Types of experiments - randomized and quasi Flashcards
one shot
-quasi
-Only one group at single point in time
-No control or comparison group
-Run an ad campaign, measure sales
one group pretest posttest
-quasi
-Only one group but tested prior to and after “treatment”
- No control or comparison group
-Measure sales, run ad campaign, measure sales again
static group
-quasi
-Comparison between one group that got a treatment and one that
hasn’t
-Measuring differences between 2 groups of people
-Exercisers, non-exercisers, dependent variable is resting heart rate
before and after with control
-randomized
-Random assignment to either experimental or control group, everyone gets a pretest, experimental starts program exercise, then measure resting heart rate again
solomon four group
-randomized
-Most robust design but harder to implement due to time and
resource constraints.
-4 groups randomly assigned – 2 groups get a pretest, 2 groups do not, 2 groups get a treatment, 2 groups do not, and everybody gets a posttest
after only with control
-randomized
-When researchers want to eliminate testing effects, so only a
posttest measurement is done
-Randomly assigned to experimental or control group, then measure some outcome
-best one
quasi experiment
no random assignment to conditions
Threats to internal validity and control.
true experiment
random assignment to conditions
random assignment
-For internal validity
- To establish control in experiments
random sampling
-For external validity
-To be able to generalize from a sample to a larger population
How to establish a cause and effect relationship:
-Covaration
-Temporal precedence
-Internal validity
covariation
(correlation) positive or negative
temporal precedence
the cause precedes the effect
internal validity
★refers to our ability to rule out plausible alternative explanations for the relationship★
threats to internal validity
-History
-Maturation
-Testing (two types)
-Instrumentation
-Selection
-Mortality
history
occurs outside the experiment. Has an effect on the dependent variable.
maturation
refers to our participants getting older, smarter, wiser, growing up
testing
(interactive testing): The Practice, practice at time 1 gets you to do better at time 2.
instrumentation
any change from O1 to O2 is due to the measuring instrument
selection
refers to whether or not if we can randomly assign people to control
mortality
Mortality: people dropping out of the study. Cannot force them.
questionaire design
-demographics last
-satisficing v. optimizing
-funnel technique
-grice’s norms of conversation
-leading/loaded questions
-ranking v. rating
-types of questions
-open v. closd
satisficing/optimizing
people tend to satisfice → typically cognitively lazy
Weak and strong forms of satisficing
Want them to optimize
funnel technique
Start with broad questions, move to more specific questions
grice’s norms of conversation
- The maxim of quality: Try to make your contribution one that is true. Specifically, do not say what you believe to be false, and do not say that for which you lack adequate evidence
- The maxim of quantity: Make your contribution as informative as is required for the current purposes of the exchange, and do not make your contribution more informative than is required
- The maxim of relevance: Make your contribution relevant
- The maxim of manner: Be perspicuous, and specifically avoid obscurity, avoid ambiguity, be brief, and be orderly
-These 4 norms are things you don’t realize till someone violates them
leading/loaded questions
→ one that implies a specific answer (leading), emotionally charged (loading)
open questions
great because its a free flowing conversation, downside: get pages of transcribed information, and you have to develop a coding scheme to condense into data,
-Anytime you want a number –use should ask an open ended question! How much, how often, how old, etc
closed questions
-minimal to statistical analyses, faster, cheaper, and easier to analyze, very efficient, downside – rigid and artificial
types of closed questions
-Simple dichotomy – yes/no, like/dislike, approve/disapprove
-Determinant choice – multiple choice questions, scales
-Frequency determination – how often
For frequency determination questions, it’s better to use open ended questions
-Checklist – which of the following
-Constant sum – allocate 100 points
ratings/rankings
Ratings allow for non-differentiation
Ranking- – force people to differentiate