When you design a questionnaire, one of the first big decisions is how to measure your key constructs. Should you use a scale that other researchers have already developed and tested? Should you adapt it to fit your context? Or should you build a completely new scale?
Each option has advantages and risks. The choice affects the quality of your data, the time your project needs, and how easily you can defend your methods in a viva or peer review. This post walks through the options, the criteria for choosing, and the steps involved in each path.
Three options: adopt, adapt or develop
- Adopt: use an existing validated scale exactly as published, with the original items, response format and scoring.
- Adapt: use an existing scale but modify it, for example by changing wording to suit a new context, shortening it, translating it, or changing the response format.
- Develop: create a new scale from scratch, based on a clear definition of the construct and a structured development and validation process.
Most PhD projects adopt or adapt. Developing a new scale is sometimes necessary, but it is a major piece of work that can become a study of its own.
Why using an existing scale is usually the first choice
Established scales offer several benefits:
- Existing evidence of reliability and validity. Someone has already tested whether the items measure the construct consistently and accurately.
- Comparability. Your results can be compared with previous studies that used the same measure. This strengthens your discussion chapter.
- Credibility. Reviewers and examiners are more comfortable with well-known measures.
- Time savings. You avoid the lengthy process of item generation, expert review and multiple rounds of testing.
Even so, "established" does not mean "suitable for every study". You still need to evaluate whether the scale fits your research.
How to evaluate an existing scale
Before adopting a scale, check it against these criteria:
1. Conceptual fit
Does the scale measure the same construct, defined in the same way, as your study? Two scales with the same name can measure different things. Read the original development paper, not just the item list, and compare the scale's definition with yours.
2. Evidence of reliability and validity
Look for reported internal consistency, factor structure, and evidence of convergent and discriminant validity. Check whether this evidence has been replicated in later studies, not only in the original paper.
3. Population and context
Was the scale developed and tested with a population similar to yours? A scale validated with university students in one country may not work the same way with factory workers in another. Look for studies that used the scale in contexts close to yours.
4. Length and respondent burden
A 40-item scale may be excellent, but if your questionnaire already measures six constructs, respondents may become tired and careless. Check whether a validated short form exists.
5. Permission and licensing
Some scales are free for research use; others are copyrighted and require permission or a fee. Check the original publication and contact the authors if unsure. Keep a record of any permission you receive.
6. Language and readability
Are the items easy for your respondents to understand? Items written decades ago or for a different culture may contain outdated or unfamiliar expressions.
When adapting a scale makes sense
Adaptation is appropriate when a scale fits your construct well but needs changes to work in your study. Common reasons include:
- Changing the context, for example from "my company" to "my university".
- Translating the scale into another language.
- Shortening the scale to reduce respondent burden.
- Adjusting the response format for consistency across your questionnaire.
- Updating outdated terms or examples.
Levels of adaptation
It helps to think of adaptation as ranging from minor to major:
- Minor: changing a context word, such as "organisation" to "hospital". The meaning of the item stays the same.
- Moderate: rewording items for clarity, changing the number of response points, or removing a few items.
- Major: translating into another language, changing the target population substantially, or rewriting many items.
The more substantial the adaptation, the more evidence you need that the adapted version still works. Minor changes may need only a pilot study and a reliability check. Major changes need expert review, pre-testing and factor analysis to confirm the structure.
Good practice when adapting
- Keep a table showing each original item and your adapted version.
- Explain why each change was made.
- Ask subject experts to review whether the adapted items keep the original meaning.
- Pre-test with a few people from your target population.
- Pilot the adapted scale and check reliability.
- In the main study, test the factor structure, ideally with confirmatory factor analysis if the original structure is well established.
- Cite the original scale and clearly describe it as "adapted from".
When developing a new scale is justified
Developing your own scale may be the right choice when:
- No existing scale measures your construct as you define it.
- Your construct is new, or it is being studied in a very new context, such as a recently emerged technology or practice.
- Existing scales have serious weaknesses, such as poor reliability, weak validity evidence, or outdated content that cannot be fixed by adaptation.
- Your research contribution is partly the development of a new measurement tool.
Be honest with yourself about the reason. "I could not find a scale" should mean you searched thoroughly, not that a quick search returned nothing. Examiners may ask you to show the scales you considered and why you rejected them.
