Constructing Teacher Made Achievement Tests: A Step-by-Step Process

Every test a teacher writes is, in effect, a claim: “This assessment reflects what my students were taught and what they should have learned.” But making good on that claim requires far more than sitting down with a blank page and writing questions. Constructing a Teacher-Made Achievement Test (TMAT) is a deliberate, structured process that begins long before the first question is drafted. When done well, a TMAT becomes a meaningful measure of student learning – fair, focused, and directly aligned with what was taught. The process unfolds across five major steps: identifying instructional objectives, making the design, preparing a blueprint, writing test items, and developing a marking scheme.

Step 1: Identifying instructional objectives 🔗

The foundation of any good achievement test is a clear set of instructional objectives. According to ERIC, the learning objectives a teacher emphasizes determine both the material to include and the form the test will take – everything else in the construction process flows from this step.

Instructional objectives describe precisely what students should know or be able to do after a period of instruction. They need to be specific, measurable, and directly linked to the curriculum. For instance, an objective like “students will be able to explain the causes of World War I” is far more useful for test construction than a vague goal like “students will understand history.” The more clearly objectives are stated in behavioral terms, the more precisely the test can be targeted to measure them.

It also helps to think about the cognitive level each objective demands. Bloom’s Taxonomy – which classifies learning from basic recall through comprehension, application, analysis, synthesis, and evaluation – is a practical framework here. Research shows that teachers tend to construct questions at the knowledge level 80 to 90 percent of the time, often neglecting higher-order thinking. Consciously mapping objectives across cognitive levels from the start corrects this imbalance.

Step 2: Making the design 🔗

Once the objectives are set, the next step is to decide on the overall structure of the test. The design is essentially a policy document for the assessment – it specifies how marks and emphasis will be distributed before a single question is written.

The design addresses four key areas of weightage: the instructional objectives to be tested and the proportion of marks each receives; the types or forms of questions to be used (multiple-choice, short answer, essay); the units and sub-units of the course content; and the levels of difficulty (easy, average, difficult). This distribution reflects the relative importance of different content areas and cognitive demands in the curriculum.

For example, if a unit on fractions was taught over three weeks while a unit on geometry took only one week, the design should reflect that difference in coverage. A well-considered design ensures the test is a fair sample of everything students were expected to learn, not a random or unrepresentative selection of questions.

Step 3: Preparing the blueprint (table of specifications) 🔗

The blueprint – also called a table of specifications – is where the design decisions made in Step 2 are given concrete, visual form. It is a two-way matrix that maps content areas against cognitive levels of objectives, showing exactly how many test items should address each combination.

A typical blueprint lists content topics along one axis and the levels of Bloom’s Taxonomy along the other. Each cell in the grid specifies the number of items – and sometimes the marks – assigned to that content-objective combination. The distribution of items across cells reflects both the emphasis given to each topic during instruction and the relative importance of each cognitive level. A topic that received more instructional time naturally commands more test items.

The blueprint serves two critical functions. First, it ensures content validity – that the test actually covers what was taught in appropriate proportions. Research confirms that preparing a table of specifications helps a test achieve high content validity and provides a fair, balanced measure of student learning. Second, it acts as a practical guide during item writing, directing the teacher to write questions in areas that need coverage rather than drifting toward topics that are simply easier to assess.

Blueprints can also be shared with students before instruction begins, making learning expectations transparent and giving students a clear framework for what they will be held accountable for. This transparency has been shown to improve student engagement and performance.

Step 4: Writing test items 🔗

With a blueprint in hand, the teacher is ready to write the actual questions. The blueprint dictates where questions must come from; this step determines what those questions look like.

Choosing item types 🔗

Achievement tests typically use a combination of objective items (multiple-choice, true/false, matching, fill-in-the-blank) and subjective items (short answer, essay). Multiple-choice questions are the most commonly used type because they can be graded objectively and are harder to answer correctly by guessing alone – but they are also the most difficult to write well. Essay questions, by contrast, are better suited to assessing higher-order thinking but require more careful scoring.

A good TMAT uses a mix of question types to assess different kinds of thinking. Long or convoluted questions should be avoided; each item should be clear, concise, and unambiguous so that the student’s response reflects their knowledge, not their ability to decode a confusing prompt. It is also good practice to generate more items than specified in the blueprint, since some items may be discarded after review or item analysis.

Principles of good item writing 🔗

Regardless of question type, every item should test a specific learning objective, use straightforward language, and avoid cultural or linguistic bias. For multiple-choice items, distractors (incorrect options) must be plausible enough to distinguish students who have mastered the content from those who have not. For essay questions, the prompt should clearly state what is expected – the scope, the level of detail, and any structural requirements. A substantial proportion of items should target cognitive levels above simple recall – application, analysis, and evaluation – to ensure the test measures genuine understanding rather than memorization.

