The most frequently used tool for field research in psychology.

Used for serious scientific inquiry.

Used to market ideas that appear scientific (marketing and politics).

Can indicate people's attitudes, behaviors, beliefs, knowledge, opinions, and personality.

Can use single item, ad hoc measures or well validated standard psychological scales.

Steps Involved In Surveying

1. Define the question/s: What do you want to know?
2. Define the population: What group do you want to know about?
3. Decide if you wish to use an interview or questionnaire.
4. Develop the survey instrument.
5. Identify the sampling frame, or group of potential respondents.
6. Choose a sampling method.
7. Conduct the survey.
8. Analyze and interpret results.
9. Disseminate results.

Question Design

Design of survey questions is an art.

Good questions will assess what is intended without introducing bias.

Clear and unambiguous

Concrete

Does not bias the respondent

Does not confuse the respondent

Respondent is able to answer accurately and honestly

1. Only ask about things a respondent knows.

2. Ask only one thing at a time.

3. Avoid priming the respondent?

4. Avoid jargon

5. Avoid biased words and phrases?

6. Be concrete.

Probability Sample: Each individual has a specifiable probability of being sampled.

Allows conclusions that generalize well to the underlying population

Random Sample: Each individual has an equal chance of being sampled.

Stratified Sample: Population is divided into two or more subgroups, and samples are taken from each.

Efficient in that it allows more representative samples with smaller numbers.

Cluster Sample: Multistage approach in which we identify existing groupings of people (cities), and first randomly select cities, and then people nested in cities.

Efficient when we have easier access to groups than individuals, e.g., school classes.

Nonprobability Sample: Select respondents from a limited group rather than the entire population.

Accidental Sample: Choosing respondents who are easily available—also called convenience sample.

Most psychology survey research is done this way, e.g., in a single clinic or organization.

Quota Sample: Choose respondents from an available population based on characteristics.

Example, being sure sample is half male and half female.

Probability sample necessary for giving a description of a population, e.g., political poll.

Nonprobability useful for showing relations among variables, e.g., people who are liberal prefer a certain candidate.

Allows branching of questions: probing

More in depth questionning

Inefficient: Costly and Time consuming

Not anonymous

Can include many questions

Can be anonymous

Doesn’t require face to face contact (or ear to ear with phone interview)

Efficient: Requires little researcher time and many respondents can be surveyed at once

Less control over who responds

Doesn’t allow feedback to clarify unclear questions