Surveys

The Data Gathering techniques of surveys include
a. individual self-administered questionnaires
b. group self-administered questionnaires
c. face-to-face interviews
d. telephone surveys

A Sample Survey surveys only a selected, representative sample of the entire population of possible respondents

An example of a Sample Survey is the Gallop Poll, who generally surveys about 1,100 people to determine the opinion of the general population of the US, which is 285 mm, on a given topic

See Also: Sampling

A Population Survey surveys the entire population of possible respondents

An example of a Population Survey is when the administration of UVaWise surveys the entire student population of UVaWise

A Cross-Sectional Survey is administered once to a sample of respondents

With a Cross-Sectional Survey, the same questionnaire may be administered again to a different sample of respondents

The repeated use of cross sections is for detecting trends

The US Census is an example of a Cross-Sectional Survey in that they typically ask the same or similar questions over a period of years or decades

A Longitudinal Survey, which is also known as a Panel Study, the same or similar survey is given repeatedly over a period of time to the same or similar group of respondents

An Establishment Survey is a survey that has, in effect, been institutionalized because of its widely recognized validity, reliability, & generalizability

An Example of an Establishment Survey is the National Organization's Survey which asks a series of standard questions about the characteristics of companies & their employees

An Example of an Establishment Survey is the Minnesota Multiphase Personality Inventory (MMPI) which asks a series of questions to quantify a person's personality

Survey Questions:

There are TWO kinds of Survey Questions:

a. Fixed-response questions are like multiple-choice exam questions

b. Open ended questions: respondent states or writes an answer to the question in their own words. (Similar to an interview)

Avoid loaded questions in surveys.

Bias is the effect you get when you have a loaded question

In closed ended questions many formats are used, but try & give a clear range of choices ( Likert Scale ) on a 5 or 7 pt scale

Example of a Likert scale:

Please mark the answer that most closely matches your belief:

I study very hard

Strongly Neutral / Strongly

Agree Agree No Opinion Disagree Disagree

______ ______ ______ ______ ______

There are FOUR Types of Questions typically used in Surveys that may either be Fact Based (Objective) or Belief Based (Subjective)
a. Background (demographics)
b. Activities
c. Knowledge
d. Sentiments (opinions, values, attitudes, feelings, etc.)

In general, the survey should try & use Fact Based, Objective type questions wherever possible, even on topics that may be subjective in nature

Example, in asking people about family values, the survey may ask a question such as: "How many times does your family have meals (include all meals, morning, noon & night) together per week?"

A Master Survey is a special survey in which only the researcher sees which indicates the function of each question

The Master Survey should label each question w/ respect to the:

- Four Types of Questions

- Thesis

- Research Objectives

- Variables

-Variables w/ respect to Independent, Dependent, & Control Variables

The Master Survey should group the questions together by Independent, Dependent Variables, & Control Variables

Make each question count
Ask, "Is this important to my thesis or hypothesis?"
Try to avoid sentiments
Use short statements

Question ordering
Start w/ safe questions to try to gain trust, build rapport
Put most controversial questions near end
End w/ safe questions

The Advantages & Disadvantages of Survey:

Compared to an Ethnography or a Case Study, surveys are more easily generalized to the population they were designed to represent

By directly questioning respondents, a survey can measure subjective indicators such as job satisfaction

Changes in facts & attitudes an be traced & studied if the same question is asked in repeated surveys

One problem w/ surveys is Selection Bias in which only certain types of people respond to a survey

Selection Bias can be minimized through strong Sampling Procedures