Before you adapt an existing questionnaire: what to check and document
Check permissions, changes, pretesting, and measurement evidence before adapting a research questionnaire. Includes a worked adaptation log and checklist.
About 8 min read · Allow extra time for the practical checks
You have found a questionnaire that seems close to what your study needs. A few words do not fit your setting, some questions feel unnecessary, and you would prefer a shorter response scale. It is tempting to make those changes and describe the result as a validated questionnaire.
Pause before doing that. You need to establish both whether the changes are permitted and what they could do to the meaning of the responses. Evidence supporting the original instrument does not automatically support every version made from it.
Adapting an existing questionnaire can be reasonable. It should be a documented measurement decision, not an unrecorded language edit.
Find the original instrument and identify what you actually need
Start with the original publication, manual, or official distribution source. A questionnaire reproduced in someone else’s dissertation may be incomplete, modified, or missing its instructions.
Identify the instrument’s version, intended construct, target respondents, recall period, response options, and scoring rules. Find the evidence supporting the use you have in mind, including relevant evaluations beyond the original development study.
Also distinguish a questionnaire from a scale. A questionnaire may contain background questions, separate subscales, and items that are not intended to form a combined score. Taking selected questions from it does not necessarily produce a shorter version of an established measure.
Write down what your own research question requires. If you want to investigate access to support, an instrument measuring satisfaction with support may only partly fit. Changing its title would not resolve that difference.
Check permission for the specific use and changes
Read the current licence or rights-holder instructions. Establish whether they cover administration, modification, translation, electronic delivery, and reproduction of items in a thesis or publication. These can be separate questions.
The International Test Commission’s adaptation guidelines place necessary permission before adaptation. Where the terms do not clearly cover your intended use, obtain clarification from the rights holder, which may be a publisher or distributor rather than the original author.
Keep the applicable licence or written permission with your study records. If the existing terms clearly permit the use, record how they apply; do not assume that every instrument requires a separate permission email. Equally, being able to download a questionnaire does not establish permission to change or republish it.
An ethics approval and an instrument licence answer different questions. Neither substitutes for the other. Check your institution’s requirements for participant work, including pretesting, before recruitment begins.
Treat each proposed change as a decision about measurement
Some changes are easier to notice than others. Translating an instrument obviously changes its presentation. Replacing “during the past month” with “usually” may look minor, but it changes the period a respondent considers.
Before editing, record what you intend to change and why:
On a narrow screen, scroll the table sideways to read every column.
| Proposed change | Question to resolve before proceeding |
|---|---|
| Replace a workplace term | Does the replacement refer to the same role or process? |
| Remove items | What part of the construct would no longer be represented? |
| Change response options | Will the responses and resulting scores retain their intended meaning? |
| Alter the recall period | Are you asking about the same experience over a different period, or a different kind of judgement? |
| Move from paper to online delivery | Do layout, routing, required answers, or device constraints change how people respond? |
| Translate the questionnaire | How will linguistic and cultural meaning be evaluated? |
This table is a planning aid, not a rule that every change has the same effect. The appropriate evaluation depends on the instrument, intended interpretation, population, and extent of modification.
Worked example: a change log that exposes hidden decisions
This example is synthetic. The instrument, wording, organisation, and decisions are invented. The item shown is not taken from a published scale and should not be treated as a validated measure.
Suppose a researcher wants to adapt a workplace-support questionnaire for staff in a service organisation. An illustrative item reads:
During the past four weeks, how often could you obtain help from your supervisor when you needed it?
The researcher proposes replacing “supervisor” with “team leader,” removing the timeframe, and deleting an item about help outside normal working hours.
An adaptation log might begin as follows:
On a narrow screen, scroll the table sideways to read every column.
| Proposed change | Reason offered | Decision and evidence needed |
|---|---|---|
| “Supervisor” becomes “team leader” | This is the local job title. | Establish whether team leaders have the same support responsibilities. Ask intended respondents whom the phrase brings to mind. |
| Remove “during the past four weeks” | The sentence reads more smoothly. | Retain the timeframe unless there is a substantive reason to change it. Smoothness alone does not justify changing the judgement being requested. |
| Delete the out-of-hours item | It seems irrelevant to office staff. | Check whether any eligible staff work outside standard hours and what the item contributes to the original construct. Do not remove it solely on an assumption. |
All changes remain subject to the instrument’s actual permissions. The log does not authorise modification.
