Research
See if independent and dependent variables measure what they are supposed to
Subject, Variable, and Setting Representativeness
How can external validity be established?
Poses a problem in interpreting the results of an experiment because while you think it may be one thing, it may not be
Each member of the population has an equal chance of being selected for the sample
What is Random Sampling?
relative vs quantitative
Magnitude of attribute
If results are generalizable across subjects, variables, settings
What is External Validity?
Events that occur while a study is being conducted (not part of experimental manipulation)
How can history pose a threat to internal validity?
when participants in each treatment condition (at the outset of the study) already differ from one another on some dimension
How could Selection pose as a threat to internal validity?
minimize confoundings of order effects and a priori differences
What do random sampling and random assignment accomplish?
Equal vs unequal/unknown
Intervals between values
infer our study demonstrated one variable had a causal effect on another (validity of making causal/explanatory conclusions)
What is Internal Validity?
way in which people change naturally over the course of a study (independent of participation)
How can Maturation pose a threat to internal validity?
Assign subjects to experimental conditions on a random basis
What is Random Assignment?
Zero point
Magnitude of attribute
Intervals between values
Zero point
Three criteria by which measurement scales differ
Whether the act of measuring one's responses affects his/her behavior on measures (practicing)
How could testing pose a threat to internal validity?
How could instrumentation pose a threat to internal validity?
Magnitude: relative/quantitative
Intervals: defined
Zero point: none
Interval
Magnitude: relative/quantitative
Intervals: none
Zero point: none
Ordinal
Magnitude: none
Intervals: none
Zero point: none
Nominal Scale
when participants fail to complete a study
How could Attrition pose a threat to internal validity?
Confounding variables (give examples)
Random Error of Measurement (give examples)
What are threats to construct validity?
all subjects get all treatments
different groups get different treatments
Between-Subjects
Explain what is involved in each of the following reliability issues:
Reliability of Measurements
Statistical reliability
Experimental reliability
Test reliability
Reliability of Measurements: Measured score (true score + error)
Statistical reliability: how likely are the results due to chance? *p < 0.05, reject the possibility of chance*
Experimental reliability: does a replication of the experiment yield the same results? (repeat experiment)
Test reliability: take the same measures on different occasions (test-retest)
Magnitude: relative/quantitative
Intervals: defined
Zero point: true zero point
Ratio