A Survey Instrument, Displayed
A transcription is only worth having if you can do something with it. This one is followed by what each question type buys you at analysis, what the option sets permit once the data are in, and the things this instrument was never able to answer.
What the document is, before what it says
Six weeks of data-collection teaching in the supplied material describe instruments, question types and response formats without showing one. This one closes that. It belongs to a single examined master's project on change management from an entrepreneurial perspective, and is observed practice, not a model.
The file is a six-page A4 PDF printed from a browser. It returns no extractable text — the pages are images — so it was rendered and read visually. Each page carries a header with the date, time and form title, and a footer with the form's editing address and a page number.
The instrument, transcribed
Everything below appears on the form in this order. Every question carries a red asterisk marking it Required and the italic instruction "Mark only one oval." Every option is a radio button, and the wording is the instrument's own.
A section header reads DEMOGRAPHIC QUESTIONS, then four questions.
QUESTIONS 1-4, AS PRINTED, WITH RESPONSE OPTIONS IN THE ORDER SHOWN
| No. | Question as printed | Response options, in order |
|---|---|---|
| 1 | "What is your age?" | 18-30 years · 31-40 years · 41-50 years · >50 years |
| 2 | "What is gender?" | Male · Female · Other |
| 3 | "What is your experience (as an entrepreneur & working as an employee combined) in the industry?" | 0-5 years · 5-10 years · 10-20 years · 20-40 years · >40 years |
| 4 | "How big is your enterprise in terms of employee strength?" | <10 · 10-50 · 50-100 · 100-250 · >250 |
Each question is numbered on the form, marked Required, and carries "Mark only one oval."
A second section header on page 2 reads ENTREPRENEURIAL RESPONSE QUESTIONS, and question 5 sits alone on page 3.
Page 4 then carries an untitled block that is neither question nor section: the heading "There are some key change management models which are found to be widely applicable and effective in change management processes", the line "Brief descriptive definitions of these models:", and five definitions with references. It briefs respondents before question 6.
THE FIVE MODEL DEFINITIONS AS THE FORM PRINTS THEM, IN ORDER
| Model, by structure | Definition as printed | Reference as printed |
|---|---|---|
| An eight-step model | "…designed to help leaders successfully implement organisational change by focusing on creating urgency … It strides through initiating, managing, and sustaining change in eight steps." | An online management-skills resource, 2022 |
| A five-letter individual-adoption model | "…a bottom-up process that focuses on individuals and their commitment to the success of a Change initiative by 'Awareness', 'Desire', 'Knowledge', 'Ability'…" | A named author's website, 2022 |
| A seven-component alignment model | "…recognises seven components of an organisation that must work together for effective change management 'Structure', 'Strategy', 'Staff', 'Style', 'Systems', 'Shared Values', and 'Skills'." | A corporate-finance training institute, 2022 |
| A three-step model | "…applies three steps 'unfreezing', 'changing', and 'refreezing' … creating the perception that a change is needed, then moving toward the new, desired level of behavior and finally, solidifying that new behavior as the norm." | An online course provider, 2022, page title recorded as "No page title" |
| A five-stage curve model | "…explains the change and personal evolution of the grievous process highlighting 'denial', 'anger', 'bargaining', 'depression', and 'final acceptance.'" | A product-onboarding vendor's blog, 2022 |
URLs scrubbed. All five references are dated 2022; none is peer-reviewed.
QUESTIONS 6-8, AS PRINTED, WITH RESPONSE OPTIONS IN THE ORDER SHOWN
| No. | Question as printed | Response options, in order |
|---|---|---|
| 6 | "Which model do you think is the most suitable for entrepreneurial change management in particular?" | The five models above, as named options: eight-step · individual-adoption · seven-component · three-step · curve |
| 7 | "What, according to you, is at the top priority as a component of change management processes?" | Leadership Alignment · Stakeholder engagement · Employee communication and training · Change readiness · Organizational structure & design |
| 8 | "According to you which is the crucial aspect to be considered for when employee communication is the concern in organizational change processes?" | Communicating the need and process for change · Communicating the action plans · Getting feedback from the employees · Conveying the influencing role that the employees are going to play · Provide the information about how the change will benefit them |
That is the whole instrument. Page 6 carries only the platform wordmark.
