Few Employee Surveys Work

January 30, 2011

We have all been exposed to an employee satisfaction survey at some point in our working lives. For some of us, the idea of filling out yet another QWL (Quality of Work Life) survey is not nearly as appealing as having a root canal. I have witnessed a significant hit to morale in many groups as a result of these attempts by management to gather information. Let’s examine why employee surveys cause problems and suggest some antidotes that can make a big difference.

1. Questionable anonymity - Nearly all QWL surveys are advertised by managers as being anonymous. This is to encourage people to share information without fear of repercussion. Unfortunately, nearly all surveys these days are conducted electronically. Most people are aware that anything online can be traced, if a smart IT technician is assigned to the case. People simply do not believe the promise of anonymity, which lowers the validity of the data. One way to give slightly more confidence in the integrity of the process is to include a statement such as this at the end of the form, “You do not have to type your name on this survey because it is anonymous. We will pay no attention to who you are if you do not sign the document. However, some people do wish to be contacted personally to give further input that might not be adequately covered on the survey. For that reason, there is a line to give your name if you wish. Someone will get back to you.”

Of course, no statement by management is going to convince the die-hard skeptics, but this explanation will assist most people. Another action that can help is to have a shop floor person (perhaps a known skeptic) participate in the data analysis phase, so he or she can verify personally that the managers do not know who said what.

If you claim anonymity but are quietly keeping track on the side out of curiosity, then there is no hope for you, and you should get out of the leadership occupation immediately.

2. Poorly worded questions - many managers believe that putting together a survey is a simple matter of writing a few questions about how people feel. Actually, survey design is a rather complex and exacting science. There are numerous ways to present questions that will yield meaningless (or at least ambiguous) information. There are also some methods that will generally produce usable data. The cure for this problem is to have someone who is trained on survey questions actually construct the instrument. If you have not had at least one course in experimental research design, then it is best to leave the matter to someone who has.

3. Long and tedious survey - It is not uncommon for QWL surveys to contain over 100 detailed questions. It is amazing that the designers of these surveys do not realize the obvious fatigue factor involved in completing one of these burdensome questionnaires. A much more accurate reading can be obtained by keeping the number of questions to 20 or less. This can be accomplished by paying attention and only asking important questions. Leaving out the fluff can cut the time to take a survey by more than half.

4. Management interprets the information as they please - It is frustrating to witness how managers wave away complaints or gripes on surveys as simple whining. Or they might shrug their shoulders and say, “there is absolutely nothing we can do about this issue,” so they consider the input as moot. The antidote here is to not ignore input regardless of how painful it is or how frivolous it seems. All input needs to be considered valid and not assumed away with some convenient rationalization.

5. “Nothing ever changes” – This is a common theme on the shop floor. “We take these stupid surveys, but nothing ever changes.” The antidote to this habitual problem is to actually take concrete actions based on the survey, then (and this is the part most managers forget) advertise that the changes are being made as a result of the QWL survey. Rather than saying, “We are going to add a second brief afternoon break,” say “As a result of your input in the recent survey, we are changing the break rules to allow a short second break in the afternoon. As always, we appreciate your candid feedback.” If managers do not make a conscious effort to communicate that changes are the result of input, people will usually not make the connection. Once a change is made and it becomes habit, people forget that there was a change, so the perception of “nothing ever changes” is common.

6. Managers try to react but do the wrong thing - It is far better to let the shop floor people be involved in decisions of how to improve conditions based on survey results. It may take a little more time, but the quality of process changes will be far better if those impacted the most have a say in their invention. They will make the changes work rather than wonder and push back at the clueless inventions of upper management. It is better to have workers inside the tent piddling out, than outside the tent piddling in.

7. Managers reacting to the vocal minority rather than the silent majority - This problem is common when surveys give the opportunity for open-ended comments. People on the fringe can give strong input, and managers might mistakenly interpret this to be the will of the majority. The simple antidote to this problem is to verify that a strong message really does come from several individuals rather than one highly disgruntled outlier.

8. Survey not tested for validity - For a survey to be useful, it needs to measure the phenomena it purports to measure. There are statistical techniques for determining if an instrument has validity. You may not have the time or money to invest in a professional survey designer to test the validity of an instrument, but at least you should ask the question of whether you are actually getting valid information.

9. No thought to reliability - The reliability of a survey is different from validity. For a survey to be reliable, it should produce a similar result if repeated and there have been no changes in processes since the last survey was taken. If survey results are all over the map when nothing in the environment in changing, it is a sign that the instrument is not reliable (repeatable).

10. Poorly communicated - When surveys are sent out, the cover letter explaining the purpose and process is a critical document. Many managers have an administration person whip out a paragraph of “management speak” like this. “It is vital that we know what people in our operation think in order to continually improve working conditions. Please take the time to fill out this anonymous survey that will give us the information. Thank you.” Here is a different note where the manager took the time to set up the survey for success.

“We are going to re-do our strategic plan, and it is important to include your input before making changes. The attached survey will begin the process. Before you take this survey, please reflect on the following points:
• The survey really is anonymous – we will have shop floor people help with the analysis to verify no names are attached to the data.
• We will summarize the data for you as soon as it is received.
• We will use the information as the basis for a series of meetings (you are invited to participate) on how this business can be improved.
• We will be making changes based on the results of this survey.
• We are all part of making this organization a success.”

With an introduction like that, employees will know this survey will likely have some impact and their viewpoints matter. It is critical to not waste credibility, time, and energy on a poorly designed and administered QWL survey. If the above 10 points are considered when designing an employee survey, it will produce results that can be the basis of solid organizational progress.


Follow

Get every new post delivered to your Inbox.

Join 494 other followers