chapter 6 Flashcards
(31 cards)
true experiments
- investigate cause and effect
- can’t always be done for ethical reasons
characteristics of an experimental design
- comparison group
- random assignment
- manipulate one (or more) variable(s) and measure the effect on one (or more) variable(s)
- strongest internal validity
three criteria for determining a causal relationship
- temporal precedence
- covariation of the cause and effect
- no plausible alternative explanations
temporal precedence - cause and effect criteria
- demonstrate that cause precedes effect
- cause predicts effect
- examine timing of the variables
- challenging
covariation of the cause and effect - cause and effect criteria
- demonstrate that the cause and effect are related
- if cause occurs effect occurs
- if cause does not occur effect does not occur
- effect increases as cause increases
no plausible alternative explanations - cause and effect criteria
- confounding variables; account for theme via control and measure
- third variable problem
- eliminating threats to internal validity
single-group threats to internal validity
- history
- maturation
- testing
- instrumentation
- mortality
- regression
- we deal with these threats by adding a control group
history - single-group threat
- relate to passage of time
- an event outside the study occurs that has an effect on our subject variable
maturation - single-group threat
- relate to passage of time
- changes that happen within subjects that occurs because of the passage of time
- e.g. physical maturation
testing - single-group threat
- when the act of measuring the dependent variable during the pretest affects participants’ responses at posttest
- e.g. reactivity due to self-monitoring
instrumentation - single-group threat
- when the basic characteristics of the measuring instrument change over time
- e.g. broken scales when measuring rate
- solution: standardize instrumentation
mortality - single-group threat
- subjects can drop out
- also need to examine if a certain group of people always tend to drop out?
regression to the mean
the statistical fact that an individual who scores extremely high or extremely low on a variable on one occasion will tend to score less extremely on the next occasion
spontaneous remission
the tendency for many medical and psychological problems to
improve over time without any form of treatment
regression - single-group threat
subjects with extreme scores over time will move the average of the group
multi-group threats to internal validity
- there really is only one multiple-group threat to internal validity in that the groups were not comparable/the same before the study
- emphasizes importance of random assignment
- random assignment creates comparable groups if sample is large enough
other ways to rule out threats to internal validity
- by argument: logically argue against the threat
- by measurement or observation: adding a threatening variable as a variable in the study so it can be measured
- by analysis: T tests to show groups are similar in confounds
- by preventative action
between subjects design (independent groups)
- different participants are assigned to each level of the independent variable
- 2 separate groups of people and comparing their results
- what effects one groups won’t necessarily affect the other because they are independent of each other
pre-test/posttest design
- ensure equivalent groups
- pretest is given to each group prior to introduction of the experimental manipulation
when are pre-test/posttest designs useful
- sample size is small; allows you to assess the equivalency of groups
- when you need specific participants for the experiment you can pre-test to screen subjects in order to select the people you need/want to answer the question
- when the likelihood of mortality (dropout) is high
disadvantages of a pre-test/posttest design
- can be time-consuming + awkward to administer
- can sensitize participant to what is being studied, thus creating demand characteristics and subject expectancy which can alter behaviour and introduce error to the study b/c most subjects try to be “good subjects”
- can reduce external validity b/c your subjects aren’t acting like normal people but are acting like good research subjects which decreases validity and reflection in the real world
- introduces unique confound: how can you know the effect is not due to the pretest?
solomon 4 group design - pre-test/posttest
- true experimental design
- purpose is to assess the impact of the pre-test
- treat the pretest as a second independent variable; run the study twice with 4 groups
- half of the participants receive the pre-test and the post-test and half receive only the post-test to examine the effect of the pre-test vs. not
within subjects design (repeated measure)
- the same participants experience all levels of the independent variable
- measuring more then once within the same group of people
- they are the independent and control
benefits of within subjects design (repeated measure)
- fewer participants needed because we only need one group
- examine questions related to change over time or due to experience b/c these tests are mainly longitudinal so it allows us to assess over time
- individual differences remain constant because it’s the same group of people
