Week 8 Flashcards
lab experiments
Take place in lab or controlled environment
Allows us to maximise control
Disadvantage: by maximising control, the researcher may create an artificial environment that may unduly influence the cause and effect relationship that the researcher is interested in
field experiments
Natural setting
IV manipulated by researcher but not with as much control as in lab
Advantage: far less artificial than the lab
between subjects design
each participant is measured under only one condition
within subjects design
each participant is measured under two or more conditions
matched-subjects design
each participant is measured under only one condition but is matched to subjects in the other condition based on certain characteristics
quasi experimental
Researcher is still comparing two or more groups, but this time they are naturally occurring groups and thus there is no true manipulation of the IV, nor is there random allocation of participants into each group
Attempts to achieve same goal as the experimental strategy, but without random assignment
Hard to establish that changes in the DV are actually caused by changes in the IV
types of quasi experimental
single group post test only
single group pretest/posttest
single group time series
nonequivalent control group posttest only
nonequivalent control group pretest/posttest design
multiple group time series
single group posttest only
Single group of participants given a treatment/intervention and then tested
Weakest type as it lacks a control or comparison group
single group pretest/posttest
Single group of participants take a pretest, then are given a treatment/intervention, before being tested again
Slightly stronger as there is now a control and treatment condition to compare
Order effect are difficult to combat with the use of counterbalancing because once the treatment group has been exposed to the intervention, they cannot go back to being ‘untreated’
single group time series
single group of participants are measured repeatedly before and after a particular treatment/intervention
Allows us to see whether the behaviour being measured is stable before the intervention and whether it changes or not across multiple time points after the treatment/intervention
nonequivalent control group posttest only
At least two nonequivalent groups are given an intervention, and then a post-test measure
There is now a comparison group added, but the control group is specified as nonequivalent because participants are not randomly assigned to either groups
nonequivalent control group pretest/posttest
At least two nonequivalent groups take a pretest, then are given an intervention before being tested again
There is now a comparison group added but is species as nonequivalent because participants are not randomly assigned to either experimental or control group
Due to lack of random assignment of participants or controlling variables, the researchers cannot be confident that the groups are equivalent
multiple group time series
Series of measures are taken on two or more groups, both before and after a treatment
Allows us to see whether the behaviour being measured is stable before the intervention, and whether is changes or not across multiple time points after the intervention
Also allows us to compare this between different groups to see whether the observed changes are consistent regardless of the group
Still subject to confounding variables
internal validity
if a research study is able to show that changes in the IV lead to changes in the DV (with no alternative variable) then it is said to have high internal validity
If there are alternative explanations then it has low internal validity
external validity
extent to which study findings are generalisable to populations other than the one from which participants have been drawn, as well as to other conditions and situations
three types of generalisability which may be of concern
- How well a sample generalises to the population
Threatened by selection bias, use of college students or volunteer participants, participant characteristics, or the involvement of a different species - How well the findings of one research study generalise to another research study
Threatened by novelty effects, multiple treatment interference, experimenter characteristics, assessment sensitisation, different measurement procedures, and time of measurement - How well the research study findings generalise to the real world
Threatened by any of the aforementioned issues, as well as by the particular setting of the research
threats to internal validity
subject effect history effect maturation effect experimenter effect testing effect regression to the mean ceiling effect instrumentation effect mortality (attrition) diffusion of treatment floor effect
subject effect
threat in which the subject, consciously or unconsciously, affects the results of the study
history effect
outside event that is not part of the manipulation of the experiment could be responsible for the results
maturation effect
naturally occurring changes within the subjects could be responsible for the observed results
experimenter effect
experimenter, consciously or unconsciously, affects the results of the study
testing effect
repeated testing leads to better or worse scores
regression to the mean
extreme scores, upon retesting, tend to be less extreme, moving towards the mean
ceiling effect
limitation of the measuring instrument that decreases its capability to differentiate between scores at the top of the scale
instrumentation effect
changes in the dependent variable may be due to changes in the measuring device
mortality (attrition)
differential dropout rates may be observed in the experimental and control groups, leading to inequality between the groups
diffusion of treatment
observed changes in the behaviours or responses of subjects may be due information received from other subjects in the study
floor effect
limitation of the measuring instrument that decreases its capability to differentiate between scores at the bottom of the scale
threats to external validity
interaction of causal effects with units
interaction of the causal effect over variations
interaction of the causal effect with outcome
interaction of causal effect with settings
interaction of causal effect with units
causal effect in one type of unit may not apply to another type
interaction of causal effect over variations
an effect for one condition/treatment may not hold when it is varied or combined with other treatments
interaction of causal effect with outcome
causal effects may not apply to other types of outcomes
interaction of causal effect with settings
causal effect in one setting may not generalise to other settings
characteristics that differentiate experiments from other types of research studies
manipulation
control
Research typically uses three basic techniques to control other variables:
Random assignment: distributes participant characteristics; controls environmental variables
Matching
Holding variables constant: using only specific age group and gender in research to hold age and gender constant
posttest only control group design
an experimental design in which the dependent variable is measured after the manipulation of the independent variable
pretest/posttest control group design
DV is measured both before and after manipulation of the IV
solomon four group design
a design with four groups that is a combination of the post-test only control group design and the pretest/posttest control group design
demand characteristics
subjects try to guess what characteristics the experimenter is in effect ‘demanding’
Hawthorne effect
type of reactivity in which subjects improve or change an aspect of their behaviour because they know that they are being studied, rather than in response to the experimental manipulation
college sophomore problem
external validity problem from using mainly college sophomores as subject in research studies; response to criticism:
Using them as subjects in a study does not negate the findings of the study
In the research conducted in many areas of psychology, the college sophomore problem is not an issue
Population of college students today is varied
correlated groups design
subjects in the experiment and control groups are related in some way
counterbalancing
mechanism for controlling order effects either by including all orders of treatment presentation or by randomly determining the order for each subject
latin square
counter balancing technique to control for order effects without using all possible orders
Have the same number of orders as we do conditions
carryover effects
a type of testing effect often present in within-subjects designs in which subjects ‘carry’ something with them from one direction to another
sequential design
Developmental design that is a combination of cross sectional and longitudinal designs
Researcher begins with participant of different ages and tests them, later they retest the same individuals
