5 Questions to Ask Yourself Before Your Next Survey Research Project
It’s really easy to just dive into a survey research project. There’s nothing like getting something going quickly, something you have may had on your mind for a while. But, hold your horses! Our recommendation is to temper your enthusiasm, take a deep breath and ponder the following questions. When you have your vision of what you may want to do, or concrete answers, you’ll be much better off.
1. What are my objectives?
Readex Research has been doing this a long time and the best survey research outcomes are always based on well-considered, thoughtful objectives. In other words, what is it you want your project to help you accomplish?
A great way to think about it is this: What decisions might you want to be able to make when all is done? Or, what data do you need to help you or your executive team make decisions?
Objectives provide guiding lights to the entirety of the plan.
2. What will be the sample frame?
Well, let’s start be defining sample frame. This is the population from which a sample is drawn and this sample should include all of those in the population. For example, if you work within a larger organization and have employees scattered around the country, you might say your sample frame is all employees or a subset, say only those in a particular city or region.
What’s important to keep in mind is this. If your sample frame is anything other than all employees, it’s not necessarily the case that your sample will be representative of all employees. In other words, don’t assume that a survey of the folks in the Chicago office is representative of everyone else. You might feel you may infer this to be the case but it’s not. If you want to position results as representative of all employees, your sample frame should be all employees!
The good news is that with employee surveys, you usually know who they are and where they are, meaning they can all be enumerated and included in the survey.
3. Who will design the questions?
We hate to say it, but designing questions, while suggested by some is an easy process, isn’t. This has to be carefully considered because well-designed questions yield the most effective results. Sure, you can look things up on the internet or ask a pal. If in doubt about your skill level or the levels of those who you may wish to use, consult with a research pro. Even if you wish to conduct the survey on your own, which is fine, get guidance from an expert. Junk in is junk out.
4. What methodology will be used?
Because of the nature and desired outcome of employee surveys, which is the TRUTH, a self-administered survey will be your best bet.
Self-administered simply means the individual takes the survey on their own. So this means online or on paper. Generally speaking, face-to-face or telephone surveys are not optimal. Why? You can probably guess. It’s much more difficult for an individual to be honest, candid and forthcoming with another live person versus an anonymous, safer channel like online or mail. While online these days is more popular (faster and also cheaper) using a mixed approach is not unusual. The mixed approach is used when the company can’t reach all employees with an email, for example when you have employees working on a production floor and they don’t have a company email, or perhaps not even computer access. It’s a bit trickier this way, but good organization can actually yield great results with a survey done using mail all by itself or mixed with online.
5. What do I plan to do with the data?
So the field work is closed and now you have the information, the data. Hopefully you will have determined in advance what you want to do with it having answered these questions.
Who will see all of the results? Who will see partial results, for example by office location? Who will have access to all of the data and be able to manipulate data for the purpose of running more analysis? How will the data be made available, for example, SPSS file or Excel? Will you need to create PowerPoint slides and who will do that?
These questions are not all-inclusive, surely there are more. Here’s the point: Have a plan! And, one final caution. If you are going to run an analysis by job titles or other segments, make sure you have enough responses in each of those segments so that respondent ID remains anonymous. In other words, if you have only one person in Accounts Payable and you run a cut by job function, then the identity of that respondent can easily be revealed.