The main stages of scale development
Scale development generally follows a sequence like this:
- Define the construct clearly, including its dimensions, using literature and theory.
- Generate items from literature, interviews, focus groups or expert input. Write more items than you expect to keep.
- Expert review for content validity: are the items relevant, clear and complete?
- Cognitive pre-testing with members of the target population to check understanding.
- Pilot testing to check item distributions and remove clearly weak items.
- Exploratory factor analysis on one sample to identify the underlying structure.
- Confirmatory factor analysis on a separate sample to test that structure.
- Reliability and validity testing, including convergent, discriminant and, where possible, criterion validity.
This process usually needs more than one data collection round and a substantial number of respondents. It can easily take several months. If you plan to develop a scale, build this time into your research plan from the start and discuss it with your supervisor.
A quick decision guide
| Situation | Best option |
|---|---|
| A validated scale matches your construct and population | Adopt |
| A validated scale matches your construct but needs context or language changes | Adapt, with re-validation |
| A scale exists but is too long for your survey | Use a validated short form if available; otherwise adapt carefully |
| Several partial scales exist, each covering part of your construct | Consider combining with care, or develop a new scale |
| No suitable scale exists after a thorough search | Develop a new scale |
A worked example: adapting a scale for a new setting
Imagine you are studying psychological safety among nurses in public hospitals. You find a well-known psychological safety scale that was originally developed with work teams in companies. The construct definition matches yours, and the scale has been used in many later studies. However, some items refer to "the team" in a general business sense, and your respondents work in shifting ward teams where membership changes daily.
Here is how you might proceed:
- Check permission. Confirm whether the scale is free to use for research, and cite it properly.
- Decide the unit of reference. After discussion with your supervisor and two senior nurses, you decide that "my ward team on a typical shift" is the most meaningful reference for your respondents.
- Revise wording minimally. Replace "this team" with "my ward team" and keep all other wording unchanged.
- Expert review. Ask three experts (a nursing academic, a ward manager and a methodologist) to confirm that each adapted item still reflects the original meaning.
- Cognitive interviews. Ask five nurses to explain in their own words what each item means. One item about "taking risks" is understood as clinical risk rather than interpersonal risk, so you add a short clarifying phrase.
- Pilot and main study. Check reliability in the pilot, then test the original factor structure with confirmatory factor analysis in the main sample.
- Report transparently. Include a table showing original and adapted items, and describe every change in your methodology chapter.
Notice that one small wording issue, found through cognitive interviews, could have changed how respondents interpreted the whole scale. This is why even minor adaptations deserve some testing.
Questions to ask before you decide
- Have I searched thoroughly for existing measures in databases, review articles and measurement handbooks for my field?
- Does the definition behind the scale match my construct definition?
- Has the scale been used successfully in a population like mine?
- What changes would I need to make, and how substantial are they?
- Do I have the time and sample size to re-validate an adapted scale, or to develop a new one?
- Can I explain and defend my choice clearly to an examiner?
If you can answer these questions confidently, your choice will be much easier to justify.
Common mistakes to avoid
- Mixing items from different scales without testing whether they form a coherent measure.
- Calling a heavily modified scale "adopted". If you changed it, say "adapted" and explain how.
- Dropping items to improve reliability without considering whether the construct is still fully covered.
- Ignoring permission requirements for copyrighted instruments.
- Assuming validity transfers automatically to a new population or language.
- Underestimating the work involved in developing a new scale.
How to report your decision in the thesis
In your methodology chapter, include:
- The name, authors and year of each scale used.
- Whether it was adopted, adapted or newly developed.
- Reasons for choosing it, including fit with your construct definition.
- Details of any adaptations, with a comparison table in an appendix.
- Reliability and validity evidence from the original studies and from your own data.
- Permission details where relevant.
Clear reporting shows that your measurement choices were deliberate and well-justified, not simply convenient.
Final thoughts
For most research projects, a well-chosen existing scale, adopted or carefully adapted, gives you stronger and more comparable data than a new scale built under time pressure. Develop your own scale only when there is a real gap, and plan enough time and data to validate it properly. Whichever path you choose, document every decision. That documentation is what turns a measurement choice into a defensible methodology.
Are you deciding between an existing scale and your own? Share your construct in the comments and we can discuss the options.