unit 3 exam Flashcards
what is factorial design?
-a design in which there are two or more independent variables
if you have an experiment with 2 independent variables (ex. noise and temperature) , how many possible combinations can you have?
4
how do you describe a factorial design?
-include each variable in the description
-indicate each variable by a number representing the number of levels of the variable
what do the number of levels of the variable represent?
number of possibilities
How do you describe this factorial design example:
2 (noise: low vs. high) x 2 (temperature: cold vs. hot)
-2 variables: noise and temperature
-noise has 2 levels
-temperature has 2 levels
How do you describe this factorial design example:
3 (time of day: morning vs. afternoon vs.evening) x 2 (precipitation type: rain vs.
snow)
-2 variables: time of day and precipitation type
-Time of day has 3 levels
-Precipitation type has 2 levels
How do you describe this factorial design example:
3 (time of day: morning vs. afternoon vs. evening) x 2 (precipitation type: rain vs.
snow) x 2 (quantity of precipitation: light vs.heavy)
-3 variables: time of day and precipitation type and quantity of precipitation
-Time of day has 3 levels
-Precipitation type has 2 levels
-Quantity of precipitation has 2 levels
the total number of conditions is the product of what?
the number of levels of each variable (ex. 2 x 2 design = 4 total conditions)
when would the number of groups equal the number of conditions?
when all variables are between-subjects
when would the number of groups be less than the number of conditions?
when at least one variable is within-subjects
the total number of groups is the product of what?
the number of levels of each between subjects variable (ex. 3 between subjects x 2 between subjects = 6 groups)
what are main effects?
on average, individuals in one condition differ on dependent variable from individuals in other conditions
what are interaction effects?
the effect of one independent variable depends on the level of another independent variable
what is intersectionality?
where the meaning of one facet of identity depends on other facets of identity
what is a statistical interaction?
the effect of one independent variable depends on the level of another independent variable
What did the Thiem (2019) study examine?
-examined how quickly guns are identified immediately after viewing images of faces that varied in gender and race
-examined how quickly tool are recognized
-tested black and non-black participants
what did the results of the Thiem (2019) study show?
no interactions between gender and race
what is a contingency table?
a table showing the distribution of one variable in rows and another in columns, used to study association b/w the two categorical variables
what does Phi measure?
-measures association between two categorial variables
what does the range of Phi represent?
-can range from 0 (completely independent) to 1.0 (perfect association)
what is the chi square test?
-statistical test used to evaluate whether you can reject the null hypothesis that two categorical variables are independent
-used to test statistical significance of phi
what is moderation?
the correlation between two
(ordinal/interval/ratio) variables depends on
the level of a third variable
what are the two types of moderation?
-for whom
-under what conditions
If the correlation between stress and eating is
strongly positive among individuals high in
neuroticism, whereas the correlation between stress
and eating is weakly positive among individuals low in neuroticism. What is the moderator?
neuroticism moderates the relationship between stress and eating
what is mediation?
-The causal link between two variables is mediated by a third variable where X at Time 1 has a causal influence on Y at Time 2,
which in turn has a causal
influence on Z at Time 3
What is this an example of?
Time 1: High test scores
Time 2: Amount of sleep
Time 3: Mood
mediation, where amount of sleep is the mediating factor
If someone experiences child abuse at one point in time, then difficulties with relationships at another point in time, and then experiences depression at time 3; what is the mediating factor?
relationship difficulties
what do correlations measure?
measure how much variation in variable A can be systematically accounted for by variation in variable B (venn diagram)
what is a partial correlation?
-measures the relationship between two variables (A and B) while eliminating the influence of a third variable (C).
Partial Correlation: The more the third variable overlaps with the first two variables, the ____ the difference between the ____ and the partial correlation.
greater, correlation
why is partial correlation used?
Used to render less plausible the possibility that the reason two variables are associated is that they are both caused by a third variable
what is multiple regression used for?
-used to see how well two (or more) predictors combined can predict an outcome
-used to see the degree to which each predictor predicts the outcome when taking the other predictors into account
Why would we conduct multiple regression in this example?
Both conflict with parents and conflict with peers predict childhood depression.
-can see how well these two predictors, when combined predict childhood depression
-how well conflict with parents predicts depression when taking into account conflict with peers and vice versa
How well A and B combined predict C depends on _________.
overlap between A and B
Less overlap between A and B leads to _____ prediction of C.
greater
Greater overlap between A and B leads to _____ prediction of C.
less
what is internal consistency?
a measure of how reliable an instrument measures something
what is internal consistency measured by? what is the range?
-Cronbach’s alpha
-range: 0-1.0 (higher is better)
what two things is internal consistency influenced by?
1.Magnitude of correlations among items (the greater the average correlation among items the higher the internal consistency)
2.Number of items (the larger the number of items the higher the internal consistency)
what is a Test-Retest correlation?
-correlation between scores at time 1 and scores at time 2
what is the purpose of a Test-Retest correlation?
used to measure temporal stability of scores
If what you are trying to measure is supposed to be very stable, test-retest correlation should be very ____.
high
If what you are trying to measure is supposed to be only moderately stable, test-retest correlation should be only _____.
modest
If measuring height in adults, test-retest correlation should be very close to 1 (even if measured one year apart). If test-retest
correlation is .5, it suggests the measure of height is _____.
not valid because height should be pretty stable
If measuring daily level of stress, test-retest correlation (measured one year apart) should not be close to 1. If test-retest correlation (across one year) is .9, it suggests the measure of daily stress is ____.
not valid because levels of stress should not be extremely stable
The quality of research depends on how specific a conclusion you can draw. The ____ specific, the better the quality.
more
Specificity is influenced by what two things?
1.the number and choice of dependent variables.
2.The number and choice of groups (in ex post facto designs)
when trying to increase specificity through measuring multiple dependent variables, what is important?
measuring multiple dependent variables that are similar in some ways by different in other ways
what is the goal of measuring multiple dependent variables?
-to examine whether there is something specific or special about each dependent variable
-examine whether other variables you are examining are associated similarly or differently with the different dependent variables
if you compare OCD with non-psychiatric controls, measuring verbal memory and the results are consistent with the hypothesis that verbal memory is associated with OCD, what remains an issue?
-the hypothesis is not specific enough and there remains plausible alternative explanations
In an ex post facto design, what is most important for determining specificity?
group selection
what are some problems researchers have with adding dependent variables?
-resources
-time (the larger the number of dependent variables, the longer the experiment)
what are some problems researchers have with adding control groups?
-resources
-the larger the number of groups, the larger the number of between group comparisons (increasing the likelihood of unreliable results)
