Construction and Validation of a Blueprint-Based Achievement Test in Class 8 Algebra

This M.Ed dissertation constructs and validates a standardised achievement test in Class 8 algebra, covering two units every Indian Class 8 child meets: algebraic expressions and identities and linear equations in one variable. Teachers write unit tests every month, but very few of those tests are planned against objectives, tried out, analysed item by item or checked for reliability. As a result, marks often reflect how a question paper happened to be set rather than what children actually learnt.

The study follows the classical sequence of achievement-test construction. A blueprint crosses content units with objectives from Bloom's revised taxonomy (remember, understand, apply, analyse) and item formats (multiple choice and short answer). About 60 items are written, reviewed by mathematics teachers and teacher educators, and tried out on 300 Class 8 students from government and private schools. Item analysis computes the difficulty index (p), the discrimination index (D) using upper and lower 27% groups, and a distractor analysis for every multiple-choice option. The best 40 items form the final test.

The final test is then validated: content validity through expert judgement against the blueprint, concurrent validity against school examination marks, reliability by KR-20 and test–retest, and norms (percentiles and z-scores) from a norming sample. The result is a test any Class 8 teacher can use, and a dissertation that shows mastery of measurement — not just statistics.

Abstract

Algebra is the gateway from arithmetic to abstract mathematics, and Class 8 is where most Indian students either make the jump or start to fall behind. Classroom tests rarely measure this transition reliably. This study constructs and validates a 40-item achievement test in Class 8 algebra based on a blueprint of content and Bloom's revised taxonomy objectives. Sixty items were tried out on 300 students; item difficulty, discrimination and distractor effectiveness were analysed to select the final items. Reliability was established by KR-20 and test–retest methods, validity by expert content validation and correlation with school examination marks, and percentile and z-score norms were prepared. The test is intended for diagnostic and summative use by teachers.

Introduction

Achievement testing in Indian schools has moved from annual examinations towards continuous assessment and, under NEP 2020, towards competency-based assessment. Both need tests that measure what they claim to measure. An achievement test is standardised when its content is planned, its items are tried out and analysed, its reliability and validity are known, and norms allow a score to be interpreted.

Review and gap

Indian M.Ed and PhD studies have constructed achievement tests in many subjects, but several report only a single reliability coefficient and skip distractor analysis or norms. International research on algebra learning documents stable misconceptions — treating letters as labels, "conjoining" terms such as 2a + 3b = 5ab, and applying inverse operations incorrectly when transposing. Few Indian tests deliberately use these misconceptions as distractors, which is what makes a multiple-choice item diagnostic.

Objectives in brief

To prepare a blueprint, write and try out items, select items through item analysis, establish reliability and validity, prepare norms, and compare achievement by gender and school management as a demonstration of use.

Feasibility

• Schools accept a 45-minute test gladly when results are returned to teachers.
• Item analysis and KR-20 need only Excel; SPSS or Jamovi speeds up correlations and norms.
• Twenty weeks cover blueprint, try-out, analysis, final testing, retest and writing.

Class 8 mathematics teachers assess algebra with teacher-made tests that are rarely planned against objectives or analysed after use. Such tests tend to over-represent recall and routine procedures, include items that almost everyone gets right or wrong, and contain distractors that no student chooses. Scores therefore say little about whether students can apply algebra or where their misconceptions lie, and results from different schools cannot be compared.

The problem of this study is stated as: "Construction and validation of an achievement test in algebra for Class 8 students." The study aims to provide a blueprint-based test with known difficulty, discrimination, reliability, validity and norms, whose distractors are built on documented algebra misconceptions so that teachers can use results diagnostically as well as for grading.

In scope

• Two units of NCERT Class 8 mathematics (state-board equivalents are similar): algebraic expressions and identities; linear equations in one variable.
• Objective (multiple-choice) and short-answer items, dichotomously scored for item analysis.
• Try-out on 300 students; final testing and norming on 400 students from government and private English/regional-medium schools in one district.
• Reliability, validity and norms; a demonstration comparison by gender and management.

Delimitations

• Geometry, mensuration and data-handling units are excluded.
• Long-answer proof or reasoning items are not included because they need rubric scoring.
• Norms are local and valid for the academic year and district of testing.

Research design

Test construction and standardisation, followed by a descriptive survey for norming and group comparison.

Stage-wise procedure

Final blueprint (40 items, 50 marks)

Sampling

Stratified random sampling of schools (government / private × rural / urban) from the district UDISE+ list; all Class 8 students present in selected sections are tested.

Timeline (20 weeks)

Weeks 1–3 review and blueprint · 4–6 item writing and review · 7–8 try-out · 9–10 item analysis · 11–13 final testing and retest · 14–15 validity and norms · 16–20 analysis, writing and viva preparation.