Psyc 295: Chapter 7
58 Cards in this Set
Front | Back |
---|---|
Bivariate associations/correlations
|
an association that involves exactly two variables
|
Association claims are _______ not ________.
|
measured; manipulated
|
An association claim describes the relationship found between _____ __________ variables
|
two measured
|
After recording data, the next step in an association claim is to describe the relationship between the two variables using _________- and the _______ ______________ (_)
|
scatterplots; correlation coefficient (r)
|
Negative r means that the relationship is __________: high scores on one variable go with _____ scores on the other
|
negative
|
R has two qualities: _________ and __________
|
direction; strength
|
Direction refers to whether the association is...
|
positive, negative, or zero
|
Strength of r refers to how closely related the ____ _____________ are or how close r is to ____ or _____
|
two variables; 1; -1
|
r = 0.10
|
small, weak
|
r = 0.30 (or -0.30)
|
medium, moderate
|
r = 0.50 (or -0.50)
|
large, strong
|
When a variable is yes-or-no (i.e. employed? yes or no) is called a _________ variable
|
categorical
|
A categorical variable is when values..
|
fall in either one category or another
|
A quantitative variable is when both variables in an association are measured on ___________ ________
|
quantitative scales
|
Quantitative variables are best represented by/on a
|
scatterplot
|
When one or more of the variables is categorical, the researchers will plot the results as a _____________
|
bar graph
|
Bar graphs show
|
the average
|
In bar graphs you examine the difference between the __________ _______ to look for an association
|
average scores
|
When one or more variables in an association claim is categorical, researchers use 3 different statistics ___ ________, ______-_______ ___________, and/or ________ ____________
|
t test, point-biserial correlation, phi coefficient
|
The t test tests whether the difference between means is ________ _______
|
statistically significant
|
A point-biserial correlation is intended for evaluating the association between ___________ and _____________
|
one categorical variable; one quantitative variable
|
A phi coefficient is designed to evaluate the association between ____ _____________ _________
|
two categorical variables
|
An association claim is not characterized by a kind of _________ or _________
|
statistic; graph
|
An association claim is characterized by a study design in which both variables are
|
measured
|
At minimum, how many variables are there in an association claim?
|
2
|
What characteristic of a study's variables makes a study correlational?
|
both variables are measured
|
When do researchers typically use a bar graph, as opposed to a scatterplot to display correlational data?
|
when one of the variables is correlational
|
When interrogating association claims, the two most important validities to interrogate are ___________ validity and ___________ validity
|
construct; statistical
|
An association claim describes the relationship between two measured variables, so you should ask about the construct validity of ______ __________
|
each variable
|
Construct validity is how..?
|
well each variable is measured
|
When you ask about the _______ validity, you are asking about factors that might have affected the scatterplot, correlation coefficient (r), bar graph, or difference score that led to the claim
|
statistical
|
To interrogate statistical validity you need to consider the ______ ____ and _______ ________, _______ ______ or outliers, and whether a zero relationship might be ______________
|
effect size, statistical significance; subgroups; curvilinear
|
Effect size describes the __________ of the association
|
strength
|
Stronger effect size means that more ________ _________- of one variable can be made
|
accurate predictions
|
Effect sizes indicate how _________ the result is- that is a larger effect size is often considered ____ important than a small one
|
important; more
|
A small effect size can be ___________ (advil)
|
important
|
However, if not as extreme as light or death an effect size of 0.03 might be _________ in importance
|
negligible
|
Obtaining a correlation coefficient, r, establishes the direction and the strength ( _____ ______) of the relationship and determines whether the correlation is __________ ____________
|
effect size; statistically significant
|
Determining statistical significance is a process of ____________
|
inference
|
The sample's result usually ________ what is happening in the population, but not always
|
mimics
|
_______ _____ calculations helps researchers evaluate the probability that results really are zero
|
statistical significance
|
Statistical significance calculations provide a _________ _________
|
probability estimate (p or sig)
|
If the probability ( p ) associated with the result is very small—that is, less than ___%—we know that the result is very unlikely to have come from a “____-association” population.
|
5; zero
|
p = 0.05
|
high - nonsignificant
|
Statistical significance depends on _________ and __________
|
effect size; sample size
|
When p is small (under 5%) the result is __________ to have come from a zero association
|
unlikely
|
Spurious- the overall relationship is attributable only to __________ mean differences or ________- within the sample
|
systematic; subgroups
|
To analyze curvilinear relationship use a __________ model
|
quadratic
|
A quadratic model tests the correlation between one variable and the ________ of the other
|
square
|
Causal temptation is an automatic tendency to make _______ inferences from any claim we read
|
causal
|
To establish causation a study must fulfill three criteria:
|
covariation, temporal precedence, internal validity
|
Temporal precedence rule is sometimes called the _____________ problem because we dont know what variable came first
|
directionality
|
Covariance
|
a <---> b
|
Temporal precedence
|
a--->b or b---> a
|
Internal validity
|
a <--> c <--> b
|
To interrogate external validity of association claim what matters is the..
|
way the sample was selected
|
When a relationship between two variables changes depending on the level of a third variable that third variable is called a..
|
moderator
|
In correlation research, moderators can inform..
|
external validity
|