Section 3
Setting up and running the survey
Everything to settle before the first invitation goes out. Each rule below affects whether the results can be read reliably afterwards, so treat this as a checklist rather than background reading.
The two forms
Build two separate forms. They share a structure but not their wording.
N−3 managers taking part in the programme
Participant survey
- 7 core questions
- 3 extended questions, used only if you run Extended
- 1 conditional clarity follow-up
- Codes P1–P7 in the export
N−2 leaders who oversee or regularly work with the N−3 cohort
Stakeholder survey
- 7 core questions
- 3 extended questions, used only if you run Extended
- 1 conditional clarity follow-up
- Codes S1–S7 in the export
Choose Core or Extended before launch and use the same version for everyone in a group. The Extended version adds questions about the intended gain and the conditions that would help, plus one optional short example. Try both versions.
Who to invite
Send each survey to the right group
Send the participant survey to the N−3 cohort. Send the stakeholder survey to N−2 leaders who directly oversee, or regularly work with, members of that cohort. N−1 leaders are not respondents in this version.
One response per leader, not per employee
Ask each N−2 leader to submit one response about the N−3 participants they know sufficiently well. Do not request separate evaluations of named individuals. An N−2 response represents one leader's perspective, not a count of the employees they oversee.
Form configuration
Keep the codes in the export
Retain the question IDs and option codes in the export. Codes are for administration; the displayed option wording is what respondents read.
Question 2 shows the question 1 selections
For question 2, show only the substantive priorities selected in question 1 where the platform permits. Otherwise display the full list and instruct respondents to choose one of their earlier selections.
Show the selected priority above the linked questions
Where possible, display the selected question 2 priority above questions 3 and 7 and the extended questions. If this is unavailable, retain the reminder in the question wording.
Limits and exclusive options
Enforce the selection limits
Questions 3 and 4 allow all applicable substantive options. Options described as “select on its own” remain exclusive. Question 3 option X (working conditions) may be combined with personal development opportunities. For capped questions, Other uses one permitted selection. Selecting Other requires a short phrase of up to 20 words; no essay or justification is required.
Require a description only for Other
Reuse the question 1 description if Other is chosen again in question 2. None and Unsure require no written justification. E3 remains optional. Make structured questions required only with their uncertainty or None options available; P4 also has Prefer not to answer. Do not force a judgement from someone without sufficient information.
Branching
Skip the linked questions when question 1 is None or Unsure
If question 1 is N or U, omit questions 2, 3 and 7 and the extended questions. If branching is unavailable, instruct the respondent to leave them blank. Questions 4, 5 and 6 remain available because they address management responsibilities overall. A U in question 1 still triggers the clarity follow-up; a standalone N does not.
Show the clarity follow-up once
Show PC or SC once, after all other applicable questions, if any U was selected. Record the triggering question IDs. If conditional display is unavailable, show “Complete this only if you selected an Unsure or insufficient-information option earlier.” If question 6 is the only trigger, use the next-role stem provided. Do not trigger from N or R, or repeat for each U.
Privacy and consent
Say plainly who will see the responses
Before launch, tell respondents who will see identifiable responses, if identities are collected, and how findings will be shared. Do not describe the survey as anonymous unless the actual settings and reporting support that statement. No names or detailed demographic questions are required by this draft.
Keep the two groups separate and do not label people
Keep participant and stakeholder findings separate. Do not use responses to label individuals or make promotion decisions. The instrument identifies reported development needs, including aspirations for excellence; it does not establish demonstrated competence.
Language
Decide on English, Arabic or both
Confirm whether the client wants English, Arabic or both. Translate the meaning of the choices faithfully and have a bilingual reviewer check workplace terminology and whether the alternatives remain distinct. Retain the same codes across languages.
Before launch
Preview one submission of each survey
Preview one participant and one stakeholder submission to check selection limits, branching and the export. Keep the wording and option order fixed once the survey opens.
During analysis
Treat the submissions as a fixed dataset
Analyse submissions as a fixed dataset, including the predesigned conditional responses. No individual contact or follow-up interview is part of administration. Record later client explanations separately; they must not replace survey answers or change their counts.
The four answers that need special handling
These options appear across the survey. They are valid answers, not missing data, and each one is treated differently in the analysis.
| Code | Option | What to do with it | What it means |
|---|---|---|---|
| O | Other | Retain the selection and its required short description. Review it without changing the submitted category. If missing despite the form rule, flag as unspecified Other. | An alternative beyond the list. The description completes the answer; no interview is required. |
| N | None / no priority identified | Retain as a valid response using the exact question wording. Do not require a justification. None is not offered in question 3. | No relevant priority, pattern or gain has been identified at present. This does not establish mastery or that development is unnecessary. |
| U | Unsure / insufficient information | Retain Unsure. Use the predesigned clarity follow-up to record what could provide clarity. | The next development step, or evidence needed to identify it, is not clear. Do not impute a capability deficit. |
| R | Prefer not to answer | Retain separately wherever offered. Do not trigger clarification from R alone or request an explanation. | Non-disclosure is not uncertainty, lack of insight or absence of need. |
Branching, in plain terms
Two rules change what a respondent sees. Both are visible in the working survey.
If question 1 is None or Unsure
Questions 2, 3 and 7 and the extended questions are skipped, because they all depend on a chosen priority. Questions 4, 5 and 6 still appear: they ask about management responsibilities in general.
If Unsure is selected anywhere
One clarity follow-up appears at the end, once only, no matter how many questions triggered it. If question 6 was the only trigger, the follow-up uses the next-role wording instead. Record which questions triggered it.
Do not trigger the follow-up from Prefer not to answer
Before you send the invitation
A short final check. If any line is not yet true, do not launch.
- Both forms are built, and each group receives only its own form.
- Core or Extended is chosen, and it is the same for everyone in a group.
- Selection limits are enforced by the form, not by instructions alone.
- None, Unsure and Prefer not to answer clear the other selections automatically.
- Other requires a short description of no more than 20 words.
- Question 2 offers only the areas the respondent selected in question 1.
- The clarity follow-up appears once, and only after an Unsure answer.
- Question IDs and option codes are retained in the export.
- The confidentiality statement matches what you will actually do with the responses.
- The wording has been reviewed for clarity for readers whose first language is not English.