Survey

toc =Definition= According to Dictionary.com, there are over 12 definitions of survey. It is both a verb (to view) and a noun (a survey). For the purpose of this course, we shall define //survey// as "a sampling, or partial collection, of facts, figures, or opinions taken and used to approximate or indicate what a complete collection and analysis might reveal." (www.dictionary.com) One of the advantageous of a survey over experiment is that surveys can use a multivariate analysis, meaning that they can gather information about a number of variables at the same time and explore how combinations of variables influence the issue of interest (Hale, 1995).

=Purpose= A survey be both a quantitative or qualitative method of collecting information from a population. It can be quantitative by using force-choiced questions or qualitative by using open-ended question. The information collected varies from opinions to facts to suggestions. A model survey is constructed with minimal bias and objective questioning. Each respondant should be presented with the same questions.

=**Factors Affecting Survey Method Choice**= There are many methods that can be used to conduct a survey, such as telephone surveys, telephone interviews, mail surveys, email surveys, web-page surveys, and so on. Choosing a survey will depend on a variety of factors. They include the following (Creative Research Systems):

Speed refers to how quickly the company can get results. The fastest method would be email and web page surveys. The second fasted method would by telephone interviewing. Generally, faster is better. However, it is also important not to sacrifice the quality of the results for speed.
 * Speed**

Cost refers to how much money it will take to obtain results. The most inexpensive method would be email and web page surveys, especially if the company needs to collect large samples.
 * Cost**

Literacy levels refer to the literacy of the survey-takers. People who are illiterate and less educated do not usually respond to mail surveys.
 * Literacy Levels**

Sensitive questions refer to questions that ask for delicate information. For instance, a question might ask, “How many times did you confuse the front with the back?” Some people may lie, especially if they are embarrassed by the truth. In this case, they may be embarrassed to admit that they were often confused. However, if people are in front of a computer, they are more likely to respond truthfully. This is because the computer gives a sense of anonymity.
 * Sensitive Questions**

=Question Design= It is essential to design survey questions carefully. Poorly designed questions will lead to meaningless results. The following are suggestions by Babbie:
 * **Clarity:** Ensure that your questions are clear. Do not presume that your survey-taker understands any special terms you use. If you are unsure, explain the term to be safe.
 * **One at a time:** Ask one question at a time. This way, the person will know precisely what question to answer.
 * **Be reasonable:** Avoid asking questions that the user will not be able to answer accurately or at least give the option to skip questions.
 * **Relevance:** Ask questions that are relevant to what you want to discover. Do not ask for unnecessary information. For example, asking for their birthday when it is irrelevant.
 * **Short:** Keep your questions straight-forward and short. If they are short, they can be quickly read, understood, and answered.
 * **Neutrality:** Watch how you word your questions. Avoid biased words and wording, as it may have an effect on their answers.

The following are hints provided by Busha and Harter:
 * Avoid slang, jargon and technical terms. The only exception would be when the survey innately needs to use them.
 * Use consistent response methods whenever possible.
 * Make questions as formal as possible.
 * Start off with general questions in the beginning, and slowly start to ask specific questions.
 * If using closed-ending questions, try to develop a thorough list of possible responses.
 * Group questions with similar content together.
 * When defining unique or unusual terms, use clear, straight-forward language.
 * Use an appropriate display when presenting your questionnaire.

Thus, designing good survey questions may be a bit difficult. Once they are designed, it is also important to test them to ensure that they are effective. This can usually be done by using a small focus group.

=Sample= An ideal survey would take into account each member of a population but in reality, this is very time consuming and expensive. By using a sample, a small proportion of carefully selected people from a wide population (Hale, 1995), researchers can more effective conduct their work. The sample should represent the range of experience and characteristics in the population (Hale, 1995). Random selection is one way of creating a sample. Here, each member of a population has an equal chance of being selected regardless of availability, status or relationship to researcher.

**Biased Samples**
A biased sample is a sample that is non-representative of the population. Although removing bias is nearly impossible, recognizing it can at least reject some of the answers.

