUT NOT VALID: It gives consistent scores but measures the WRONG construct or content. Example: A ruler that is always 1 cm too long gives consistent (reliable) but inaccurate (invalid) measurements.
• A VALID TEST IS ALWAYS RELIABLE: If a test truly and accurately measures what it should, it must do so consistently. Validity implies reliability.
• A test can be NEITHER RELIABLE NOR VALID: Random, inconsistent scores that also miss the target completely.
• A test CANNOT be VALID BUT UNRELIABLE: An unstable, inconsistent test cannot meaningfully measure the right thing.
• THE DARTBOARD ANALOGY: Arrows clustered ON the bullseye = VALID AND RELIABLE. Arrows clustered OFF-CENTER (tight group, wrong spot) = RELIABLE BUT NOT VALID. Arrows scattered EVERYWHERE = NEITHER valid nor reliable.
• This relationship is one of the MOST FREQUENTLY TESTED concepts in the Assessment of Learning section of the LET.

Definitions

Section Title

The Validity-Reliability Relationship: A Critical LET Topic

Common Mistakes

• Saying a test that is reliable is also valid. WRONG. Reliability does NOT guarantee validity.
• Thinking a valid test could be unreliable. WRONG. If a test is truly valid, it MUST also be reliable.
• Forgetting the order of logic: Reliability must come first; validity builds on it.
• Confusing the dartboard analogy: Scattered arrows = NEITHER valid NOR reliable (not 'valid but unreliable,' which is impossible).