This is just one possible workflow to consider. Follow the tracks and click the stations for context and sources of help.
What sort of data are you collecting? This will determine your approach.
How much data are you collecting? Are there access, security or backup implications to consider?
Are you storing personal or sensitive data?
What tools meet your data needs?
A survey needs to flow logically. If you have branching in your survey, is everyone seeing the questions they need to see?
Test your survey thoroughly before making it live. Are the results coming through as you expected?
How are you going to distribute your survey? Who is allowed to fill it in? Are you accepting multiple responses? Can you use a mailing list? Do you need to advertise for participants?
Will you send reminders? How many responses do you need?
Where do you need to be searching to find the information you need?
Evaluate what you find to determine what is worth using.
How you collect your data depends upon what data you're collecting. Is it a survey, pre-existing statistics, literature,
experiment...? Whatever you're collecting, you inevitably need to collect it!
You may need to transfer your data from one tool to another.
You'll likely need to clean up and tidy your data for it to be useful.
It's time to start organising what you've got. This is especially important with text analysis, but whatever data you've got, you need to manage it in some way, shape or form.
In order to analyse your data, you must first understand it. How else will you know whether or not you're analysing it properly? A small error in a formula, for instance, can cause an awful lot of damage
should it go unchecked!
What are the best methods for working with your data? What calculations do you need to apply? What criteria do you have to work with?
Once you're on top of your data, it's time to do some actual analysis!
You want data you can analyse, and therefore you want it to be consistent. Use validation to limit the choices people can make in your survey. But don't limit people too far that the responses lose their meaning!
How you ask a question will affect the response you receive. Make sure that you're asking what you think you're asking.
Don't be afraid to ask other people. It's a survey after all!
What methods will you be using to crunch your responses? Because that will inform how you design your questions...
What information do you actually need? Don't waste people's time on redundant questions.
What are you trying to achieve? Get a clear understanding of this before you start. It will inform your choice of tools, your question design, and your analysis.