CHPT 9: Studying People Quantitatively Flashcards
(29 cards)
Surveys
Involves asking questions of individuals selected from a population to measure relationships between variables
no manipulation of variables
Goals of Surveys: Descriptive
Identify/describe attitudes or behaviors in a given population.
ex: Does X predict/relate to Y?
Example: Does exposure to alcohol ads (X) predict teen drinking (Y)?
Goals of Surveys: Explanatory
Examine relationships between the attitude or behavioral variables measured.
Do many factors together predict Y?
Example: Do alcohol ads (X.1), parent drinking (X.2), peer drinking (X.3), & risk-taking (X.4) together predict teen drinking (Y)?
Interview-Style Survey
a researcher/research assistant asks questions and record the responses given
Advantages: rapport building, get more detailed responses
Disadvantages: more $$ and time costly, each respondent should have a nearly identical experience
FTF survey
involve the researcher and the survey respondent directly interacting with one another in the same place
Advantages: develop rapport, interpret nonverbal communication, and clarify
Disadvantages: resource intensive and geographically limited
Telephone survey
provide a greater sense of privacy while still building rapport and engage in probing
Self-administered survey
rely on the participants to complete the questionnaire on their own without any assistance.
Internet Survey
use a web-based interface to collect responses from survey participants.
Mail-survey
often called a “paper-and-pencil”, involves sending an easily completed questionnaire to sampled individuals
Opinion Polls
refers to a survey that samples from a large, geographically defined population and primarily measures the overall distribution of public opinion
Causality
- must be connection between IV & DV
-must establish time order (IV-> DV)
-Must rule out other explanations/causes
Experimental design
- participants will be exposed to at least one version of the stimulus
-participants will be randomly assigned to condition
Random assignment
random determination of which condition each participant falls into
-each person is equally likely to be placed into a given condition
Helps address 2/3 requirements for causality
-establishment of time order
-elimination of alternative explanations (confounds)
Matched (or paired) assignment
intentionally linking similar people into a group or pair, then making sure one member of each group is assigned to each condition
ex: testing whether a new website is more effective than an existing one for e-commerce, you want to be sure that you have equal numbers of prior site users in each group. You could ask participants whether and how often they visited the old site and identify pairs of individuals who made roughly equal use of the site. Then, you would randomly assign one of those individuals to each condition
Between-subjects Design
each participant is assigned to just one condition and comparisons are made between the groups exposed to each condition
ex: you have 3versions of magazine content showing different types of models. You might measure changes in eating behavior by having participants choose from different snack options after viewing the images. Each person would see only 1 of the 3 layouts, but you would have three groups of people whose eating behavior could be compared.
Within-suibjects Design
focuses on comparing participants with themselves (presents participants w multiple stimuli)
ex: magazine example w 3 models (stimuli). Each participant views all 3 (random time order).
Now can compare them to themselves, person chooses healthy snack after attractive model (bc of low self-esteem)
Double-blind design
when both the researcher and participants are not aware of their treatment group (both do not know which condition they are in, preventing them from changing behaviors)
ex: is a medical study in which the effectiveness of a drug is being tested. The treatment group receives the medicine (the treatment), and another group, the control group, receives a placebo pill.
Pre-test/Post-test
those evaluated before the stimulus vs those evaluated after the stimulus
ex: in evaluating the effects of thin-ideal media, you would probably want to measure beliefs about ideal body image. If you want to know exactly how people’s attitude changed as a result of exposure, you could measure those attitudes before and after image viewing
issues: sensitization
Solomon four-group design
Tests whether a pre-test affects results
Involves 4 groups:
-Pre-test + Treatment+ Post test
-Pre-test + No Treatment + Post test
-Treatment + Post test
-No Treatment + Post test
Quasi-Experimental Design
involves comparing responses to different treatments, does not include random assignment to treatment conditions
ex: you might want to study the effectiveness of a new approach to teaching media literacy. One instructor might use this approach, while other classes at the university do not. Give the students in every class a test on media literacy at the end of the semester. If those in the class using the new approach score better, that might be because of the new approach. BUT could be due to them being more skilled
Field Experiment
takes place outside of the laboratory (real world setting) and manipulates the independent variable in a natural way
ex: want to test a fundraising campaign in which postcards are sent out requesting donations. You would randomly assign people to 2 groups and send each condition a diff. postcard–> see which group sends more money
Adv: increases external validity
Dis adv: reduces control over variables
Treatment Groups
individuals are randomly assigned to a group that represents their experimental conditions
Stimulus
the specific experience presented to participants in a given condition of an experiment
ex: if participants are viewing a movie that has a few scenes altered, the [blank} is the entire movie
Stimulus sampling
an approach that creates a set of stimuli that all represent the same specific treatment condition and then randomly assigns people to one stimulus from that set
