Test Development

Main ideas / concepts / lessons

• High-quality large-scale tests (e.g., international assessments like ILS and TOL) produce reliable, useful results, but reaching that quality is time-consuming and expensive.
• Major test development follows a cycle of phases; classroom test development is essentially the same process, with differences in:
  • Content
  • Timing
  • Format
  • Reporting
• For classroom testing, teachers:
  • Know their learners well, so they may not need to write extensive profiles from scratch.
  • Still should maintain a class profile (and update it yearly) to make future test creation easier.
• Even for classroom quizzes, teachers should do basic post-test checks so items measure student ability appropriately and the test can be reused/improved.
• Well-monitored tests can support building a test bank for repeated use.

Test development methodology (cycle of phases)

1) Planning phase

• Decide why a new test is needed.
• Determine what kind of test it will be.
• Identify potential candidates (who will take it).
• Identify test users (who the test results will be for/use by).

2) Design phase

• Develop test specifications.
• Define:
  • Test features
  • Range of topics
• Write sample materials.

3) Development phase (try-out + feedback)

• Try out the materials.
• Collect feedback from everyone involved.
• Use feedback to make changes.
• Changes may require returning to planning and/or design to “get it right.”

4) Operational phase

After specifications are agreed and finalized, proceed to:

• Construct the full test.
• Administer the complete test.

5) Monitoring phase

• Review performance on the new test.
• Revise and update materials so they still match the original aim.

Outcome: Completing the cycle yields evidence that can support future test changes. In classroom settings, the same logic applies: develop, administer, analyze, and refine.

Classroom test development: key steps and decisions (detailed checklist)

A) Build a class profile (recommended)

Briefly note personal information such as:

• Age
• Class level

Consider knowledge/background:

• What the students know about the topic
• Whether learners share roughly similar background knowledge or if it varies

Estimate language strength/weakness (how strong/weak the group’s language is).

Keep the profile so it can be updated every year.

B) Use a “decision table” to plan the test (record details)

Include decisions such as:

• Purpose of the test
• Which skills to test
• Which language elements to test
• Time constraints
• Test organization
  • Number of sections
  • Length of each section
• Test methods by section (e.g., multiple choice, gap fill)
• Number of items per section
• Scoring/weighting
  • Whether all sections count equally
  • Or whether some sections earn more marks

Record as much detail as possible to make the test easier to replicate for another class later.

C) Plan operational/admin details

Make notes about:

• Timing (is there enough time, too much, too little?)
• Instruction clarity
• Whether students need to ask for clarification
• Problems observed during administration
• What you said/clarified to resolve issues

D) Do basic analysis after students take the test

Ensure questions are not:

• Too easy
• Too difficult

Rationale:

• If items are too easy/hard, they may not provide useful information about student ability.

Review and align again:

• Test content
• Specifications
• Administration

Use monitoring results to support a test bank (a resource that can be reused repeatedly).

E) Final aim

• A well-developed, monitored test provides reliable and useful information.
• That information helps teachers support learners to learn more effectively.

Speakers / sources featured

• No individual speakers or specific source organizations are identified by name in the subtitles (beyond general references to international tests such as ILS and TOL).