Name: Score:
93 Multiple choice questions
Definition 1 of 93
measures inter-item reliability by comparing each item in a scale to every other item individually
Correlational method
Cronbach's Alpha
Convergent validity
Stratified sampling
Term 2 of 93
Pretest sensitization
used to measure effect size with non-linear and ordinal data
make inferences about a population from a sample
ensures that people rate the same thing consistently
exposure to a pretest influences people's behavior
Definition 3 of 93
the amount of overlap in the variability of two variables (r^2)
Phi coefficient
Social desirability bias
Convenience sampling
Coefficient of determination
,Term 4 of 93
Name nine types of confounds
Z-scores convert any singular score into the number of SDs it is away from the mean. We
use them because they are standardized, which means we can compare different sample
means regardless of scale.
Likert scales (strongly agree-strongly disagree), semantic differential scales (bipolar
adjectives), and nominal scale (T/F or Y/N)
requires less subjects and eliminates individual differences which decreases the error term
and makes it more powerful and more likely to reject the null
biased assignment, differential attrition, pretest sensitization, history effects, maturation,
regression to the mean, instrumentation, diffusion, lack of standardization
Term 5 of 93
Interval data
a correlation coefficient used when there are two dichotomous variables
How far a score falls from the mean on average; the square root of variance
has equal intervals but does not have a true zero (i.e., degrees farenheit)
a design with multiple IVs. We use them because they are more realistic; in the real world
IVs do not occur in isolation
Term 6 of 93
Pearson correlation
used to measure effect size with non-linear and ordinal data
a design with multiple IVs. We use them because they are more realistic; in the real world
IVs do not occur in isolation
a coefficient determination in which we assume the relationship is linear and either ratio or
interval data
allows you to explore the influence of a third variable by looking at how r changes when
the third variable is factored in
,Term 7 of 93
When is the mean not the best measure of central tendency?
a study can be reliable and not valid, but it cannot be valid without also being reliable
mean, median, and mode
when there are outliers; they will skew the mean
make inferences about a population from a sample
Term 8 of 93
Quota sampling
everyone scoring high; test is too easy or not sensitive enough to measure correctly
convenience sampling, but the experimenter has certain quotas of how many people they
need from different demographics
participants are randomly chosen from a population and randomly assigned to a condition;
everyone has an equal chance of being chosen
consistency of measurements; getting the same answer every time
Term 9 of 93
Range
the distance between the lowest and highest scores (taking into account the upper and
lower real limits)
a coefficient determination in which we assume the relationship is linear and either ratio or
interval data
things that should correlate do (i.e., one person's scores on two different depression
measures)
allows you to enter in multiple predictors for one outcome variable, based on partial
correlation
, Term 10 of 93
Describe the relationship between reliability and validity
simple random, stratified, and cluster
a study can be reliable and not valid, but it cannot be valid without also being reliable
when an experimenter influences the outcome of a study because they are looking for a
certain effect
can only be used to answer a limited amount of research questions, and ethical concerns
also limit what can be observed
Term 11 of 93
Test-retest reliability
how consistent each item is a scale is, measures whether or not participants answer all items
in a consistent manner
allows you to explore the influence of a third variable by looking at how r changes when
the third variable is factored in
a design with multiple IVs. We use them because they are more realistic; in the real world
IVs do not occur in isolation
if a test is administered as different points in time and the scores correlate highly it has high
test-retest reliability
Definition 12 of 93
the manipulated variable
Inter-item reliability
Independent variable (IV)
Conceptual definition
Interrater reliability
93 Multiple choice questions
Definition 1 of 93
measures inter-item reliability by comparing each item in a scale to every other item individually
Correlational method
Cronbach's Alpha
Convergent validity
Stratified sampling
Term 2 of 93
Pretest sensitization
used to measure effect size with non-linear and ordinal data
make inferences about a population from a sample
ensures that people rate the same thing consistently
exposure to a pretest influences people's behavior
Definition 3 of 93
the amount of overlap in the variability of two variables (r^2)
Phi coefficient
Social desirability bias
Convenience sampling
Coefficient of determination
,Term 4 of 93
Name nine types of confounds
Z-scores convert any singular score into the number of SDs it is away from the mean. We
use them because they are standardized, which means we can compare different sample
means regardless of scale.
Likert scales (strongly agree-strongly disagree), semantic differential scales (bipolar
adjectives), and nominal scale (T/F or Y/N)
requires less subjects and eliminates individual differences which decreases the error term
and makes it more powerful and more likely to reject the null
biased assignment, differential attrition, pretest sensitization, history effects, maturation,
regression to the mean, instrumentation, diffusion, lack of standardization
Term 5 of 93
Interval data
a correlation coefficient used when there are two dichotomous variables
How far a score falls from the mean on average; the square root of variance
has equal intervals but does not have a true zero (i.e., degrees farenheit)
a design with multiple IVs. We use them because they are more realistic; in the real world
IVs do not occur in isolation
Term 6 of 93
Pearson correlation
used to measure effect size with non-linear and ordinal data
a design with multiple IVs. We use them because they are more realistic; in the real world
IVs do not occur in isolation
a coefficient determination in which we assume the relationship is linear and either ratio or
interval data
allows you to explore the influence of a third variable by looking at how r changes when
the third variable is factored in
,Term 7 of 93
When is the mean not the best measure of central tendency?
a study can be reliable and not valid, but it cannot be valid without also being reliable
mean, median, and mode
when there are outliers; they will skew the mean
make inferences about a population from a sample
Term 8 of 93
Quota sampling
everyone scoring high; test is too easy or not sensitive enough to measure correctly
convenience sampling, but the experimenter has certain quotas of how many people they
need from different demographics
participants are randomly chosen from a population and randomly assigned to a condition;
everyone has an equal chance of being chosen
consistency of measurements; getting the same answer every time
Term 9 of 93
Range
the distance between the lowest and highest scores (taking into account the upper and
lower real limits)
a coefficient determination in which we assume the relationship is linear and either ratio or
interval data
things that should correlate do (i.e., one person's scores on two different depression
measures)
allows you to enter in multiple predictors for one outcome variable, based on partial
correlation
, Term 10 of 93
Describe the relationship between reliability and validity
simple random, stratified, and cluster
a study can be reliable and not valid, but it cannot be valid without also being reliable
when an experimenter influences the outcome of a study because they are looking for a
certain effect
can only be used to answer a limited amount of research questions, and ethical concerns
also limit what can be observed
Term 11 of 93
Test-retest reliability
how consistent each item is a scale is, measures whether or not participants answer all items
in a consistent manner
allows you to explore the influence of a third variable by looking at how r changes when
the third variable is factored in
a design with multiple IVs. We use them because they are more realistic; in the real world
IVs do not occur in isolation
if a test is administered as different points in time and the scores correlate highly it has high
test-retest reliability
Definition 12 of 93
the manipulated variable
Inter-item reliability
Independent variable (IV)
Conceptual definition
Interrater reliability
