Survey design and writing
A survey is the cheapest way to hear from a lot of people, and the easiest way to collect a pile of numbers nobody trusts. The difference is in the writing and the build. Between us we have written, built and fielded surveys on paper, online and on tablets for as long as we have been evaluators, and we still argue about question wording. That argument is the service.
What does Plotline Evaluation’s survey service include?
The whole run, under one quote: writing the questions, building the instrument on paper or online, fielding it, collecting the data, analyzing it and delivering the findings. The technical work is included: branching logic, piping, validation, calculated fields, scoring, randomization and survey queues. Every survey is written for the people who will answer it and built to your specifications.
What we write and build
- Paper surveys laid out so people can fill them in and your staff can key them in without mistakes, with a scanning or data-entry plan.
- Online surveys for desktop, mobile, tablet or kiosk, reached by link or QR code, on any survey platform you prefer, or built by us as a custom survey with a personal link for each respondent and secure data collection.
- Mixed mode: the same instrument on paper and online, so every answer lands in one dataset.
- Standardized instruments, scored the way their authors intended, alongside the questions only you need.
- Every kind of survey a program runs: needs assessments, client and patient experience, staff and climate surveys, pre, post and follow-up surveys, community surveys, event and training feedback.
- A survey you already have that is not working. We mark it up and rebuild what needs rebuilding.
All of it to your specifications. You tell us whom you need to hear from, what you need to decide, and what the funder or the board expects to see. We design to that, not to a template.
Written for the people who will answer it
A survey written for the wrong audience gets polite nonsense back. We write for the population in front of us: the reading level, the language, the examples, the response options and the way the questions are ordered all change with who is answering. Between us we have surveyed Indigenous and tribal communities, veterans and military families, first responders, people who are incarcerated and people returning home from incarceration, people with low incomes, communities of color, immigrant and refugee communities, people with disabilities, older adults, young people, patients and clients, whole communities, the staff of nonprofits and public agencies, and the employees and customers of companies. If your audience is not on that list, it is only because the list had to end somewhere. Cultural review is built into the design, not bolted on at the end.
- Questions people can answer: one thing at a time, in plain language, with response options that cover every case and overlap nowhere.
- Length people will tolerate. Every question earns its place against the analysis plan, which is written before launch.
- Tested before it goes out, with a handful of real respondents, and fixed before anyone else sees it.
- Paper where paper works. Some populations, and some rooms, still answer more honestly on a page than on a screen.
The technical side
- Branching and skip logic, so nobody sees a question that does not apply to them.
- Piping, so the survey remembers what people already told it.
- Validation and ranges, so a birth year cannot be 2091.
- Calculated fields, scoring and randomization of items or blocks.
- Survey queues, scheduled invitations and reminders, and longitudinal designs with events and repeating instruments.
- Embedded consent, anonymous or identified modes, and bot and duplicate screening on public links.
- Coding of open-ended answers, a codebook and data dictionary, and exports that open clean in your analysis software.
If you can describe it, we have very likely built it. If we have not, we will say so before you pay for it.
One quote, start to finish
Design, build, deployment, data collection, analysis and the deliverables are quoted together, so there is no second conversation about who analyzes the results. The people who wrote the questions analyze the answers, which is how it should be.
- A scope call, free of charge: whom you need to hear from, what you need to decide, by when, and on paper, online or both.
- A written scope with the deliverables and dates, and one quote for the whole run.
- Draft instrument and analysis plan for your review, then a pretest with real respondents and the fixes that follow.
- Build, launch, reminders and follow-up, with a response target agreed up front.
- Analysis, then the deliverables: the dataset and codebook, a summary written for the people who have to act on it, and the charts, dashboard or report you asked for. See sample deliverables.
Surveys are one of our services, and often part of a larger evaluation. Many of ours run in REDCap.
Who does the work
John and Scott, personally, with more than twenty years of combined experience in surveys. A medical anthropologist with a PhD and a qualitative evaluator and data scientist, both certified by the CITI Program in human subjects research; the rest of the background is on the About page. We write the questions ourselves; nothing is handed to someone who has never met your program.
Questions people ask
Paper or online?
Whichever your respondents will actually complete. Often both, built as one instrument so the data lands in one place. Paper still wins with some populations and in some settings, and we say so when it does.
Can you fix a survey we already wrote?
Yes. The usual problems are a question that asks two things at once, a scale that changes direction halfway through, or branching logic that strands people. We review it, mark it up and rebuild what needs rebuilding.
Do you analyze the results too?
Yes, and we would rather. The analysis plan is written before the survey launches, so every question has a job. You get the dataset, a codebook and a summary written for the people who have to act on it, all under the same quote.
Can you survey a population or culture you have not worked with?
Usually, with the right review. We adapt wording, examples and response options for the audience and pretest with people from that population before launch. Where a community has its own review process, we work through it.
Which platforms do you build in?
Any of them. If your organization has a preference or already pays for a platform, we work in it. If not, we build a custom survey of our own, with a personal link for each respondent and secure data collection. Paper is always an option.
How long does a survey project take?
It depends on the length of the instrument, the sample and how fast your people can review drafts. The written scope gives dates, and a real deadline can be met when we know about it early.
How is it priced?
One written quote for the whole run: design, build, deployment, collection, analysis and deliverables. The scope call is free.
Tell us about your survey
Whom you need to hear from, roughly how many, by when, and whether it is paper, online or both. You will hear back within one business day.
Every inquiry is confidential. We never sell or share your information. Privacy policy.