Suppose pretesting then shows that some employees treat the team leader as the person who allocates tasks, while others think of the manager who approves leave. The replacement has not clarified the intended meaning for everyone. The next step is to revisit the role definition and wording, then evaluate the revised version. Recording “wording checked” would conceal the problem.
Keep earlier versions of the log. Include the decision, its basis, the version affected, and the person responsible for approving it. A later reader should be able to distinguish an initial proposal from a change that was actually made.
Use pretesting to investigate how people answer
Asking colleagues whether the questionnaire “looks fine” can reveal obvious mistakes, but it does not show how intended respondents interpret the questions.
Cognitive interviewing examines the processes people use to answer: what they understand, what they recall, how they judge an answer, and how they select a response. The CDC’s explanation of cognitive interviewing describes this as question evaluation rather than population estimation.
For the synthetic item, useful probes would ask which person the respondent considered, which incidents they recalled, and how they chose between response options. Do not coach them towards the interpretation you hoped to hear. The misunderstanding is information you need.
Pilot administration can address additional practical issues such as routing, missing responses, completion burden, and whether the delivery process works. Plan these activities around specific questions. There is no single pilot size that makes every adapted questionnaire acceptable.
Keep findings as well as actions. “Participants interpreted the role label differently; wording was revised and examined again” is more useful than “A pilot study was conducted.”
Do not treat translation as a complete validation process
A translation can read naturally while changing what an item means. A familiar phrase in one setting may refer to a different practice in another.
The ITC guidance addresses adaptation beyond literal translation, including cultural fit, administration, score interpretation, and supporting evidence. Translation and review should involve the relevant language, subject, and measurement expertise. Back-translation can contribute to that process, but it does not by itself demonstrate equivalent measurement.
If you plan to compare language or population groups, ask what evidence supports that comparison. Similar-looking questionnaires and acceptable reliability estimates in each group do not, on their own, establish that score differences mean the same thing. Seek methodological advice about the equivalence evidence needed for your intended comparison.
State what the evaluation establishes—and what remains uncertain
“Validated” is often used as if it were a permanent label attached to a questionnaire. A more useful report identifies the evidence available for particular scores, respondents, and uses.
For example, distinguish evidence about item relevance, respondent interpretation, internal structure, consistency, and relationships with other measures. Boateng and colleagues’ scale-development and validation primer treats these as parts of a broader process. A reliability coefficient alone does not establish that the adapted questionnaire measures the intended construct.
If your work only examined comprehensibility and practical administration, report that. Do not imply that a small pretest established all the measurement properties required for consequential individual decisions or cross-group comparisons.
A study can be transparent about limited evaluation without presenting the instrument as either unquestionably valid or entirely useless. Explain the evidence obtained, why it is relevant, and which interpretations remain tentative.
Report the version you administered
Your methods account should identify the source instrument and explain the changes actually used in data collection. Report the adaptation rationale, relevant permissions, pretesting and evaluation, scoring decisions, and limitations.
Make sure the questionnaire, participant-facing instructions, scoring notes, and methods section describe the same version. If further changes occur after data collection starts, preserve the record and consider their implications with your supervisor or methods adviser. Do not quietly combine different versions as though every respondent received identical questions.
Before administration, check:
- Have I located the correct instrument and supporting documentation?
- Is it a good fit for the construct and intended respondents?
- Are the planned uses and changes permitted?
- Have I documented each change and its reason?
- Have intended respondents helped evaluate comprehension where appropriate?
- Are scoring and missing-response decisions specified?
- Does my description of evaluation stay within the evidence obtained?
If the adaptation would remove much of what the instrument measures, reconsider whether another instrument would be more suitable. Our research support can help you examine those methodological choices and understand what further evaluation may be needed. It does not turn an adapted questionnaire into a validated measure by declaration.
