Correlation
R Studio
Methods
Variable Types
Which Test?
100

This type of graph is used for correlational studies

What is a scatterplot?

100

lm(house_cost ~ size)

In this equation, the independent variable.

BONUS: What is predicting what?

What is size?


BONUS: size is predicting house_cost

100

Boxplots can reveal this property of data

What are outliers?

100

Satisfaction ratings (very, neutral, not)

What is ordinal?

100

This test can be used to assess normality.

What is the shapiro test?

200

True or false: Nominal variables can be used for correlations.

What is False?

200

the command: summary(dataset) 

will give this information.

What is mean, min, max, 1st/3rd Quartiles?

BONUS: # of NAs

200

Identify the dependent variable from the following vignette:

A researcher wants to study the effect of visual acuity on reading comprehension. Students are given vision and reading tests.

What is reading comprehension?

200

Mental health scores

What is interval?

200

This test uses a categorical variable of more than 2 types to predict a continuous variable.

BONUS: This test can be used to assess equal variances between these input variables.

What is an ANOVA?


BONUS: What is Levene's test

300

Finding the line that best fits a correlation is this type of statistical test.

What is linear regression?

300

Describe this model:

lm(happiness ~ money*job_satisfaction)

What is: a linear model predicting happiness from the interaction of money and job satisfaction?

300

This term describes the assumption that errors should be roughly equal size across the independent variable

What is homoskedasticity?

300

Education level (high school, BA, MA)

What is ordinal?

300

This test uses one or more continuous variables to predict a categorical outcome.

BONUS: name the glm family

What is logistic regression?

BONUS: binomial

400
An r-value of 0.6 explains this amount of variance.

What is 36%?

400

This predictor explains the most variance:

             Estimate Std. Error t value Pr(>|t|)    

(Intercept) 37.105505   2.110815  17.579  < 2e-16

hp          -0.031157   0.011436  -2.724  0.01097

wt          -3.800891   1.066191  -3.565  0.00133

disp        -0.000937   0.010350  -0.091  0.92851    

What is 'wt'? 

400

Residuals from a linear model can be tested for normality using this approach

What is a q-q plot?

[Also: histogram, also Shapiro test]

400

Species type (mammal, fish, reptile)

What is nominal?
400

This test uses one or more continuous variables to predict a continuous outcome.

BONUS: name the glm family

What is linear regression?

BONUS: Gaussian

500

True or false: Comparison of the strength of correlation between A & B and A & C is best assessed with multiple regression.

What is true?

500

This model explains the most variance:

A: R2=.67, AIC = 850, F=85.12

B: R2=.67, AIC = 950, F=83.80

C: R2=.76, AIC = 480, F=126.3

D: R2=.76, AIC = 520, F=116.7

What is C? (though also need more information)

500

This type of Chi-Square test compares distributions across categorical variables

BONUS: The other type of chi-square test

What is an Independence Chi-Square test?

BONUS: What is a "goodness of fit" Chi-Square

500

Temperature (in Celcius!)

What is Interval?

500

True or false: A t-test and ANOVA of two groups on a continuous variable will produce the same result.

What is true?

BUT you shouldn't do this because of the equal variance assumption of ANOVA, and ANOVA doesn't reveal direction of significance: TWO GROUPS = t-test