Quiz 3 - Inferential Statistics & Correlation + Regression

Statistical probability

the odds that what we observed in the sample did occur because of error

in inferential statistics, we are always testing the _____ hypothesis

null

inferential statistics

uses sample data to evaluate the credibility of a hypothesis about a population

Type I Error

rejecting a true hypothesis

Type II Error

accepting a false hypothesis

which type of error is more of a problem?

Type I

alpha

the probability of making a type I error

what is the numeric value of alpha

0.05

if p value is greater than alpha, we _____ the null hypothesis

accept

if p value is equal to or less than 0.05, we _____ the null hypothesis

reject

statistical significance

the observed group difference unlikely to be due to error (p-value)

practical significance

looks at whether the difference is large enough to be of value in a practical sense

practical significance involves _______ ______, while statistical significance involves _____-_____.

effect size; p-value

Cohen's Effect Size Guidelines:

2, 5, 8

significant differences (i.e., reject the null hypothesis) means that the differences in group means are ____ likely due to _____ _______.

not; sampling error

true or false: the problem is that statistically significant differences can be found even with very small differences if the sample size is large enough.

true

Is this a practically significant difference?

A study of 14,000 children ages 6-17 showed a highly significant (p < 0.001) and effect size d = 0.11 between height and IQ.

No; effect size is small

Testing for Mean Differences: If you have one IV and 2 groups, you will do a....

t-test

Testing for Mean Differences: If you have one IV and more than 2 groups, you will do a....

one-way ANOVA

Testing for Mean Differences: if you have two or more IVs, you will do a....

factorial ANOVA

Testing for Mean Differences: if the groups are NOT related (i.e. between-subjects), you will do a.... (2 answers)

1. independent t-test

2. between-measure ANOVA

Testing for Mean Differences: if the groups ARE related (i.e., within-subjects), you will do a... (2 answers)

1. dependent (paired) t-test

2. repeated measures ANOVA

true or false: the first step is to do a test of statistical significance, and the second step is to compute descriptive statistics

false

what is MANOVA?

multivariate analysis of variance; more than one DV

for t-test, what should the sample size be?

more than 30

F-test is used with...

ANOVA

what does f-test tell us?

not all groups are equal

Post-hoc test is used with...

ANOVA

what does post-hoc test do?

tells us where the difference is between groups

ANCOVA

tests whether there is a significant difference between groups after controlling for variance explained by a covariate

what is a covariate?

a continuous variable that correlates with the dependent variable

what is the pictorial representation of correlation?

scatter plot

Correlation: associations between 2 scores involves what 3 factors?

1. direction

2. form

3. strength

What is positive correlation?

as one variable increases, so does the other

What is negative correlation?

as one variable increases, the other decreases

What is a correlation coefficient?

indicates the degree to which 2 variables are related

What is the range for the correlation coefficient (R)?

-1 to +1

Correlation Strength (R):

0.00-____ weak or none

0.2-______ weak

0.4-_____ moderate

0.6-______ strong

0.8-______ very strong

0.2, 0.4, 0.6, 0.8, 1.0

what three factors influence correlations?

1. sample size

2. outliers

3. restricted range

In correlation, with very _____ samples, even ______ correlations can be significant

large, small

Outliers can make correlation look ____ or ____ than it actually is.

stronger, weaker

Correlations tend to be weaker when data represents _____ range.

narrow

Regression is an extension of _______

correlation

What are the two types of regression models?

simple and multiple

Regression is used to explain the impact of changes in an ________ variable on the _________ variable

independent, dependent

Using regression, you can predict the value of a _______ variable based on the value of at least one _______ variable

dependent, independent

What is the regression equation?

y = A + Bx

y= dependent variable

A = y-intercept

B= the slope of the line

x = independent variable

Effect size is also depicted as what in a regression table?

R-squared

What does r-squared mean?

the amount of variance in the dependent variable that can be predicted from the independent variable

In a regression table, the constant represents...

the constant coefficient (y-intercept)

multiple regression equation

Y = m1X1 + m2X2 + b

Regression issues: with too ____ a sample, you lack ______ _______ to generalize your results to other samples/the whole population.

small, statistical power

In multiple regression, the more variables you use in your model, the greater your risk of Type _____ Error

II

Sample Size calculation: Tabachnick and Fidell (1996) equation

N > 50 + 8m

N= appropriate sample size

m= number of IVs