**Examples**
If the sample consisted of only existing customers, the sample would be bias because the existing customers would give favorable responses (Creative Research Systems). For example, if a question was, “how would you rate the ease of use of our on a scale from 1 (lowest) to 5 (highest)?” then existing customers would probably rate it highly. After all, the customers would have switched to a different company if they were unhappy with services.
 * Existing Customers**

In contrast to existing customers, a sample consisting of ex-customers would be bias towards giving unfavorable responses (Creative Research Systems). It should be noted, however, that it is important to know why they discontinued using your product or services.
 * Ex-customers**

“Phone in” samples consist of people who call in to give their opinions on something. This is a biased sample because those who call in often have extreme views on the subject – they will either be passionately for or against it (Creative Research Systems).
 * “Phone in”**

A daytime sample consists of people who are called during the day, usually to perform a phone survey. This is a biased sample because a majority of people who respond to daytime phone surveys do not work (Creative Research Systems). Thus, their opinions will not reflect the working population.
 * Daytime**

An Internet sample consists only of people who have Internet access. This becomes a problem because not everybody has Internet access (Creative Research Systems). For example, if there was a survey about redesigning a chair, and it was only available online, then those who do not have net access would be underrepresented. Moreover, since the object is something as typical as a chair, their valuable opinions would not be voiced. Unless the subject needing opinions is Internet-based (such as web-based product), then the sample should consist of more than just Internet views.
 * Internet**

=General Disadvantages= For research involving a large population, it is difficult to have a representative sample. Not all members of a population may be registered and not everyone wants to participate in a survey. Mailed questionnaires may be useful for collecting sensitive or embarassing data and are also quick and relatively inexpensive (Hale, 1995). But written questionnaires exclude the illiterate members, some less-educated members and those with a different first language. The questions must also be specific and not bias yet accomadate the varied responses (Hale, 1995). The researcher must keep in mind the context of the questions and be able to interpret a respondant's answer correctly. Surveys can provide valuable data but they must be admistered effectively.

=**Confidentiality and Integrity**= When filling out a survey, one natural concern would be whether the data supplied will be held confidential. This is important to both suppliers and survey distributors. Thus, there are safeguards in some countries that protect data collected. For example, in the United States, the U.S. Bureau of the Census protected the data they collect by law. In Canada, the Statistics Act ensures that data collected by Statistics Canada is kept confidential (Scheuren).

Some professional organizations have codes of ethics (such as the American Statistical Association) that deal with keeping surveys responses private. The following list illustrates recommended policies regarding the safeguarding of surveys:


 * "Using only number codes to link the respondent to a questionnaire and storing the name-to-code linkage information separately from the questionnaires
 * Refusing to give the names and addresses of survey respondents to anyone outside the survey organization, including clients
 * Destroying questionnaires and identifying information about respondents after the responses have been entered into the computer
 * Omitting the names and addresses of survey respondents from computer files used for analysis
 * Presenting statistical tabulations by broad enough categories so that individual respondents cannot be singled out" (Scheuren)

=Survey Tools= There are many online tools to create, publish and analyze surveys, such as: http://www.zoomerang.com/ http://www.surveymonkey.com/ http://www.questionpro.com/?siteRef=135000 http://www.websurveyor.com/gateway.asp http://www.confirmit.com/

=References= Babbie, Earl R. //Survey Research Methods//. Belmont, CA: Wadsworth Pub. Co., 1973.

Busha, Charles H., and Stephen P. Harter. //Research Methods in Librarianship: Techniques and Interpretation//. Orlando, FL: Academic Press, Inc., 1980.

Hale, Syliva M., //Controversies in Sociology: A Canadian Introduction (2nd Edition).// Toronto: Copp Clark, 1995.

Scheuren, Fritz. What is a Survey. //American Statistical Association//. 2004. 30 Nov 2006 <[|http://www.whatisasurvey.info/>.]

Survey Design. //Creative Research Systems//. 2005. 28 Nov 2006 <[|http://www.surveysystem.com/sdesign.htm>.]

"Survey." Dictionary.com. 2006. 2 December 2006 <[|http://dictionary.reference.com/browse/survey>.]

//Survey Methods.// 1999. 29 Nov 2006 <[|http://www.ischool.utexas.edu/~palmquis/courses/survey.html>.]