Pre- and Post-Training Test Design: Comparable Forms and Score Changes

This guide concentrates on the measurement design of pre- and post-training tests: comparable tasks, timing, practice effects, paired scores and uncertainty. It does not replace a wider evaluation of training effectiveness, business outcomes or programme costs.

An organisation pays for a training programme. Employees attend, the trainer delivers the sessions, and participants receive certificates. But one question remains: Did employees actually learn what the organisation needed them to learn?

Pre- and post-training assessments can help answer that question. A pre-training assessment checks what participants know or can do before training begins. A post-training assessment checks their knowledge or skills afterwards. Comparing the two results can show evidence of learning, provided the assessments are well designed and interpreted carefully.

However, a higher score does not automatically mean that employees will perform better at work. Effective evaluation combines assessment results with observations, practical tasks and relevant business measures.

This guide explains how to do that without complicated terminology.

Why measure learning before and after training?

A pre/post design asks whether observed performance on a defined measure changed between two occasions. It does not by itself establish that training caused the change. Specify the measured skills, test-form comparability, administration conditions, missing observations and the workplace outcome to check later. Diagnosing which employees need training is a separate task addressed in the training-needs analysis guide.

Without a starting measurement, an employer may know that employees performed well after training but not how much they improved. Without a follow-up measurement, the employer may know employees attended but not whether the programme achieved its learning objectives.

A well-planned before-and-after approach helps employers:

• Identify the knowledge employees already possess.
• Avoid teaching material everyone has mastered.
• Match training to genuine skill gaps.
• Check whether participants understood key lessons.
• Identify topics that require additional support.
• Compare training approaches more responsibly.
• Decide whether to repeat, revise or discontinue a programme.
• Demonstrate the basis for development decisions.

These benefits depend on assessment quality. Poor questions and inconsistent conditions can produce misleading conclusions.

Start with the job, not the test

Before writing questions, define what employees should be able to do differently after training.

Consider an accounts team learning bank reconciliation. A vague objective is “understand reconciliation.” A more useful objective is “identify outstanding deposits, unpresented payments and recording errors in a sample bank reconciliation.”

The second objective can be tested through a practical task. It also gives the manager something observable to check later at work.

A good training plan links three elements:

The assessment should measure the stated learning objectives, not unrelated general intelligence or obscure facts.

What should a pre-training assessment include?

The first assessment should cover the knowledge and skills the training is intended to develop. It should be taken before participants receive the teaching materials or coaching being evaluated.

For example, a 40-question assessment for payroll administrators might contain:

The employer should record the assessment version, instructions, time allowance, scoring rules and date. Participants should understand why the results are being collected and how they will be used.

A pre-training test is not necessarily a performance appraisal. Someone may have a low starting score because the material is new to them. That is often exactly why training is needed.

What should a post-training assessment include?

The follow-up assessment should measure the same learning objectives as the first assessment. It should not simply reward remembering the trainer's examples word for word.

Good post-training questions ask employees to apply what they learned to realistic situations. For payroll training, an employee might examine a fictional payroll schedule, identify an inconsistency and choose the correct next step.

The second assessment should have comparable coverage and demands. If the first test asks straightforward calculations while the second asks complex multi-step cases, a direct comparison may be unfair.

Employers have two common options:

Use the same questions. This makes the content identical, but participants may remember answers from the first attempt. A higher score could partly reflect familiarity rather than new learning.

Use different questions covering the same skills. This reduces answer recall, but the employer must examine whether the forms are sufficiently comparable. Having 40 questions in each form does not guarantee equal difficulty.

For more detail, see Comparing Assessment Results Across Different Test Forms: A Practical Employer Guide.

Worked example: Measuring improvement in a 40-question test

Suppose a fictional organisation trains five payroll assistants. Each completes a 40-question test before training and a suitably comparable assessment afterwards.

For Ada, the starting percentage is 20 ÷ 40 × 100 = 50%. The later percentage is 30 ÷ 40 × 100 = 75%. Her improvement is 25 percentage points.

The team answered 120 questions correctly before training and 155 afterwards, out of 200 possible answers at each sitting. Its average score increased from 60% to 77.5%, a rise of 17.5 percentage points.

These are accurate arithmetic comparisons for the fictional example. They are not, by themselves, proof that training caused all the improvement. Practice, form differences, outside learning and changes in testing conditions may also affect results.