After drafting, the test items should be reviewed – ideally by a colleague – to catch errors, ambiguities, or unintended clues. A question-wise analysis at this stage checks whether each item aligns with the blueprint in terms of objective, content area, question type, difficulty level, and marks. This also helps determine the content validity of the paper before it is administered.

Step 5: Developing a marking scheme 🔗

The final step is to create a marking scheme – a clear, written guide that specifies exactly how each question will be scored. A marking scheme is not an afterthought; it is as important as the test itself, because it determines whether grading will be fair, consistent, and transparent.

Assigning marks 🔗

Each question should be allocated marks proportional to its complexity and the depth of response it requires. A straightforward multiple-choice item might be worth one mark; a multi-part essay requiring analysis and synthesis could be worth ten or more. These allocations should be consistent with the weightage decisions made in the design stage.

Rubrics for open-ended questions 🔗

For short-answer and essay items, a rubric is essential. A rubric is an assessment tool that clearly indicates achievement criteria across all components of student work, describing what performance looks like at each level. An analytic rubric breaks the response into separate criteria – for example, accuracy of content, quality of reasoning, and clarity of expression – and scores each independently. A holistic rubric assigns a single overall score based on the general quality of the response. Clear, descriptive performance labels such as “exemplary,” “proficient,” “developing,” and “needs improvement” help both the teacher and the student understand exactly where a response falls and why.

Answer keys for objective items 🔗

For objective questions, a complete answer key must be prepared alongside the test – listing the correct answer for every item. Sharing the marking criteria with students before the assessment – not as a surprise after the fact – increases fairness and helps students understand what is expected of them.

Together, the answer key and rubric constitute a complete marking scheme that makes scoring systematic and defensible. When multiple teachers grade the same paper, a detailed marking scheme significantly improves consistency across graders.

Why the process matters: alignment and validity 🔗

These five steps are not bureaucratic formalities – they serve a fundamental purpose. A well-designed test blueprint ensures that assessments cover a well-rounded mix of content, match the cognitive levels required by learning outcomes, avoid redundancy and trivial questions, and fairly represent the course material. When a test is built this way, its scores carry real meaning: they reflect what students actually learned, not the teacher’s unconscious biases or the arbitrary selection of whatever questions were easiest to write.

This alignment between objectives, instruction, and assessment is what distinguishes a purposefully constructed TMAT from a hastily assembled quiz. Research indicates a positive relationship between the use of a table of specifications and student academic achievement, partly because students who know that all domains of learning will be tested tend to study more comprehensively and deeply. The test, in other words, shapes learning – which is precisely why its construction deserves this level of care.

What do you think? If you were to construct a TMAT for a subject you teach or have studied, which of the five steps do you think teachers are most likely to skip or rush – and what consequences might that have for the accuracy of the test results? And how might sharing the test blueprint with students in advance change the way they prepare for an assessment?

References

1. https://eric.ed.gov/?id=ED315426
2. https://www.yourarticlelibrary.com/statistics-2/teacher-made-test-meaning-features-and-uses-statistics/92607
3. https://www.anthology.com/blog/using-blueprints-to-align-course-objectives-with-assessments
4. https://testbook.com/question-answer/the-first-step-in-the-construction-of-an-achieveme–63aec7a477c9cdf96b04582c
5. https://assess.com/test-blueprints-specifications/
6. https://blog.ansi.org/anab/creating-table-specifications-test-blueprint/
7. https://www.seahipublications.org/wp-content/uploads/IJISSHR-M-8-2022.pdf
8. https://files.eric.ed.gov/fulltext/ED315426.pdf
9. https://electronicinformationallibrary.medium.com/bloom-taxonomy-and-table-of-specification-36d28d699500
10. https://www.proftesting.com/test_topics/steps_3.php
11. https://uwaterloo.ca/centre-for-teaching-excellence/catalogs/tip-sheets/rubrics-useful-assessment-tool
12. https://teaching-resources.delta.ncsu.edu/rubric_best-practices-examples-templates/
13. https://www.niu.edu/citl/resources/guides/instructional-guide/rubrics-for-assessment.shtml
14. https://tlconestoga.ca/assessments-with-purpose-leveraging-test-blueprints-to-organize-question-libraries-and-optimize-question-pool-selection/
15. https://eajournals.org/wp-content/uploads/Table-of-Specification-and-its-Relevance-in-educational-development-Assessment.pdf