Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
49 Cards in this Set
- Front
- Back
sample
|
the people we collected our data from
|
|
population
|
larger group that we think our sample represents
|
|
outlier; what measure of central tendency do they most affect?
|
score that is far from most other scores; mean
|
|
z-score
|
a centered, standardized score
|
|
explained variance
|
When one or more predictor variables help us make a better guess at an outcome variable, we say that we have explained variance in the outcome variable
|
|
unexplained variance
|
The discrepancy between this guess and the actual outcome is the unexplained variance --- some of which might include systematic relationships we haven't thought of, and some of which is just random noise.
|
|
why is variance important with psychological research?
|
The more variance we can explain, the better we can claim to understand why people's responses differ, making it a useful goal for psychological research.
|
|
noise
|
random influences on a variable that are equally likely to occur in one direction as the other
|
|
bias
|
influences on a variable that are more likely to occur in one direction
|
|
Type I error
|
when the null hypothesis is actually True, but you conclude that it is False
|
|
how we explain variance, or the General Linear Model
|
1) single sets of groupings = t test, one way ANOVA
2) cross classified groupings = factorial ANOVA 3) continuous predictors = correlation, regression |
|
power
|
probability of finding a difference that does exist
|
|
type II error
|
failing to reject the null hypothesis when you should have
|
|
2 branches of statistics? examples?
|
descriptive: central tendency, variability, correlation, regression, z-scores
inferential: z-tests, t-tests, ANOVA, factorials, HYPOTHESIS TESTING |
|
score
|
value of a variable that applies to a particular observation
|
|
if two researchers make the same measurement but get very different values, we would say that the measure isn't _________.
|
reliable
|
|
manipulation check
|
making sure that whatever is supposed to affect the subjects' behavior actually affects the subjects' behavior. (an experimenter tries to affect participants' moods by watching sad movies. manipulation check would be useful)
|
|
nominal variable
|
no relationship between the categories, the variable categories are just names for things. example: race. also called categorical
|
|
ordinal variable
|
also called rank order. assigns numbers or ranks things in a particular order. example: shirt size S M L
|
|
equal interval variable
|
has to do with spectrums, dichotomies. an ordering that makes all separating intervals the same. example: scale of 1 to 10, date of birth (day month year)
|
|
ratio variable
|
says something about the presence or absence of something, there IS a zero point. example: height, age, temp in kelvin
|
|
variance (formula in words)
|
average squared deviations from the mean
|
|
Standard Error of the Mean, or Standard Error
|
the standard deviation of the distribution of sample means
|
|
what does the 95% confidence interval tell you?
|
the range of true values of the population mean that, when compared to your sample, would not lead to rejecting the null hypothesis
|
|
characteristics of the T Distribution
|
fatter tails than the Z distribution, more extreme scores because it estimates the population variance based on the sample scores. critical values on the t-distribution will always be greater than the z-distribution
|
|
one way ANOVA - the Within Groups variance
|
has the same value regardless of whether the null hypothesis is true
|
|
one way ANOVA - both the Between and Within Groups variances
|
they both include the effects of random variation, and they both require the assumption that the groups all have the same population variances
|
|
one way ANOVA - the Between Groups variance
|
estimate is done by treating sample means as scores taken from the distribution of sample means
|
|
alpha
|
probability of making a Type I error (probability of getting statistical significance if the null hypothesis is actually true)
|
|
beta
|
probability of making a Type II error (probability of not getting statistical significance if the research hypothesis is actually true)
|
|
Type I sums of squares (Two Way ANOVA)
|
sensitive to difference in cell group sizes - it focuses on weighted cell means. Marginal means reflect these differences
|
|
Type III sums of squares (Two Way ANOVA)
|
not sensitive to differences in cell group sizes - it focuses on unweighted cell means. relative cell sizes are ignored and the means tend to be roughly even.
|
|
Which test do you use to determine whether a correlation is significant?
|
t- test
|
|
when we ______ for another variable we are trying to see if our variable of interest still explains variance even when taking a secondary/nuisance variable into account
|
control
|
|
can you use correlation, regression, both or neither when: you want to test the relationship between a categorical IV and a categorical DV?
|
neither! this is a non-parametric test, the chi-squared test
|
|
can you use correlation, regression, both or neither when: you want to look at the strength of the relationship between 2 variables
|
Both correlation and regression can be used to look at the strength of the relationship between 2 variables
|
|
can you use correlation, regression, both or neither when: you want to compare multinested models to determine which best explain your DV
|
regression only
|
|
criterion variable
|
the variable that is predicted (usually Y)
|
|
when we ______ for another variable we are trying to see if our variable of interest still explains variance even when taking a secondary/nuisance variable into account
|
control
|
|
can you use correlation, regression, both or neither when: you want to test the relationship between a categorical IV and a categorical DV?
|
neither! this is a non-parametric test, the chi-squared test
|
|
can you use correlation, regression, both or neither when: you want to look at the strength of the relationship between 2 variables
|
Both correlation and regression can be used to look at the strength of the relationship between 2 variables
|
|
can you use correlation, regression, both or neither when: you want to compare multinested models to determine which best explain your DV
|
regression only
|
|
criterion variable
|
the variable that is predicted (usually Y)
|
|
predictor variable
|
variable that is used to predict scores of individuals on another variable (usually X)
|
|
general null and research hypotheses for the Chi-squared test of Independence and Goodness of Fit
|
There are 2 populations. Null= The two populations are the same. Research= The two populations are different.
|
|
Whats the difference between the Chi-squared tests for Independence and Goodness of Fit?
|
Independence involves 2 nominal variables, each with several categories. The Goodness of Fit test simply involves one nominal variable with several categories.
|
|
predictor variable
|
variable that is used to predict scores of individuals on another variable (usually X)
|
|
general null and research hypotheses for the Chi-squared test of Independence and Goodness of Fit
|
There are 2 populations. Null= The two populations are the same. Research= The two populations are different.
|
|
Whats the difference between the Chi-squared tests for Independence and Goodness of Fit?
|
Independence involves 2 nominal variables, each with several categories. The Goodness of Fit test simply involves one nominal variable with several categories.
|