The Purpose of Employee Surveys

"Gain Insight About Your Organization "

Changing policies at your workplace? Trying to find out why your turnover is so high? Hoping to gain more insight into what your employees would like in terms of benefits? Good idea! Planning on taking the easy way out, by just talking to a few key members of your workforce? Not one of your better ideas! If you want to know what your employees think, you need to ask them. All of them.

"There is one main difference between using a survey or basing your decisions on what a focus group or a couple of people have told you," says Paul M. Connolly, president of Performance Programs Inc. (PPI), a consulting group in Old Saybrook, Connecticut. "That difference has to do with getting honest answers. If a survey is done properly, it gives some anonymity to the people answering the survey so they are positioned to answer as honestly as possible. "When you actually talk face-to-face with employees, you have a relationship with them, and they may feel that they need to tweak their answer one way or the other," Connolly says. "They may believe they need to give you the 'right' answer, perhaps out of fear or out of the desire to look like a team player. Their reasons aren't necessarily dishonest; it's just a dynamic that interferes with getting a clean answer.

It Matters How You Ask

Surveys of the entire workforce provide the most accurate indicator of employee attitudes. Of course, having an outside vendor conduct a survey can be costly. For that reason, many employers choose to administer surveys on their own. Unfortunately, employers are unaware of the pitfalls they face when trying to self-conduct an employee survey. Some of the most common mistakes, according to Connolly, are:

• A tendency to "over question" one topic. This can give the employee an inaccurate perception of that topic's importance, relative to the others.
• Loading too many components into one question. Long, compound questions frustrate the employee and make it hard for the employer to be sure what part of the question the employee was answering.
• Filling questions with jargon. This is acceptable only if the employer is absolutely sure that everybody in the organization understands that jargon. But, Connolly notes, you would be surprised at how often that is not the case.
• Asking questions that cannot be addressed. For example, an employer might want to know, "How satisfied are you with your salary?" But what if the employee responds: "I'm not satisfied"? They've answered your question, but you have no idea of the reasons for that dissatisfaction, or if it is justified. There is little, if anything, that you can do with the answer.
• Asking questions about issues you don't want to address. Employers run a danger when they ask about issues they do not intend to address. The salary question is a good example. If you ask employees whether they are happy with their salaries, you may create an expectation that you will make changes based on the results of the survey. This can lead to increased dissatisfaction if, after the survey, no changes are made.

We've Got Our Answers, Now What Do They Mean?

Connolly also points out that organizations need to have a good understanding of how to read survey results. Even the best survey will not be helpful to your organization unless you know what the data are telling you. For example, what if your survey asks, "Are your coworkers friendly and supportive?" and you receive a favorable response from 70 percent of your workers, while the question "Do you think your salary is adequate for the work you do?" receives only 30 percent to 40 percent favorable answers. Many organizations faced with that result might assume that workplace cooperation is not an issue, but salary adequacy is. However, if the organization were to compare their results with those of other organizations, they would likely find that their 30 percent to 40 percent favorable rate on the salary question was completely normal, while the 70 percent favorable rate on the coworker support question was below the norm. Self-conducted surveys, therefore, can often be misleading because they lack the normative data comparison that is essential to an accurate understanding of survey results.

Getting Help

As demonstrated above, a self-conducted survey, while inexpensive, is unlikely to get you the information you need. Surveys conducted by outside consultants can avoid the pitfalls discussed, but the costs are significant. So what's a small to midsize employer with a limited budget to do? Self-service, partially customized surveys may be the answer. These are standardized survey products drafted by experts, often with a limited amount of customization capability. Because the entire survey is not drafted "from scratch" for your workplace, these products are significantly less expensive than custom surveys. Typically, these products can be provided to you in hard-copy form, over the Internet, or through a dial-in, voice-response system. The survey is distributed through one of these methods to your employees, who complete the survey and return their results directly to the survey provider. This affords the employees the anonymity and security they need in order to feel free to give honest answers. Once the results are in, the survey provider analyzes the data, compares it with normative data, and returns a report to you. Connolly suggests that whether you decide to use a fully customized custom survey or a survey product, you should first ask the provider the following questions:

• How much experience does your organization have in conducting surveys?
• Have you conducted many for organizations in our industry?
• What methods will you use for surveying our employees: phone, paper, Internet, e-mail?
• How often will we need to survey our employees? This question is important, because the whole purpose of having a survey is to make the need for one go away. So, while you should have some kind of follow-up survey to see if things have changed, you don't want to constantly be surveying your employees.
• Can your research tell what the changes between surveys mean? If we see a three-point difference, can you detect whether that is due to changes we have implemented or is it just the amount of movement that would normally be expected?

Finally, regardless of how you choose to conduct your survey, remember that it's important that afterwards, you act on the results. Your employees will expect some reaction from you, and if they don't perceive one, may reach the conclusion that their opinions don't matter after all. Even if you cannot justify big changes, find a way to convey to the employees that when they speak, you listen.

This article was written by Nancy Hatch Woodward, a contributing editor to Business & Legal Report's Best Practices in Human Resources and Best Practices in Compensation & Benefits. The interview with Paul M. Connolly, Ph.D., appeared in the April 2001 edition of "Best Practices in HR," published by Business & Legal Reports, Inc. Dr. Connolly is president of PPI.