Term
| Discrete variable definition |
|
Definition
|
|
Term
| Continuous variable definition |
|
Definition
|
|
Term
|
Definition
discrete
nO order
example: gender |
|
|
Term
|
Definition
discrete
can be ordered
ex: birth order |
|
|
Term
|
Definition
0 doesn’t mean 0 it is arbitrary
intervals of same size
ex: temperature |
|
|
Term
|
Definition
ordered values
ex: number of siblings |
|
|
Term
|
Definition
central tendency - mode, median, mean
variability |
|
|
Term
|
Definition
Hypothesis testing
sample testing to whole population |
|
|
Term
|
Definition
|
|
Term
|
Definition
| order least to greatest then the middle |
|
|
Term
|
Definition
|
|
Term
|
Definition
range
iqr
variance and standard deviation |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
outliers strongly affect correlations
strong v weak
neg v pos |
|
|
Term
|
Definition
| two unrelated groups compared |
|
|
Term
|
Definition
| compare 1 group to population |
|
|
Term
|
Definition
| compare 3 groups with each other |
|
|
Term
|
Definition
| compare nominal and ordinal data |
|
|
Term
| Frequency claim v. association claim v. Causal claim |
|
Definition
frequency - rate or degree of a single variable
ex: 2/5 or 57%
association - supported by correlational study, 2 variables associated
ex: people with higher income spend less time socializing
causal - supported by experimental study, not just related but causal
ex: music lessons enhance IQ |
|
|
Term
| Four types of validity for a claim |
|
Definition
construct
statistical
external
internal |
|
|
Term
|
Definition
is the operationalization good enough for the construct ?
Ex: is IQ a good test for the brain ? |
|
|
Term
|
Definition
extent statistical conclusions are accurate
accurate analysis avoid type errors |
|
|
Term
|
Definition
| can result be generalized to the population |
|
|
Term
|
Definition
| how well can study rule out confounds to make causal claims ? |
|
|
Term
|
Definition
| scientist manipulates this |
|
|
Term
|
Definition
|
|
Term
|
Definition
| flaws that make alternative explanations possible |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
| How correlational studies are different from true experiments |
|
Definition
| doesn’t determine CAUSATION |
|
|
Term
| How quasi experiments are different from true experiments |
|
Definition
| use non equivalent groups |
|
|
Term
|
Definition
Different participants get assigned different conditions
Need more participants
20 points of data = 40 participants
more noise
|
|
|
Term
|
Definition
each participant gets all conditions
less noise
Need less participants |
|
|
Term
|
Definition
| participants tested once before and after the IV is manipulated |
|
|
Term
|
Definition
| participants tested multiple times at different time points |
|
|
Term
|
Definition
participants are measured on dependent variable multiple times
must for within subject design
coukd happen in between |
|
|
Term
|
Definition
look at different age groups
between subject design |
|
|
Term
|
Definition
start the same
can do with random assignment |
|
|
Term
|
Definition
Used to get rid of subject selection bias
to run large sample sizes
randomly assign participants to different experimental conditions
ex: flip a coin |
|
|
Term
| Threats to internal validity |
|
Definition
these are confounds
all biases and effects |
|
|
Term
|
Definition
| biases and expectations a participants brings to the study themselves |
|
|
Term
|
Definition
| Participant will act differently upon knowing the thesis or purpose of experiment |
|
|
Term
|
Definition
| Effect of the experimenter expectations on the outcome of the study |
|
|
Term
|
Definition
| behavioral differences in study caused by the participation if different experimenters |
|
|
Term
|
Definition
a measurement device fails to measure in the same manner across various observations |
|
|
Term
|
Definition
when subjects withdraw from a study resulting in loss of data
Systematic - withdraw from some conditions more than others
Nonsystematic - withdraw due to reasons not related to the experiment |
|
|
Term
|
Definition
| change in performance due to event that occurred during the experiment |
|
|
Term
|
Definition
| change in performance in participant due to passage of time |
|
|
Term
|
Definition
| participants improve due to repeated to testing |
|
|
Term
| Regression towards the mean |
|
Definition
| 1st is extreme then the second result is more average |
|
|
Term
|
Definition
| being in condition A may effect condition B |
|
|
Term
|
Definition
| do better in condition B because you learned |
|
|
Term
|
Definition
| do worse on condition B because you get tired |
|
|
Term
|
Definition
| too easy so majority people do well |
|
|
Term
|
Definition
| too hard so majority people do poorly |
|
|
Term
Counterbalancing with within subject design
COMPLETE |
|
Definition
ABBA - experience condition ABBA
Downside is participant may predict order
Block randomization - experience blocks of conditions in random orders
Ex: If 3 conditions then 6 blocks |
|
|
Term
Counterbalancing with within subject design
INCOMPLETE |
|
Definition
random order w rotation - experience one order of all conditions then next experiences an order in which the first condition in the order is shifted to the back
Latin square design - each condition is presented in each ordinal position and is presented before and after each condition
*for even #s the conditions = number of order
*for odd #s the conditions = 2timesN |
|
|
Term
|
Definition
use of nonequivalent groups
may be able to support causal claims
no random assignment |
|
|
Term
|
Definition
measure score before and after an intervention for one setting
no control condition
no controlling for extraneous variable bc in natural setting
Ex: effect of sex ed on teen pregnancy
Ex: effect of prenatal program on infant mortality |
|
|
Term
| Nonequivalent groups design |
|
Definition
control group is COMPARABLE but not EQUIVALENT (bc did not randomly assign)
ex: residents in 2 similar locations
Ex: kids in 2 similar classrooms/schools |
|
|
Term
|
Definition
No control group
multiple observations for single group
extended version of pre-experimental
Ex: pre-pre-pre-INTERVENTION-post-post-post |
|
|
Term
| Multiple Time Series Design |
|
Definition
combination of time series and nonequivalent group
multiple observations of 2 nonequivalent groups |
|
|
