Final exam Flashcards
Threats to internal validity: History
an external event occurs that
affects the results of the study
Threats to internal validity: Maturation
respondents change over
time, which affects the DV/outcome
Threats to internal validity: Experimenter bias
attitudes or behavior
of researcher affect results
Threats to internal validity: Testing/sensitization
participant is
familiar with the measure (increase their
scores)
Threats to internal validity: Regression to the mean
people may score higher or
lower on a measure, then move toward the mean when
measured again
Threats to internal validity: Experimental mortality
people dropping out of the
study
Threats to internal validity: Contamination
people who have been part of the study
tell others what the study is about
Threats to internal validity: Sample bias/non-equivalent groups-
groups in conditions
are not equivalent before starting study
Key Elements of an Experiment: Manipulation of independent variable(s)
▪ Create different conditions/groups that
receive different treatments
▪ Control group- no treatment
▪ Treatment group- treatment
▪ Measure the effect (DV) after exposure to the
cause (IV)
Random assignment of participants
In theory, this creates equivalent (similar)
samples in different conditions
Problems with Post-Test Only
Control Group: sample bias
No way to assess if these
groups were different to begin
with
Factorial Design
Experimental studies with two or more
independent variables
To understand whether the combination of
two or more variables increases/decreases
effects (i.e., look at interactions)
Interpreting 2X2 Factorial Design:
Two IVs, each IV has two levels
So we end up with four conditions/groups
Within-Subjects Designs
Advantages
* Don’t have to worry about individual differences
* Fewer participants are required
Disadvantages
* Order effects
* Fatigue
o People could get sick/tired of experiment
o People get worse at game because they don’t want
to play anymore
* Solution: Counter-balanced design
o Participant 1: X1O1 X2O2 X3O3 X4O4
o Participant 2: X2O2 X3O3 X4O4 X1O1
Uses of Content Analysis
Describe how much or what kind of
certain messages exist
❖E.g., How frequently are “dogs” and
“trucks” mentioned in top country songs?
❖What type of messages do Fortune 500
CEOs communicate on Twitter? (e.g.,
company vision, personal life etc.)
Compare media content to the “real
world”
❖E.g., How many criminals in prime‐time
dramas are minorities vs. actual U.S.
convicts that are minorities?
Uses of Content Analysis 2
Assess the “image” of a particular
group
* E.g., How is Nike portrayed by its
competitors in their annual report?
* How are Syrian immigrants portrayed
in French TV media?
Discussion Question: What kind of
content analysis do we do in
PR/COM/Marketing? Examples?
Content Analysis Steps
- Formulate a research question or
hypothesis - Define the population
- Select a sample
- Define the unit of analysis
- Construct the categories of interest
- Train the coders
- Assess reliability
- Analyze the data
Formulate a research question &
Defining the Population
RQ: How is drunk driving portrayed
in primetime television shows?
Need to define:
Drunk driving
Implied or explicit
Primetime
What hours?
Television shows
Cable, broadcast, streaming services?
Code data
Typically need 2-3 people
They must understand and agree on
coding rules.
So, we need to create a codebook to
provide these rules…
Individuals must be trained to use
codebook.
Compute Reliability
Coders should agree in what they see. If
not, that’s a problem.
To ensure that the coding scheme is
reliable we have to test it:
Coders analyze identical content
Results are compared using statistical
tests for reliability
Intercoder reliability: 70% low bottom
Intercoder Reliability
Agreement
between coders
Select a Sample: Example
Often multi-stage cluster sampling
* Selecting Fortune 500 CEO communication
channels
* All media -> Social media -> Twitter
* Picking a sample of content of a workable
size
* Randomly pick 5 years’ content
* Sampling tweets
* Every 3rd tweet