week 5 / chapter 10 cards Flashcards
what does rigor mean?
the strength of the methodology of the study
what does bias mean?
unseen errors/ design flaws
looking at how much bias was involved
validity means in terms of quantitative study
the appropriate truth of an inference
what is meant by a threat to validity
means hen the inference could go wrong
what is a general idea of a threat to validity
confounding variable
how do we best handle confounding variables
randomization
other then randomization, how can we eliminate confounding variables
crossover, matching, blocking, statistical control
what is the definition of homogeneity
standardization of a study (can be done by restricting to a specific domain, and is done to eliminate variability on the confounding characteristics)
name the 4 types of validity that affect the rigor of a quantitative study:
- statistical conclusion validity
- internal validity
- external validity
- construct validity
what are we looking for when thinking about statistical conclusion validity
a true empirical relationship!!!
is there actually a relationship between cause and effect
how could the statistical conclusion validity be impacted?
- low statistical power
- low precision
- if the independent variable operailation is impacted
what is the main thing to think about with the internal validity?
was it actually due to the intervention!!!
what does independent variable stand for?
intervention
how can the internal validity be impacted?
temporal ambiguity selection history maturation mortality testing instrumentation
what does construct validity deal with?
thinking about the setting / other people / and their possible impact on the study
what effect do we need to be aware of in terms of construct validity
Hawthorne effect
what is the Hawthorne effect
how the participants may react different knowing they are being watched/ studied
what word should we think about when we hear external validity
generalizable
so external validity is concerned with:
can this study be recreated
what is the number one threat to statical conclusion
statistical power
reliability and validity can also be known as
trustworthiness
internal validity can also be known as
credibility
reliability can also be known as…
dependability
confirmability is also known as …
objectivity
external validity is also known as …
transferability
the extent to which the appropriate inferences can be made is known as
validity
why your inferences may be wrong is due to
threats to validity
the conclusion that there is actually a relationship that exists between variables is known as
statistical conclusion validity
how confident are we in the independent variable
internal validity (credibility)
does your test accurately assess what it is supposed to do
construct validity
generalizability of results (is the sample representative)
external validity
what are 6 ways to control the confounding variable
- randomization
- crossover
- homogeneity
- blocking/ stratification
- matching
- statistical control
the best way to control confounding variables is through
randomization or crossover
threats to internal validity include
temporal ambiguity selection history maturation attrition
