PSYC-2018 Quiz Part Two
Random sampling
Every member has an equal chance of being selected
Stratified sampling
Proportions of an important subgroup of the population must be represented (ex. 75% female, 25% male)
Cluster sampling
Randomly choosing from a cluster of people with the same trait
Convience sampling
Simply asking the people around you, non-representative
Quota sampling
Only recruiting until you reach the number you need
Snowball sampling
Asking participants to tell people they know to participate in the study
Types of non probability sampling
Quota
convience
snowball
Reliability
Reliable and consitent results
produce the same results over and over
Validity
Measures what it is supposed to measure
Face validity
How effective a study LOOKS
Criterion
Will the test be able to predict what happens for future events
Construct
extent that the test properly meaasures what it is designed to
Nominal Scale
Assigning numbers to classify into groups (numbers as names for people)
Oridinal Scale
Numbers indicate rank ordering
- Doesn't indicate distance between two points
Interval Scale
Scores indicate quantities BUT zero is just a point on the continume (doesn't mean an absence of something)
Ratio scales
Scores idicate quantities BUT zero means nothing is there
Descriptive statistics
Describes the sample data
Inferential statistics
Lets us infer conclussions from that data
Null Hypothesis
Relationship between variables in the population expected
Type 1 error (null hypothesis)
Rejecting null hypothesis but being wrong
Type 2 error (null hypothesis
Sailure to reject null error but you're wrong
Alternative hypothesis
Relationship between variables in population is expected given the sample
Confidence intervals
range that the population is likely to be found
Power
Chance of rejecting a null hypothesis
3 components of establishing the independant
1) Manipulated IV
2) Experimental Groups
3) Control groups
Confounds (controlling extraneous variables)
Any uncontrolled conditions
Ceiling effect
Task is too easy so everyone scores high
Floor effect
Task is too hard so everyone scores low
Subjective variables
Things already existing within participants that cannot be controlled or manipulated
External validity studies (3 things)
1) Generalize to other populations
2) Generalize to otehr envoirnments
3) Generalize to other times
External validity
How well does the study generalize to contexts outside of the study
Internal validity
How well the study actually answers the proposed question
3 things needed for internal validity
1) Valid operational defintions
2) Valid measures
3) No confounds
Maturation (internal validity threat)
how one matures over time of the study
Regression to the mean (internal validity threat)
Checking at later times may change results
Testing anf instrumention (internal validity threat)
issues with expired/not working equipment or methods
Importance of using a control group (internal validity threat)
CLusters of people who drop out with the same characteristics
Between subject design
Comparing the difference between two sets opf subjects
main problem to solve in between subject design
Creating equivalent groups
Matching
Deliberate control over a potential confound
Within subject design
Looking for a difference within a group
aka repeated measure design
Main problem to solve within-subject design
Difference in results based on the order contition is presented in (order effect)
Counterbalancing
Altering the order of experiemental conditions
Partial counter balancing
Random sampling of all possible combinations is selected
Complete counter balancing
Testing in every possible order at least once
Cross sectional design
between subject design (potential for cohort effect)
Longtudinal design
Within-subject design, Subjects tested at differentg points in their life
Cohort sequential design
testing different generations for the same thing
Experimenter bias
experimenters expectations can influence subject behavoir
Controls for experimenter bias
double blind procedure
Hawthorne effect
Knwoing one is in a study
good subject effect
Participants tend to be cooperative to please researcher
Evaluation apprehension
Participants behave in ideal ways to avoid negative evaluation
Demand bias
Cues giving away true purpose of the study