Chapter 9 Flashcards
What is the most frequently used mode of gathering data in the social science? what does it involve?
Survey research
-it involves selecting a sample of respondents and then collecting info from them with a questionnaire or interview
Surveys are versatile, what three purposes can they be used for..
- exploratory
- descriptive
- explanatory purposes
The unit of analysis in surveys
- usually the individual
- sometimes used to report on the household
What do surveys study
the attitudes, beliefs, values, and orientations of populations
Research with low legitimacy is sometimes carried out as survey research, what are some examples…
marketing research
political push polls
Most survey concepts are measured with what two things?
questions or statements
Examples of statements that are presented to respondents who are asked to respond in some way:
for example, asked to indicate how much they agree or disagree with a position expressed as a statement, how often they engage in behavior described in statement, etc.
What forms of questions can be asked? Explain both
Closed-ended: format is easier to process but provides answers with less depth
Open-ended: more in depth
Good questions should be (7 things)
- relevant to the respondent
- clear
- as short as possible
- possible for respondent to answer
- not biased
- not double barreled
- not subjected to social desirability issues
Questionnaire formatting
- well organized
- uncluttered ( a lot of white space)
- attractive
Formatting for close-ended questions
use boxes for each response, evenly and adequately spaced
when should contingency questions be used
should be used when needed to avoid asking irrelevant questions of the respondent
When is the matrix format useful
when asking several questions that use the same response format (answer categories).
What is one problem with the matrix format?
respondents will sometimes select the response from the same category without reading the question. This is called a RESPONSE SET
How to order questions in the questionnaire
– Question order effects are real. An order effect happens when
answers to a question affects response to later questions.
– Questionnaires should begin with a set of very interesting questions
to generate respondent interest in the study.
Why and who pretests questionnaires?
- need to identify issues before “going live”
- should be tested on respondents who are similar to the kinds of people who will respond to the questionnaire in the real investigation
How are self-administered questionnaire surveys done?
they are most often administered by mail, but can also be administered to a group of respondents
another strategy is the “drop off/pick up” approach
Process of mail surveys
- arrangements should be made in mail surveys for the questionnaire to be returned at the researchers expense
- the researcher needs to monitor the return of questionnaires, this allows for follow up mailings to be sent out as well as the calculation of response rates
- follow up mailings are essential in mail surveys. two follow-ups mailed two or three weeks apart are best
What advantages do interviews have over self administered questionnaire surveys?
they tend to have higher response rates, they decrease the number of “idk” and “no answer” responses, they help correct confusing items, and they provide the opportunity to observe the social situation in which the data are obtained
What are interviewer “musts”?
- must be neutral
- must have pleasant demeanor
- must dress in a manner similar to the person being interviewed
- must be familiar with the questionnaire being used
- must read questions as worded and record responses exactly
- must be good at probing incomplete answers
- must be thoroughly trained and monitored by supervisors
- must be taught how to handle difficult and confusing situations
What do telephone interview surveys save?
money and time
What may respondents be more willing to give over phone rather than in person?
socially disapproved or embarrassing responses
what is known about safety concerns when interviewing over phone?
they have less safety concerns rather than having to enter unsafe communities, etc.
What makes the obtainment of an acceptable response rate challenging for phone interviews?
they can hang up
What is necessary to gain access to respondents with unlisted phone numbers?
(RDD) random digit dialing
Why is the obtainment of representative populations for phone surveys growing increasingly difficult?
because most cell phone numbers are excluded from telephone sampling frames
some households are now “cell phone only households”
how has telemarketing influenced people?
it has made people less willing to participate in phone surveys, many people assume that a survey is a hidden attempt to sell products and services
What is CATI and why do we use it?
- computer-assisted telephone interviewing
- telephone surveys usually use computers to enhance reliability and validity
What is increasingly using computers?
computer-assisted face-to-face (personal) interviewing
What is the chief problem with web-based surveys?
-obtaining a sample of respondents that is representative of the population of interest. web surveys tend to have response rates similar to what are obtained in mail surveys
Relative advantages overall
–self administered surveys are less expensive to carry out, faster to complete, and are especially appropriate when the survey is not about sensitive topics
–interviews tend to lead to more complete data and higher response rates, but are expensive and take a lot of time to complete. interviews are especially appropriate for topics that are complicated. Interviews can also be carried out when all one has is a sampling frame of addresses
strengths of surveys over other methods
- -better when one wishes to describe the characteristics of large populations
- -flexible and the use of standardized procedures and questions enhances reliability
weaknesses of surveys over other methods
–standardization often results in the investigator using closed-ended questions, overlooking other appropriate responses to questions
–in short, surveys are weak on validity but strong on reliability
secondary analysis of survey data
–survey researchers often do not have the resources to carry out their own large-scale surveys. for this reason, they may rely on secondary analysis of survey data gathered for other purposes
–secondary data analysis is low cost
–a key problem of secondary data analysis is validity. data collected for one purpose may not always be relevant to another researchers interests. a question asked in the original study, for example, might be different than how the secondary data analyst would have asked it if she/he had the resources to carry out a survey