Overview
A questionnaire is a measurement instrument, not a list of questions. Each construct in your framework (for example perceived usefulness, burnout or financial literacy) needs a clear definition, a set of items that cover that definition, and a response format that suits the analysis you plan. When this mapping is loose, no amount of statistics later can rescue it: you may find low reliability, factors that do not separate, or a key hypothesis that no item actually tests.
We start by building a construct-to-item matrix with you, deciding for each construct whether to adopt an established validated scale, adapt one, or write new items. Where you use published scales, we help you trace the original source, check whether the author or publisher requires permission or a license, and record every wording change so your methodology chapter can report it accurately.
Before you collect your main sample, we help you pretest the draft through cognitive interviews and a small pilot. You receive a revised instrument, a pretest log explaining what changed and why, and preliminary reliability estimates that tell you whether each scale is behaving as expected.
Who it’s for and the problems it solves
This is for you if you are…
- Master's and doctoral students designing a survey for a thesis built on a conceptual framework
- Researchers adapting a validated scale to a new country, language, age group or sector
- Supervisors or ethics committees who have asked for evidence of instrument validity or a pilot study
- Institutional and nonprofit teams building a questionnaire for program feedback or needs assessment
- Anyone whose draft questionnaire has grown long and unfocused and needs trimming against the research questions
Common problems we help with
- You are not sure which items measure which construct, or some constructs have only one item
- You found a scale in a journal article but do not know whether you may use or modify it
- Items are leading, double-barreled or use jargon respondents will not understand
- You cannot decide between a 5-point and 7-point Likert format, or whether to include a neutral midpoint
- Your supervisor asked for content validity evidence and a pilot, and you have neither
- A previous survey produced low Cronbach's alpha and you want to avoid repeating the problem
What’s included — and what isn’t
Included
- Construct operationalization: working definitions and a construct-to-item matrix linked to each research question and hypothesis
- Search for established validated scales and a comparison of candidate instruments (length, reported reliability, population used, license terms)
- Guidance on requesting permission from scale authors or publishers and documenting adaptations
- Item writing and review for clarity, single focus, neutrality, reading level and cultural fit
- Response scale design: Likert and semantic-differential formats, number of points, midpoint and 'not applicable' options, anchor labels
- Reverse-worded items, attention-check placement and question order to reduce order and social desirability effects
- Content validity support, such as an expert review form and content validity index calculation
- Cognitive pretesting protocol (think-aloud and verbal probing) and a pretest findings log
- Pilot study plan and preliminary reliability analysis (Cronbach's alpha, McDonald's omega, item-total correlations)
- A final instrument with a codebook showing variable names, item wording, scale sources and scoring rules
Not included
- Using copyrighted or licensed instruments without the permission their owners require; we help you ask, but the permission must be yours
- Inventing reliability or validity figures for a scale that has not been tested
- Collecting data or recruiting respondents on your behalf unless separately agreed as part of an institutional project
- Writing the methodology chapter for a graded thesis; we give you documentation you can describe in your own words
What you receive
- Construct-to-item matrix
- A table linking every research question and hypothesis to its constructs, definitions, items, source and planned analysis, so gaps and redundancies are visible at a glance.
- Draft and revised questionnaire
- A clean instrument in Word or your survey platform's import format, with instructions, consent text placeholder, section introductions and response scales.
- Scale source and permissions log
- For each adopted or adapted scale: full citation, license or permission status, and a record of wording changes to report in your methods.
- Pretest and pilot report
- Summary of cognitive interview findings, item changes made, pilot descriptives and reliability estimates, with recommendations before main data collection.
- Codebook
- Variable names, labels, value codes, reverse-scored items and scoring instructions ready for SPSS, R or Stata.
How it works
Framework review
You share your research questions, framework and any draft items. We identify constructs, flag ones without measures and agree the target respondent group and length.
Source and draft
We locate candidate scales, check permission requirements, and assemble or write items into a first draft with a construct-to-item matrix.
Expert and cognitive review
We help you run a content-validity review with subject experts and a handful of cognitive interviews, then revise wording based on what respondents actually understood.
Pilot and reliability
You pilot the instrument with a small sample from your target population. We analyze completion times, item distributions and reliability, then recommend final edits.
Finalize
You receive the final instrument, codebook and documentation, plus a short call to go through any open decisions.
Design and testing techniques we use
- Construct operationalization
- Defining each construct from theory and literature, choosing reflective or formative measurement, and setting a minimum of three to four items per reflective construct where latent-variable analysis is planned.
- Likert and rating scale design
- Choosing 5- or 7-point agreement, frequency or satisfaction scales with fully labeled anchors, and deciding deliberately about midpoints and 'don't know' options.
- Content validity index (CVI)
- Structured expert ratings of item relevance, summarized at item and scale level, often requested by health and education committees.
- Cognitive interviewing
- Think-aloud and scripted probes to check comprehension, recall, judgment and response mapping, the approach described in Willis's widely used cognitive interviewing guidance.
- Reliability estimation
- Cronbach's alpha, McDonald's omega and corrected item-total correlations from pilot data, interpreted cautiously given small pilot samples.
- Translation and adaptation
- Forward and back translation with reconciliation for multilingual instruments, coordinated with our research instrument translation service.
Adopt, adapt or write new items?
