PSYC 202: FINAL 1
24 Cards in this Set
Front | Back |
---|---|
type I error
|
occurs when a researcher erroneously concludes that the null hypothesis is false and thus rejects it
|
type II error
|
the researcher may mistakenly fail to reject the null hypothesis when in fact it is false
|
Confound Variance
|
when a variable other than the independent variable differs between the groups
|
Treatment Variance
|
the portion of the variance in participants' scores that is due to the independent variable
|
Error Variance
|
the result of systematic differences among participants
-due to time, weather, mood, etc.
|
Problem of multiple t-tests
|
the probability of a type I error increases as tests are run more than once
probability of making a type I error on at least one of 10 t test is 4 out of 10 or .40. probability of making a type I error on multiple tests can be calculated by 1-(1-alpha)^c
|
Bonferroni Adjustment
|
researchers divide their desired alpha level such as .05 by the number of tests they plan to conduct so the probability of making a type I error on any test would be very low and not exceed the desired alpha level
|
ANOVA
|
analysis of variance, used when one wants to test differences among many means. used to analyze data from designs that involve more than two conditions simultaneously. can perform the work of lets say 10 tests at once with a .05 chance of a type I error
|
F-test
|
the ratio of the variance among conditions to the variance within conditions. the larger the between-groups variance relative to the within-groups variance, the larger the F value. F=MSbg/MSwg
testing the likelihood that the differences between the condition means are due to error var…
|
sum of squares
|
reflects the total amount of variability in a set of data
|
TOTAL sum of squares
|
calculated by subtracting the mean from each score squaring the differences, then adding them up
SS total = sigma (xi - x bar) ^2
|
sum of squares between groups
|
involves systematic variance that reflects the influence of the independent variable
subtract the grand mean from the mean of each group, square the differences, multiply each squared difference by the size of the group then sum across groups
|
sum of squares within groups
|
reflects error variance
calculate the sum of squares separately for each experimental group then add these sum of squares together
|
Mean Square between groups
|
when some of squares between groups is divided by its degrees of freedom. which is is the estimate of between-groups variance
|
mean square within groups
|
by dividing the sum of squares within groups variance by the within groups degrees of freedom
|
within-subjects ANOVA
|
used for multilevel and factorial within-subjects designs (when the same group of subjects serve in more than one treatment)
just as a paired t-test was used to analyze data form a within-subjects 2-group experi
|
main effects
|
if an AVOVA reveals a significant effect for an independent variable that has only two levels, no further stat tests are necessary
if there are more than two levels further tests are needed to interpret the finding
|
post hoc tests
|
follow up tests, identify which means differ significantly with the main effect that has more than two variables
|
MANOVA
|
tests differences between the means of two or more conditions on two or more dependent variables simultaneously
|
ANCOVA
|
-controls for/takes out extraneous variables (covariates) that you aren't interested in
ex: pre-test variables or extraneous variables
-reduces error by explaining some of the variance
-increases error
-greater experimental control
-gives F value
|
If the calculated value of F is less than the critical value...
|
the differences among the group means are no greater than we would expect on the basis of error variance alone, we fail to reject our null hypothesis and conclude that the IV does not have an effect
|
ANOVA in factorial designs
|
the systematic, between-groups portion of the variance can be broken down further into other main components to test for the presence of different main effects and interactions
with more than one IV, we can ask whether any systematic variance is related to each of the IV's as well as w…
|
interactions
|
if the effect of Variable A is different under one level of variable B than it is under another level of a variable B, an interaction is present
|
mixed-factorial ANOVA
|
used when doing both a within and between subjects analysis and you have 2 or more IVs
|