Percentage points are not the same as percentage growth

An increase from 50% to 75% is a rise of 25 percentage points. Relative to the starting score, it is a 50% increase because 25 ÷ 50 × 100 = 50%.

For clear management reports, percentage-point changes are often easier to understand. Always label the measure so readers do not confuse the two.

What can the results tell the employer?

The example suggests that employees performed better on the assessed content after training. It also shows that improvement varied across participants.

Dami improved from 45% to 65% but may still need help with some skills. Efe began at 75% and improved to 85%; the smaller gain does not necessarily mean the training was ineffective for Efe. Someone who starts with strong knowledge has less room to improve on a test with a maximum score of 100%.

A responsible report looks at both starting level and later performance, not just the size of the increase.

It should also examine results by topic. A team may improve substantially in payroll calculations while still struggling to respond correctly to employee payroll queries.

Look beyond the overall score

Suppose the average post-training score is 78%. That sounds encouraging, but a single total can hide important gaps.

These illustrative topic results show why a follow-up plan matters. The employer may need to revise the section on payroll queries, add practical exercises or investigate whether that topic was adequately covered.

Do not automatically label a training programme a failure because one topic did not improve. Check the questions, starting level, teaching approach and workplace relevance.

Five mistakes that can distort before-and-after comparisons

1. Testing different skills

A pre-test that measures basic knowledge and a post-test that measures advanced problem-solving may be useful for separate purposes, but its scores are not straightforward measures of improvement on the same scale.

2. Reusing questions without considering recall

Employees may remember answers, especially when tests are taken close together. Where the same questions are necessary, acknowledge this limitation rather than treating every extra mark as new learning.

3. Changing the conditions

A quiet pre-test taken on a laptop and a rushed post-test taken on a phone during a busy shift may not provide a fair comparison. Instructions, access arrangements and time limits should be appropriately consistent.

4. Treating every increase as proof of training impact

Employees can learn through practice, colleagues, personal study and work experience. Without an appropriate evaluation design, the employer should say results are consistent with learning, not that training alone caused the change.

5. Ignoring employees who did not complete both tests

If 30 employees take the pre-test but only 20 complete the post-test, comparing the average of all 30 with the average of the 20 may be misleading. The missing participants could differ from those who finished.

For a clear matched comparison, report results for people who completed both assessments, and separately disclose participation and non-completion.

When should the post-training test take place?

The best timing depends on the learning objective.

An immediate post-test can check whether participants understood the material at the end of a course. It is useful for identifying topics that need clarification.

A later follow-up assessment can examine whether employees retained the knowledge after returning to work. A follow-up after several weeks may be appropriate for some programmes, while others need a different interval.

A workplace review checks whether employees actually use the skill in their job. This may involve reviewing completed work, observing a task or measuring relevant errors.

A useful evaluation plan may include all three. Do not assume a fixed interval suits every role or training programme.

Measuring workplace application

A pre/post assessment estimates change in performance on the assessed tasks under the stated testing conditions. Workplace application is a separate outcome: decide in advance which observable work behaviours or error patterns matter, when they can reasonably be observed and who will review them.

Compare relevant workplace evidence before and after the learning activity where feasible, while documenting changes in workload, supervision, systems or task difficulty. Do not attribute a reduction in errors solely to training without considering these competing explanations. Broader training effectiveness and organisational impact belong in the existing learning-and-development evaluation guide.

Example: Assessment scores and workplace errors

A fictional team records 20 avoidable payroll corrections in the month before training and 12 in a comparable month afterwards.

The observed reduction is 8 corrections. Expressed relative to the starting figure, that is 8 ÷ 20 × 100 = 40% fewer recorded corrections.

This may be encouraging, but the employer should ask whether the workload, number of payroll transactions, staff mix or checking procedures changed. A month with fewer transactions could produce fewer errors even without better skills.

A more informative measure may be errors per 1,000 processed payroll entries, if reliable transaction counts are available. The employer should also consider trends over several periods rather than relying on one month.

How to evaluate training when groups are small

Small teams are common in growing organisations. An employer might train only six accounts staff or four new supervisors.

Small groups can still provide useful feedback. Their results may reveal unclear questions, persistent learning gaps or the need for extra practice. But a few scores should not be presented as strong evidence that the same programme will work equally well across all departments.

Report the number of participants. Show individual or skill-level results where appropriate and privacy permits. Explain that conclusions are limited by the size and characteristics of the group.

See Small Sample Sizes in Assessment Pilots: What Employers Can Conclude for further guidance.