Using an established scale saves time and lets you compare results with earlier studies, but only if it fits your population and definition. A scale validated with nurses in one health system may need wording changes for teachers elsewhere, and every change weakens the link to the original validation evidence.
We help you weigh the options item by item. Minor adaptation (changing 'my organization' to 'my school') is usually acceptable if reported. Substantial rewording, dropping items or changing the response format means you should treat the scale as newly adapted and plan to re-examine its structure, for example through confirmatory factor analysis.
Permissions and licensing for published scales
Appearing in a journal article does not make a scale free to use. Some authors release items openly for research; others ask you to request permission; some instruments (many clinical and psychological measures) are commercially licensed and must be purchased or registered. Translating a scale often needs separate permission.
We help you identify the owner and the terms, draft a polite permission request, and keep the reply on file for your ethics application and appendix. If a license is out of reach, we suggest openly available alternatives with comparable evidence.
Common item-writing mistakes we fix
Most weak items fall into a small number of patterns. Catching them before the pilot avoids wasted responses.
- Double-barreled items: 'My manager is supportive and fair' asks two things at once
- Leading wording: 'How much did you enjoy the excellent training?'
- Undefined time frames: 'Do you exercise regularly?' instead of 'In the past 7 days…'
- Overlapping or incomplete response categories, especially for age and income bands
- Negations inside agreement scales that confuse respondents ('I do not dislike…')
- Sensitive questions placed too early, before rapport and context are established
Pilot testing: what a small sample can and cannot tell you
A pilot of a few dozen respondents shows whether people finish the survey, how long it takes, which items get skipped, and whether responses cluster at one end of the scale. It gives a first look at internal consistency. It is too small to confirm a factor structure, so we report pilot reliability as provisional and plan the full psychometric checks for your main sample.
What the work can look like
Illustrative example — fictional data
Adapting a burnout and engagement survey for secondary-school teachers
A fictional master's student wants to test whether workload predicts teacher burnout through reduced work engagement, using two published scales originally developed for corporate employees. The draft questionnaire has 64 items and no pilot.
Matrix and trimming
The construct-to-item matrix showed that 14 items measured nothing in the hypotheses. Removing them and using the short form of the engagement scale cut the questionnaire to 38 items, with an estimated completion time of about 10 minutes.
Permissions and adaptation
One scale was openly available for non-commercial research; the other required emailed permission, which the student requested and received before the ethics submission. Six items were reworded from 'company' to 'school', each change recorded in the adaptation log.
Cognitive interviews and pilot
Five teachers took part in think-aloud interviews; two items about 'clients' were rewritten as 'students and parents'. A pilot with 32 teachers gave provisional alphas between 0.79 and 0.88 for the three scales, with one item showing a corrected item-total correlation of 0.18, flagged for review.
Illustrative, invented scenario and figures. Real instruments depend on your field, population and your supervisor's or committee's expectations.
What you’ll need to provide
- Research questions, hypotheses and a diagram or description of your conceptual framework
- Any existing draft questionnaire or scales you intend to use, with their sources
- Details of your target respondents: language, literacy, setting and how they will complete the survey
- Your institution's or ethics committee's requirements for instruments and pilots
- Pilot data (anonymized) when you reach that stage
You don’t need everything ready to ask for a quote — tell us what you have and we’ll say what’s missing.
What affects price and turnaround
Price depends on
- Number of constructs and whether scales are adopted, adapted or newly written
- Whether content validity review and cognitive interviews are included
- Number of languages or versions of the instrument
- Whether pilot data analysis and reliability reporting are part of the scope
- Complexity of skip logic and respondent groups
Turnaround depends on
- Time to hear back from scale authors about permissions, which is outside anyone's control
- Availability of experts and pilot respondents on your side
- Number of revision rounds with your supervisor or team
Every project is quoted individually. How we price our work.
Questions people ask before starting
Can I just copy a questionnaire from a published paper?
Sometimes, but check first. Published items may be copyrighted or licensed, and the paper may report only part of the scale. We help you find the original source, confirm the terms and record any permission you receive so you can cite it properly in your methods and ethics documents.
Should I use a 5-point or 7-point Likert scale?
Both are common and defensible. Seven points can capture finer distinctions for motivated, literate respondents; five points can be easier on mobile screens and for lower-literacy samples. The more important choice is consistency with the original validated scale, because changing the number of points alters the instrument.
How many people do I need for a pilot?
Pilots are about feasibility, not significance testing, so there is no single number. Many studies use somewhere between ten and fifty respondents from the target population. We help you pick a size that answers your pilot questions and explain why a pilot is too small to confirm factor structure.
Is a pilot the same as a pretest?
They overlap but differ. Cognitive pretesting uses a few in-depth interviews to see how people interpret each question. A pilot runs the whole survey with a small sample under realistic conditions to check timing, flow, drop-off and preliminary reliability. Many committees like to see both.
Will you write the instrument for my graded thesis?
The instrument is a research tool, and designing it together is legitimate methodological support. You make the decisions, approve every item and describe the development process in your own words. Follow your institution's rules on acknowledging help with instrument development.
Can you set the questionnaire up online too?
Yes. Once the content is final we can build it in Qualtrics, Google Forms, SurveyMonkey or KoboToolbox through our online survey setup service, including skip logic and attention checks.