Chapters 6,7 Flashcards
experimentors
investigations where researcher manipulates an independent variable
nuance for experimements
groups of participants must be equal
correlational studies
an investigation that explores the effect of a subject variable on a dependent variable
how to make it experimental
randomly assign participants into groups, placebo control
between groups experiment
both groups are very similar and are treated equally
control group
exposed to the same conditions as the experimental group, except for the manipulation of the independent variable
experimental group
exposed to the specific treatment or manipulation of the independent variable being investigated
equivalent groups
control and experimental groups must be equal
selection bias
the confound arising when there are differences between the comparison groups in a study
matching
identifying alike participants then randomly assigning them to different groups
pretesting
identify similar characteristics prior to matching
ceiling effect
a measure yields scores near the top limit of measurement for one or all groups
floor effect
dependent variable measurements yields scores near the lower limit
within-subjects design
an investigation where every participant receives every level of the independent variable at least once
pretest-posttest
one group of participants is tested twice using the same measurement tool, once before and once after the independent variable is manipulated
repeated-measures design
involved multiple measurements per participant
longitudinal design
within-subjects design where participants are tested multiple times except it looks for changes over a long period of time
advantages of within-subjects design
often requires fewer participants, takes less time, subject variables remain constant across the experimental conditions, error variance is reduced
why fewer participants?
because participants perform in each of the experimental conditions
lower error variance
less error variance = more powerful test of the IV’s effect
type 1 error
occurs when you incorrectly reject a true null hypothesis
type 2 error
occurs when you fail to reject a false null hypothesis
extraneous variable
anything that could influence the dependent variable
confounding variable
influences the dependent variable and also correlates with or causally affects the independent variable
operational definition
clear and specific description of how a variable will be measured, observed, or manipulated in a research study
reliability
the consistency with which the same results are obtained from the same test, instrument, or procedure
validity
the extent to which a measurement actually measures what it is supposed to
internal validity
the extent to which the design of an experiment ensures that the IV caused the change in DV
external validity
the generalizability of the study
population
all the organisms to which the researcher wishes to generalize their research results
sample
a subset of the population, a sample of the population used in the study
random sampling
randomly selecting participants from the population to be part of your sample
convenience sampling
participants are not randomly chosen, they just happen to be in the right place at the right time
hypotheses vs. prediction
hypothesis is more general than a prediction, prediction is derived from the hypothesis
null hypothesis
the prediction that nothing is different, there is no difference between groups
alternative hypothesis
the hypothesis upon which the researchers prediction is made- there is a difference in the groups
two-tailed hypothesis
researcher does not predict a specific direction of the difference between groups
one-tailed alternative hypothesis
researcher predicts the direction of the difference between groups
region of rejection
a special zone that helps researchers decide whether their data provides strong evidence to support a new idea or hypothesis
cons of within-subjects
demand characteristics, carryover effects, unrelated event impact
demand characteristics
when a participant derives information about what is expected
carryover effects
having a participant repeat some measure multiple times
history effect
the result of an event that occurs outside the experiment at the same time the independent variable is being changed
maturation effect
a change in performance due simply to the passage of time
regression towards the mean
people who scored really high or low tend to score closer to the mean the next time they take the test
counterbalancing
presenting experimental conditions to participants in different orders so that carryover effects are controlled
ABBA counterbalancing
participants get all conditions
block randomization
used when you have 3 or more conditions
random order with rotation
the experimental conditions are ordered randomly and the first participant receives this order and its changed for every participant
downside of matching participants
can be hard to find matching criteria for every participants, resource and time intensive, loss of generalizability
resource and time intensive
may require additional data collection, data cleaning, and careful consideration of matching variables
loss of generalizability
you could end up with groups that are so homogenous that they no longer represent the broader population accurately
cross-sectional study
data is collected from different groups of participants at a single point in time with the aim of comparing differences between these groups
pros of cross-sectional study
efficient, no participant drop out, good for population trends
cons of cross-sectional study
does not generate causal results, can’t track the changes over time, susceptible to cohort effects
temporal priority
the independent variable must come before the dependent variable is measured
control extraneous variables
there should be no confounds that could act as alternate explanations