From PALS Flashcards
What is | in R
‘OR’
Three types of R variables
Numeric
Characters (also called ‘strings”
Logical (eg true/false)
A variable with more than one variable
Vector
Character you use to select variables out of a dataframe
$
Types of scores - untouched metric
= raw/observed
Scaled to the metric mean
= Deviation scores (always sums to zero)
Scaled to a specific mean, with specified amount of st dv per unit
= standardised score
Mean is set to 0 and 1 standard deviation of 1 per unit
= Z score
What are random variables
perpetually unknowable variables pertaining to the real world
Two types of variable
Continuous (decimal points make sense)
Discrete (or factor) (ps categorical variable = discrete variables
Standard Dev of a sampling distribution
= Standard error
Examples of observed test statistic
t-statistic, chi-squared, F-statistic
Theoretical probability distribution
gives you the equivalent of a sampling distribution without yo having to construct it
Are CIs a function of p values
Yep
What defines the CI
standard error + alpha criterion value
Which error types do alpha criterions control
Type 1 (false positives)
Effect size..
= strength of relationship
Pearson’s r is…
an effect size measure
Cramer’s V is…
an effect size measure
R squared is…
an effect size measure
Measuring systematic co-occurrance is done with..
association
Correlation coefficient is a standardised…
covariance (essentially covariance of Z-scores)
Range for Pearson’s r
-1 to 1
What is r in correlation
sample coefficient
What is rho in correlation
population parameter
When do you use Cramer’s V
association between categorical variables
Cramer’s V range
0-1?
What is the value of a 0% CI
The sample statistic itself
What does simple regression mean
1 x DV and 1 x IV
How you calculate St Dev
Sort of variance
Variance
Average variation
Variation
Total variation…
Covariance is standardised?
False
Linear regression model equation
Y(hat) = a + bX + e
What is OLS in regression
Ordinary Least Squares (smallest sum of squares)
R-squared
proportion of variance explained by the model
Under what conditions do you need to use an adjusted R squared
Small sample size
Large number of IVs
Assumptions for regression (4)
- Independence of observations
- Linearity
- Homoscedasticity
- Normality of residuals
How do you test Homoscedasticity
Breusch-Pagan (want a big p value)
What is homoscedasticity anyway
residuals distributed equally above and below zero
How can you tell if you’re looking at standardised multiple regression analysis results
there is no intercept reported (or it is 0)
