What You Should Know about Using Surveys
Aiman-Smith, Lynda, Markham, Stephen K., Research-Technology Management
Surveys are commonplace in organizations, with more than 70 percent of U.S. organizations using them regularly (1). R&D managers can use surveys to understand the perceptions, thoughts, feelings, and behavior of scientists, project managers, teams, customers, suppliers, and others. Information from some surveys can be used as a measure of employee attitudes and morale. Other surveys can provide baseline measures for benchmarking of processes and behaviors. Still others can provide information for identifying organizational gaps and for driving organizational change (2).
The quality of the information you will receive from a survey depends on the quality of the questions you ask. It may seem simple and easy to develop interesting questions, send them to people and ask them to answer; but developing a good survey is a complex effort and requires a great deal of attention, thought and care (2,3). This article offers advice on how to develop good surveys, as well as some tips on administering them and analyzing the information you will get.
There are three important guidelines you should keep in mind with any organizational research:
* First, confidentiality should be scrupulously maintained. Even though it might seem useful to know exactly who responds to surveys, we recommend collecting anonymous data so employees know their input cannot be traced.
* Second, share the results regardless of whether they seem favorable or critical. If your survey has been constructed with care, all information you get will be useful.
* Third, if issues arise from surveys, you should have a commitment to respond to them. A well-constructed survey, with an appropriate sample response, will provide useful information that should be a base for action.
Here now are out six key steps:
1. Decide Precisely What You Want to Measure and Whom You Want to Survey
By definition, a survey is a systematic process of data collection, intended to quantitatively measure specific aspects of organizational concepts. Because a survey is often the result of some burning issues in an organization, those burning issues must be clearly articulated.
Different concepts can be assessed in a survey, for example: Do you wish to know whether employees feel empowered in their jobs? Are you interested in what motivates your scientists? Do you want to assess the current state of commitment to the organization? Do you care how employees see the organization's culture?
One way to identify and clarify concepts is through discussions with stakeholders, including employees, professionals and focus groups. After you have identified concepts, a thorough review of the literature of previous research on those concepts and associated areas is vital to developing clear definitions.
Different people represent different viewpoints. Give thought to your intended respondents and make sure your questions are targeted specifically to those people. Questions to consider: Do you want responses from people in different functional areas? Different geographical areas? Different hierarchical levels?
2. Write Survey Questions
The quality of the survey questions affects the quality and quantity of responses. In this step, you generate questions or survey items that reflect the concepts you wish to measure. The items should be specific, clear, concise, free of jargon, and appropriate for the respondents to whom they will be administered (3, 4). This step is usually the most complicated and labor-intensive phase of survey development.
Where to start
You can use focus groups representing the target sample, or area experts, to help generate survey items. Explaining the construct and then asking different people to articulate what kinds of questions they would ask, should lead to 5-10 statements about each construct (4). You can ask questions like, "Have you felt empowered in your organizations? …