Creating Surveys

Surveys are a popular instrument for collecting information in many Management Analysis and Development projects. A major advantage is their cost-effectiveness: Surveys can collect representative information about a large population through a small sample of the population's members. Surveys can collect information that might not be available elsewhere. Surveys also can confirm or complement data collected through interviews and focus groups.

Management Analysis and Development consultants can assist you through all steps of the survey process. We work with clients to develop surveys that will meet their needs. We administer surveys by hosting them on our survey web site, subcontracting telephone surveys with trusted, experienced vendors, or distributing paper questionnaires. We receive and tabulate the results, analyze the data, and report the findings (either in summary form or as raw data, depending on the client's preference). Our role in the survey process ensures that we deliver a high-quality survey to our clients. We also act as a neutral third party so respondents know their identities will not be revealed.

Uses

Survey results can be used for making decisions about changes to internal operations, services, and policies. Surveys can collect information on demographics; attitudes, opinions, and values; past behaviors and experiences; intended behaviors; demand for services and willingness to pay; and evaluation of services and programs.

Specific survey uses include: service assessment; measuring customer satisfaction; collecting employee opinions; forecasting and trends analysis; marketing; and benchmarking.
Characteristics
Most surveys have three general characteristics:
1. Random sampling of a population to produce a representative group to take the survey
2. Standardized questions and procedures to increase reliability of the information
3. Quantitative analysis of the responses
Determining how to ask questions: web, mail, or phone

Web surveys, telephone surveys, and mail surveys have their unique advantages and disadvantages. For example, telephone surveys are expensive but yield richer data because clarifying questions can be asked. Telephone surveys also generally yield higher response rates than mailed surveys. Web surveys have almost completely replaced paper surveys, but paper ones are needed in some situations.
Planning and implementation

Survey planning is important so that the questions accurately measure and collect the information you want and so that the respondents truly represent the population of interest. Steps must also be taken to reduce the many types of bias that can arise. Important steps in creating and implementing a good survey are knowing when to sample a larger population, writing good survey questions, and achieving a good response rate.
Sampling from a population

"Sampling" means gathering data from a representative subgroup of a larger population. One of the great advantages of sampling is that it enables you to make valid assumptions about the attitudes and opinions of a large population without having to spend resources on interviewing or surveying each member of that population. The key to sampling is making sure that your subgroup is indeed representative of the larger group. Sampling is not difficult but must be done correctly in order to yield valid data.
Writing good survey questions
First, a Management Analysis and Development consultant will help you clearly outline the information you need to collect. Don't worry about "how" before defining "what." We will help you decide:
- What information is necessary and help you separate "need to know" from "nice to know" information
- What you will do with the information, who will see it, and how much detail they will want
- Who will receive the survey-that is, define the population
- Once we have defined your data and analysis needs, we will help you write the questions. Every question on a form should have a purpose. If you cannot clearly identify how the information will be used, don't bother the respondent with the request. Keep things simple, concise, and clear.
Caveats for interpreting surveys

Survey interpretation is affected by response rates, how accurately the questions measure what is desired, the questions' specificity and construction, and the sampling method. The more carefully constructed and administered the survey is, the more confidence you can have in interpreting the results.

As all the above information shows, it is not easy to design a good survey. Management Analysis and Development consultants have conducted many types of surveys and can make sure your survey results are of the highest quality.

Excerpted from internal survey guides written by management consultants Tom Miller and Donna Koren.