An alternative to magic jumping statistics
So you’ve written a great questionnaire, you’ve sent it out to hundreds of people, and you’ve got a big pile of completed questionnaires sitting on your desk. Hurray!
But what are you going to do with them now? Have you thought about that?
All too often people forget that *something* needs to come after the data collection stage of a survey. The stats don’t magically jump out of the pile of papers and present themselves to your stakeholders.
If you are going to do a survey yourself can make your life so much easier by thinking about the end of the project at the beginning of the project, by building ‘what next’ into your questionnaire and project plan.
Think about data entry as you write your questionnaire
- Will you need to do data entry? If you will be printing off hard copy questionnaires (either for postal, telephone or face-to-face surveys) you will end up with a pile of surveys that you will need to do something with.
- What data entry system will you use? Will you type responses into Excel, or Access, or SPSS? Do you know how to set this up?
- How long will it take to enter all the data? Say you have 100 completed questionnaires, and it takes ten minutes to enter each one, that is 1,000 minutes / 16 hours / 2 working days.
- Have you got the skills, time and inclination to do this? Do you know how to do data entry? Can you spare the time to do it? Can you be bothered to do it?
Think about analysis as you write your questionnaire
- What key statistics do you want to get out of the survey?
- Will your questions allow you to get the key statistics you need through analysis? Have you presented the questions in a way that will help you get the numbers you need? Do you want to get percentages, sums or mean scores?
- Do you know how to get the calculations you need through analysis? Will the data entry system you used enable you to calculate the right percentages, sums or mean scores?
- Have you got the skills, time and inclination to do this? Do you know how to do data analysis? Can you spare the time to do it? Can you be bothered to do it?
Think about reporting as you write your questionnaire
- What do you want your report to look like? Who is it for? How much detail will they need? Will they want graphs or tables or commentary? How might you meaningfully present the results from each of the questions?
- Will your questions / data entry / analysis allow you to get the data to make the graphs and tables you need to do this?
- Think about every question you write, and what you plan to do with it.
- Estimate data entry, analysis and reporting time at the start of the project, and build this into your plan.
- Work out what systems you will use, who will set them up, and who will actually do the data entry, analysis and reporting.
- Consider doing a web survey, so that respondents do their own data entry!
- Consider using Surveymonkey (or similar web survey software!) as a data entry system – you set your questionnaire up like a web survey, but rather than emailing it out you just use it for data entry. It is quick and easy to set up, it is so much more user-friendly for entering data than Excel or Access because it looks like a questionnaire, and it does all the adding up for you so that you can download it in Excel, SPSS or graphic format.