Survey Research: Maria Lucia Barron
Survey Research: Maria Lucia Barron
Survey Research: Maria Lucia Barron
Outline
Introduction
Classification of surveys
Steps in survey research
Some problems in survey research
References
Introduction.
Survey (noun).
STATISTICS analysis of poll sample: a statistical analysis of
answers to a poll of a sample of a population, for example, to
determine opinions, preferences, or knowledge. [1].
A method of gathering information from a sample of individuals. [2]
A powerful, scientific tool for gathering accurate and useful
information. [8]
Gathering information, asking questions, and sample of population.
Definitions
Census.
Any count: any systematic count or survey.
Count of population: an official count of a population carried
out at set intervals.
Sample.
STATISTICS group selected for testing: a representative
selection of a population that is examined to gain statistical
information about the whole.[1]
Random
STATISTICS equally likely: relating or belonging to a set in
which all the members have the same probability of occurrence.
[1]
4
Classification of Surveys
By size (social, community and school surveys).[9]
By result type (quantitative or qualitative).
Time and population.[4]
Qualitative vs Quantitative
Focus group [2]
Provide qualitative information
A wide range of information can be gathered in a relative short time
span.
The moderator can explore related topics.
Do not require complex sampling techniques.
The sample is neither random nor representative of the population.
The quality of the data is influenced by the skills of the moderator.
Allow researchers to use actual words and behavior of the
participants rather than counting response options.
The analysis is truly qualitative.
10
Types of Surveys[4]
(Time and Population)
A.
B.
Trend study.
Cohort study.
Panel study.
11
Example.
A researcher might be interested in the attitudes of high
school principals toward the use of flexible scheduling.
Population: high school principals in Brevard county.
Instrument: questionnaire.
Number of samples: 3 in a school year.
12
13
Disadvantages
Cross-sectional
Trend Study
Members of population
can change over time
without affecting the
result.
Cohort study
It is possible the
selection of different
sample.
Members of population
can not change.
Panel study
Loss of individuals
could happen.
15
2.
3.
4.
5.
6.
7.
8.
9.
Data analysis
10.
Results
16
17
18
19
21
22
23
24
The Questionnaire[5]
A effective questionnaire will follow this rules:
$_________
Cheese
$_________
Cottage cheese $_________
Sour cream
$_________
Other dairy prod. $_________
26
27
YES
NO
UNSURE OR UNDECIDED
FAVOR
NOT FAVOR
UNSURE OR UNDECIDED
FAVOR
NOT FAVOR
UNSURE OR UNDECIDED
28
Types of Errors.
Coverage error occurs when the list (or frame) from which a
sample is drawn does not include all elements of the population
that researchers wish to study.
Sampling error occurs when researchers survey only a subset
or sample of all people in the population instead of conducting a
census.
Measurement error occurs when a respondents answer to a
giving question is inaccurate, imprecise, or can not be
compared in any useful way to other respondents answers.
Nonresponse error occurs when a significant number of
people in the survey sample do not respond to the questionnaire
and are different from those who do in a way that is important to
the study.
31
Errors Examples.
Prior to the 1936 united states presidential election, pollsters for the
magazine Literary Digest mailed postcards to more than 10 million
people who were listed in telephone directories or as registered
owners of automobiles. The cards asked for whom they intended to
vote. Based on the more than 2 million ballots that were returned,
the Literary Digest digest predicted that republican candidate Alfred
M. Landon would win in a landslide over democrat Franklin D.
Roosevelt. At the time, however, more republicans than democrats
owned telephones and automobiles, skewing the poll results. In the
election, Landon won only two states.
32
Errors Examples.
Interviewer:
Respondent:
Interviewer:
Respondent:
No.
33
Errors examples.
When asked about welfare, a majority of Americans
in one survey said that the government spends too
much money. But when asked about assistance to
the poor, significantly fewer people gave this
response.
34
61% favor
25% oppose
14% no opinion
35
Telephone
Personal
Interview
Comparative cost
Lowest
SAME
SAME
High
Facilities needed?
Yes
No
No
Yes
Yes
Yes
No
Yes
Data-collection time
Shortest
Short
Longer
Longest
Response rate
Very high
Good
Poorest
Very high
Yes
No
No
Yes
Possibly
Yes
Yes
Yes
Yes
No
Yes
No
No
Yes
No
Yes
Somewhat
Somewhat
Best
Weak
Standardization of responses.
Easy
Somewhat
Easy
Hardest
37
Quick results
Less expensive than PI
Disadvantages
Nonresponse
Incomplete list
Type of questions to ask
Direct
administra Short response time
tion
Personal
interview
Expensive
Training needed
More sensitive to bias
Noncoverage errors
Low response rate
Long response time
Dont know who respond
38
Disadvantages
Internet
39
Close-ended questions.
Open-ended questions.
Partially close-ended questions.
Missing data.
40
41
42
Vocabulary used.
45
References
1.
2.
3.
4.
5.
6.
7.
8.
Encarta world English dictionary [north American edition] & (P) 2001
Microsoft corporation. All rights reserved.
Brochure, what is a survey?, Bill Kalsbeek, 1995 publications officer, ASA
section on survey research methods.
http://www.deakin.edu.au/~agoodman/sci101/#RTFToC2.
Developed by Albert Goodman
school of computing and mathematics Deakin university
copyright Deakin university (1995, 1999).
Frankel, J. And Wallen, N.. How to design and valuate research in education.
McGrawHill (2000).
Hillway, T. Introduction to research. H.M.Co. (1964).
Http://www.chssc.salford.Ac.uk/healthSci/resmeth2000/resmeth/ch2surv.htm
Research methods for health science.
"Psychology". Microsoft Encarta online encyclopedia 2001
http://encarta.Msn.Com (13 mar. 2002).
Salant, P. and Dillman, D. How to conduct your own survey. Wiley (1994).
46