How to protect fairness and employee trust

Employees may worry that a pre-training score will be used to embarrass them or harm their promotion prospects. Employers should explain the purpose before testing begins.

Good practice includes:

• Tell participants what is being assessed and why.
• Explain who can see individual results.
• Separate development feedback from disciplinary decisions unless a different lawful, clearly communicated process applies.
• Provide reasonable accessibility arrangements where needed.
• Give consistent instructions and appropriate testing conditions.
• Limit access to identifiable results.
• Keep records only as long as necessary for legitimate purposes.
• Allow employees to raise concerns about errors or interruptions.

A development-focused programme should help people improve. It should not turn every training activity into an unexpected high-stakes examination.

A practical results report for management

Managers need a short, accurate account of what changed and what remains uncertain.

A useful report might include:

The numbers are illustrative. Such a report should explain whether the assessments were sufficiently comparable and avoid presenting the score increase as proof of workplace impact.

For managers reviewing employee results, see Assessment Score Reports for Hiring Managers: A Practical Guide for broader principles of clear score reporting.

A four-week implementation plan

Week 1: Define success

Meet with department leaders to identify the business problem. Write specific learning objectives and decide what evidence would show improvement. Agree on the intended use of individual and team results.

Week 2: Prepare the assessments

Develop or select questions that match the objectives. Check accuracy, clarity, difficulty and scoring. Plan whether the same form or a reviewed alternative form will be used. Establish fair administration rules.

Week 3: Run the pre-test and deliver training

Administer the baseline assessment before teaching begins. Record participation and relevant testing conditions. Use the results to adapt training, but avoid unfairly labelling employees based on one score.

Week 4: Assess immediate learning and plan follow-up

Administer the post-test, compare matched results, review skill-level gaps and prepare a clear report. Schedule later checks of retention and workplace application. Four weeks may not be enough to establish lasting improvement.

Employer checklist

Before launching a before-and-after training evaluation, confirm that:

• The training addresses a clearly defined work need.
• Learning objectives describe observable knowledge or skills.
• Pre- and post-tests cover the same intended skills.
• Questions and answer keys have been reviewed.
• Any differences between test forms have been considered.
• Participants receive clear instructions and fair conditions.
• Scores can be matched to the same participants appropriately.
• Non-completion is recorded and explained.
• Reports distinguish percentage points from percentage growth.
• Results are examined by topic, not just overall score.
• The limits of the comparison are stated.
• A workplace follow-up measure is planned.
• Individual results are handled confidentially.
• Managers have a plan for remaining skill gaps.

Frequently asked questions

What is the difference between a pre-test and a post-test?

A pre-test measures knowledge or skills before training. A post-test measures them afterwards. Their comparison can provide evidence of learning when the tests and conditions support a fair comparison.

Should both tests contain the same questions?

Not necessarily. Identical questions make the content consistent but can encourage answer recall. Different questions reduce recall but require checks for comparable skill coverage and difficulty.

Is a higher post-training score proof that training worked?

It is evidence that assessed performance improved. It is not automatic proof that the training caused the entire increase or that workplace performance improved.

How many questions should a training assessment contain?

There is no universal number. The assessment should cover the important objectives with enough suitable questions to support its intended use. A 40-question format is one possible design, not a guarantee of accuracy.

What if an employee's score falls after training?

Investigate before drawing conclusions. Check for unclear questions, different form difficulty, technical problems, stress, misunderstanding or genuine remaining learning needs. Offer constructive feedback and appropriate support.

Can employers compare results across departments?

Only when the assessments measure suitably comparable skills and the comparisons make sense for the roles involved. A 75% result in payroll should not automatically be treated as equivalent to 75% in customer service.

Should training scores determine promotions?

A training score may contribute relevant evidence, but promotion decisions should use a broader, fair process based on job requirements, performance and other appropriate information.

What is the best way to measure training success?

Combine evidence of learning with relevant workplace outcomes. A useful evaluation asks what participants learned, whether they retained it and whether they applied it in their work.

Conclusion

Pre- and post-training assessments give employers a practical way to examine whether employees are learning. The most useful approach begins with clear objectives, uses carefully reviewed questions, compares results responsibly and follows up on actual workplace performance.

A higher score is a useful signal, not the entire story. Organisations gain the greatest value when they use assessment results to improve training, support employees and solve the business problems that made training necessary in the first place.

For the broader treatment, see the existing Talentrenza article: Learning and Development Assessment: How to Measure Training Effectiveness.