What each question type buys you
Every question here is what the supplied material calls a structured question: the respondent chooses from options the researcher wrote — see structured and unstructured questions. Nothing needed coding after collection, and nobody could say anything the form had not anticipated.
WHAT EACH QUESTION YIELDS, AND WHAT IT RULES OUT
| Question | Data it produces | Supports | Forecloses |
|---|---|---|---|
| 1, 3, 4 — age, experience, size | Ordered bands | Counts per band; a sample profile; cross-tabulation | Any statistic needing a number: no mean age, no median years |
| 2 — gender | Three unordered categories | Counts and percentages of a known base | Anything outside the three labels offered |
| 5 — importance | An intended rating scale | Aggregation of adjacent categories whose order is clear | Any claim about the distance between two ratings |
| 6, 7, 8 — model, priority, communication | One unordered choice from five | A most-selected option and a split summing to 100% | Ranking, weighting, multiple selection, and anyone who would choose none of the five |
Data types as at levels of measurement and attribute, ordinal and numerical data. Column three is what you can write; column four is what you gave up before the first response arrived.
What the option sets permit once the data are in
Because every question was compulsory, the fifty responses arrived with no blank cell anywhere. The base is fifty on every question and every question's categories sum to one hundred per cent: no imputation rule, no missing-data convention, no differing bases.
Two moves follow legitimately: a straight count per category, which is what all eight tabulations are, and aggregation of adjacent categories whose order is clear. This thesis combined the top two importance categories, reported thirty-five of fifty (seventy per cent), and said which two — which is what makes the figure checkable. It survives a full trace back to the raw data.
Three properties an option set either has or does not
- Exhaustive
- Everyone has an option. Question 6 is not: a respondent who knows none of the five models has no route through it.
- Mutually exclusive
- Nobody qualifies twice. Questions 3 and 4 fail at their boundaries: 5, 10, 20 and 40 years, and 50, 100 and 250 employees, fall in two bands.
- Monotonic
- Options run in one direction. Question 5 does not — "Important" is printed below "Fairly Important".
The two versions, and which one governed
The same eight questions also exist as a word-processed questionnaire: tables with a row of option labels and an empty row for marking, and the consent acknowledgement above the title "Survey Questionnaire". It is not the same instrument.
FIVE DEMONSTRABLE DIFFERENCES BETWEEN THE TWO VERSIONS
| # | The form as issued | The questionnaire document |
|---|---|---|
| 1 | Q5: "How important change management tools are for implementing…" | Q5: "How vital are change management tools for implementing…" |
| 2 | Q6 ends "…in particular?" and is marked Required | Q6 ends "…in particular? And if you don't have experience in any of these models, please skip this question." |
| 3 | Q2 reads "What is gender?" | Q2 reads "What is your gender?" |
| 4 | Section headers present | No section headers |
| 5 | "Mark only one oval" and a required marker on every question | No response instruction and no required markers anywhere |
Five differences, counted for this library; the source states no total.
The second is the one that matters. One document grants a permission — skip question 6 if you have no experience of the models — and the other withdraws it, because a required question cannot be skipped. A respondent who knew none of the five chose one anyway or abandoned the survey.
What this instrument cannot answer
An instrument's limits are fixed when it is fielded, and every one is visible on the page beforehand.
Five things this form was never going to establish
Whether change management helps entrepreneurs succeed
Both hypotheses are about entrepreneurial success, and no question measures success in any form — not revenue, growth, survival or self-assessment. Question 4 measures enterprise size, which is not the same.
Whether a respondent was an entrepreneur
The method chapter states a criterion of a year's experience as an entrepreneur. Question 3 merges entrepreneur and employee experience, so the criterion is unverifiable from the data.
How many people were asked
No invitation count, no field dates, no timestamp column in the export, so no response rate can be computed from this project.
How strongly anyone held a choice
Questions 7 and 8 force one priority from five plausible items: no ranking, no weighting. The thesis concludes all five matter, which this design could not measure.
The reason behind any answer
There is no free-text question anywhere. Every finding is a count of choices between the researcher's own categories.
Reading your own instrument the same way
The supplied material prescribes no procedure of this kind; what follows is this library's synthesis. Run it on a draft: half of it stops being actionable once the first response arrives.
Six passes over a draft instrument
Print it as configured, not as drafted
Export the live form with its required flags and instructions. A draft cannot tell you what the platform will enforce, and that gap is where this project's divergence sits.
