Qualitative research & Correlational research Flashcards
what is the purpose of qualitative research
attempts to capture complexity of human behaviour in its natural environment. generally comprised of detailed narratives of behaviour that focuses on the big picture
associate quantitative research with:
deductive logic
associate qualitative research with:
inductive logic
positivism
a philosophical position on how knowledge can be obtained. stresses the importance of OBSERVABLE FACTS (implies universal laws) in the valid accumulation of knowledge.
psychology today values:
empiricism but not positivism because laws cannot be established
one should conduct quantitative studies if:
if addressing very clearly specified research question, if substantial body of theory has already been established, when quantitative measures provide a good way to collect info
one should conduct qualitative studies if:
wish to study the complexity of something in its natural surroundings, if there is a lack of clarity about research question/hypothesis, if not enough body of theory to develop hypothesis
key words in correlational research:
association, relationship, how does x vary with y
in correlational research, both variables have to be:
continuous
correlation cannot specify:
causation
correlational research determines if:
there’s a relationship between two variables
what are possible explanations for correlation?
causation, reverse causation, another variable is involved
positive correlation
A increases as B increases
what is the value of r to signify a strong relationship?
magnitude of 0.5
negative correlation
A increases as B decreases
zero correlation
there is no LINEAR RELATIONSHIP between A and B
the difference between quantitative and qualitative research arises in:
data collection and data analysis
repeated measures study
participants tested for IV condition multiple times
internal validity is a determinant of:
experimental control, the extent to which we know DV is caused by the IV
external validity is a determinant of:
generalizability, the extent to which we can generalize our results to different populations
What are threats to internal validity?
selection, history, maturation, testing, instrumentation, regression to the mean, experimental mortality, interaction with selection
selection (as a threat to internal validity)
if groups are not equal prior to the experiment (ex. non-random selection and/or assignment)
history (as a threat to internal validity)
when events occur in the course of an experiment (after IV manipulation but before DV taken) that affects DV
maturation (as a threat to internal validity)
when changes occur in the participants over the time during their participation in the experiment which affects DV. Usually associated with repeated measures studies that take place over extended periods of time
testing (as a threat to internal validity)
when measuring the DV inadvertently causes a change in future measurements of DV (thus test can only be taken once by each participant)
instrumentation (as a threat to internal validity)
when the measurement criterion (procedure, coding) is changed during testing
regression to the mean (as a threat to internal validity)
when people with extreme scores are measured multiple times, their scores tend to drift closer to the mean (“ceiling/floor” effects)
experimental mortality (as a threat to internal validity)
when participants from different groups drop out of the experiment at different rates, AKA experimental attrition
interaction with selection (as a threat to internal validity)
when systematic differences exist between among groups and another threat acts with the pre-existing interaction (ex. selection and history, selection and maturation)
what is a way to solve for systematic unaccounted differences?
random assignment
pre-test-post-test control group design
random assignment-Observation-Control/Treatment-Observation
when is a pre-test-post-test control group design appropriate?
when testing a physiological aspect, likely to get consistent results for the control.
what does pre-test-post-test control group design control for?
selection, history, maturation, testing, interaction with selection, regression to the mean (but not maturation of over extended period of time)
post-test-only control group design
random assignment-treatment/control-observation: same protection from testing effects but less protection from selection. No check to see if manipulation worked, thus needs a large sample size
solomon four group (REQUIRES LARGE N)
random assignment for four groups, observe two groups, control/treatment for all four groups, observe all four groups (group with two observations allows for comparison, group with one observation contributes to generalizability=high in external validity)
single-case experimental design
one participant (N=1), involves a manipulated variable and attempts to exert some control over variables. 1. baseline measurement, 2. change one variable at a time, 3. repeated measures
when is a single-case experimental design suitable?
limitations on opportunity observe behaviour, no/limited variability between people, when a single negative instance would refute a theory
single-case experimental design examples
(A=baseline and Test/Observation, B=treatment and Test/Observation):
- AB
- ABA
- ABAB
quasi-experimental designs
cannot use random assignment but manipulates an IV to measure DV
when is quasi-experimental design suitable?
if studying a certain group, social condition, if it is unethical to withhold treatment, random assignment is too expensive/time-consuming (thus at least four conditions: pre-existing difference + control/treatment)
issue with quasi-experimental design?
not fully experimental (no random assignment) so cannot infer causation)
what are two types of quasi-experimental designs?
non-equivalent group design (AB, ABA, ABAB) and interrupted-time series (OOOOOXOOOOO)