Survey Research Flashcards
Surveys useful to directly observe;
Cognitions e.g. stereotypes
Emotions e.g. attachment
Behaviours hard to observe directly e.g. sex
What is a survey
Field study
Collection of data from sample of people from well-defined pop
Through the use of a questionnaire
Distinguishing characteristics
Used to quantitatively deserve specific aspect of given pop
By asking people questions (self-report)
Collect data from a subset of pop
Advantages
Enables to study large variety of topics in real life natural settings
Increased confidence in generalising findings to gen pop
Provides ideal conditions to examine moderating role of personal, social or cultural variables
Disadvantages
Correlational nature
We usually do not manipulate IV’s & therefore cannot infer causal relationships between variables
Cross-sectional design
Collection of data at a single point in time from a sample drawn from a specified pop
Often used to assess- frequency with which people enact certain behaviours, number of people who have particular attitudes or beliefs
Also used to assess differences between subgroups in pop
Also provides opportunity to assess relations between variables
Repeated cross-sectional design/ successive independent samples design
Date collected from independent samples drawn from the same pop at 2 or more points in time
Useful if wanting to describe changes in additives or behaviours over time
May provide evidence to support causal relation- changes over time in IV should be mirrored by changes in DV
Panel Survey (longitudinal design)
Data collected from same individuals at 2 or more points in time
Test causal hypotheses in at least 2 ways-
1) Examine whether changes over time in an IV correspond to change in DV
2) examine whether initial levels of an IV predict changes in DV
Experiments within surveys
Respondents randomly assigned to different versions of questionnaire
Differences in responses between groups can be attributed to specific elements that were manipulated
Hypothetical scenarios
Manipulate certain characteristic of actor or circumstances
Examine Pp reactions & attitudes
Survey sampling
Sampling is a procedure in which a specified number of individuals are drawn from a sampling frame that represents an actual list of the pop
Probability sampling
Selection procedures in which individuals randomly selected from sampling frame
Equal chance per pp
Advantages of probability sampling
Increased representativeness
Generalisability
Types of probability sampling
Simple random sampling
Stratified sampling
Cluster sampling
Simple random sampling
Individuals drawn from the pop at random, equal chance
Stratified sampling
Sampling frame is divided into subgroups (strata) & sampling process conducted separately on each strata
Cluster sampling
Sample of individuals is drawn from natural groups (clusters) rather than one-by-one
Then all individuals within a given cluster are sampled
Evaluation of cluster sampling
\+ve= reduces time & cost -ve= reduces accuracy/representativeness
Measurement error
What can go wrong
Increasing sample size tends to reduce sampling error but cannot correct methodological problems such as non-response & coverage errors
Sampling error
Discrepancy between sample data & true pop data
Caused by random differences between sample & sampling frame
Chances are dependent on sample size, variance & pop size
Can calculate estimates of sampling error & know the magnitude of uncertainty regarding our data
Non-response error
Often some individuals chosen for the sample are unable or unwilling to participate
Error occurs when sampled individuals who did not respond differ systematically from those who did
Need to minimise apparently costs of responding & maximise rewards for doing so
Low response rate does not equal non-response error
Coverage Error
Discrepancy between sampling frame & the pop
Non-probability sampling
Selection procedures in which individuals not randomly selected
Advantages of non-probability sampling
Convenience
Cost
Convenience Sampling
Pp selected solely on basis of convenience
e.g. through newspaper ads
Problem with convenience sample
People who happen to be in certain place or respond to ads do not represent the pop
Purposive Sampling
Convenience sampling among members of particular subgroup within pop
Snowball sampling
A few members of subgroup located, each asked to suggest other members of subgroup for researcher to contact
Quota sampling
Members of various subgroups of pop selected to create sample that reflects known characteristics of pop
Relevant demographics, plan in advance how many individuals recruited from each demographic category
Base quotas on accurate info about composition of pop
Relatively representative sample compared to convenience sampling but not probability sampling