Write the sentence you intend to publish
Draft each results sentence in advance: "X of N respondents (Y%) said …". If you cannot write it from the option set in front of you, the question is not ready.
Test every band endpoint, and ask who has no honest answer
Name a respondent sitting on each boundary and ask which option they tick. If anyone has no true answer, add a route — "none of these", "don't know", "prefer not to say" — or accept that you are compelling a choice and say so.
Match every hypothesis to a question number, and reconcile every copy of the wording
A variable with no question number beside it will not be measured. Check the wording in the form, the working document, the column heading, the tabulation title and the chapter text: drift survives into the write-up.
Every defect above is a defect in the instrument, visible only because the document survived alongside its data — not a judgement about a researcher whose thesis passed examination. One instrument from one project is no evidence about instruments in general. The constructive counterpart is designing a data collection instrument.
What to carry forward
- An instrument is a set of decisions about what you will be able to say. Eight required single-choice questions bought a complete dataset and forfeited ranking, intensity and explanation.
- Boundary overlaps, non-monotonic scales and non-exhaustive option sets are visible before any data exist and cannot be repaired afterwards.
- Where a form and a questionnaire document disagree, the platform's configuration wins.
- Every variable named in a hypothesis needs a question number beside it. The variable this study was about had none.
- One instrument, one project, observed practice — not a standard and not a claim about instruments in general.
Frequently asked questions
Can I use this instrument as a template?
No. It is displayed because the library had no instrument at all, not because it is exemplary. Ten defects are demonstrable from the document alone, including overlapping bands on two questions, a scale that does not run in one direction, and no route through question 6 for a respondent unfamiliar with the models. Use it as a worked object to practise reading, and design your own option sets from your own analysis plan.
Why does it matter that all eight questions were compulsory?
It produced fifty complete responses with no missing cell, which makes every base identical and every percentage checkable. It also removed any way to decline a question, so a respondent whose gender was not one of the three offered, or who knew none of the five models, had to give an answer that was not true or abandon the survey. Both consequences follow from one setting.
Which version of the instrument is the real one?
The evidence points to the printed form. It is the version whose configuration the data reflect — all fifty responses answered the question the other version said could be skipped — and the workbook reproduces its wording rather than the document's. That inference is this library's, drawn from the required flag, the fifty answers and the workbook heading. The project's own documents never address the difference.
Does a non-monotonic scale mean the findings are wrong?
No. The counts are counts and they are correct. What the scale prevents is interpretation of the two middle categories, which cannot be ordered relative to each other and so cannot be pooled or read as a midpoint. The thesis aggregated only the top two options, which are unambiguously above the rest, and stated that it had done so.
Is it a problem that the instrument briefed respondents on the models before asking about them?
Briefing is a design choice with a cost, not an error: it made the question answerable without prior knowledge, and it fixed what respondents knew at the moment of answering. The defects worth recording are different — one definition was given with four of its five parts, and the definitions came from commercial web pages while the thesis applied a scholarly-source rule elsewhere in the same project.
What single check from this page is most worth adopting?
Match every variable in your hypotheses to a question number before you field anything. This study's hypotheses are about entrepreneurial success and no question measures success in any form. That absence is visible in one pass over a draft, and it is not recoverable once the data are in.
References and source attribution
- A complete research dataset supplied as student work: one examined master's thesis with its survey instrument as issued (a six-page image-based PDF, transcribed here) and as a working questionnaire document, its participant information sheet, its raw response file and its ten-sheet analysis workbook. Researcher, supervisor, institution, protocol number and form addresses scrubbed. Used as observed practice, not as a model answer.
- The five model definitions printed on the instrument, referenced there to commercial web publications dated 2022; publishers and URLs scrubbed, model names not reproduced under this library's naming rule.
- The supplied teaching source: weekly study notes, slide decks and assessment activities for a master's-level research methods subject in project management, including six weeks of data-collection material that names instrument types without displaying an instrument. Author, institution and year not stated in the supplied files.
Suggested questions for Ask KEVOS
- How do I decide response options for a survey question before I write the question?
- What can I analyse from single-choice questions, and what needs a different question type?
- How do I check that my bands and categories have no overlaps or gaps?
- What should I do about respondents who have no true answer to one of my questions?
- How do I make sure every variable in my hypotheses is measured by my instrument?
- What is the difference between a questionnaire document and the form as configured?