Deck 2: Statistical Analytics With R Part 1

Question 1

Which R package was used for the Levene's test in ANOVA and for hypothesis testing?

A) library(car)
B) library(caret)
C) library(cluster)
D) library(corrplot)

Answer 1

A
Others were:
library(caret) # Used for logistic regression
library(cluster) # Used for cluster analysis
library(corrplot) # Used for visualizing correlation matrices

Question 2

Which R package is NOT used in conjunction with crosstabulation?

A) library(expss)
B) library(gmodels)
C) library(plyr)
D) library(vcd)

Answer 2

C
The "plyr" package which is used, for example, to apply a statistical function to each subset of a data frame. The "expss" and "gmodels" packages contain crosstabulation commands (the cro and CrossTable commands respectively). The "vcd" package supports measures of association for crosstabulation through its assocstats command. There are also other crosstabulation programs, such as xtabs from the "stats" package in R's System Library.

Question 3

Given that a data frame called "survey" is loaded and has a column labeled "marital", what is wrong with this command?: table(marital)

A) The table command requires two variables, not just one.
B) Should be TABLE(survey$marital)
C) Should be table:survey$marital
D) Should be table(survey$marital)

Answer 3

D
is correct. A is incorrect because table does support 1-way tables, which are frequency distributions. B is incorrect because R is case-sensitive in most contexts and will not recognize the table command if capitalized. C is incorrect because commands like table require the argument in parentheses, whereas the double-colon is used in relation to packages.

Question 4

What command in R is used most frequently to implement ordinary least squares (OLS) regression? Fill In: __________________________

Question 5

Give a short answer: What is the "base::" part in the command below and what does it do? base::table(survey$marital,survey$sex)

Question 6

Cronbach's alpha, a leading measure of reliability, is produced by which R package?

A) reliability
B) psych
C) alpha
D) alpha

Question 7

Which was NOT used for cluster analysis in the text?

A) clusplot from the "cluster" package
B) cluster.stats from the "fpc" package
C) NbClust from the "NbClust" package.
D) All of the above were used

Question 8

The ANOVA family has four main variants: ANOVA, ANCOVA, MANOVA, and MANCOVA. In the text, which command(s) was/were used to implement them?

A) anova
B) anova and manova
C) aov and manova
D) glm

Question 9

In logistic regression, what is a synonym for a "confusion table", created by the confusionMatrix command?

A) a regression table
B) a logistic regression table
C) a table of logistic coefficients
D) a classification table

Question 10

The "processR" package implements which type of statistical analysis?

A) mediation and moderation analysis, as per the work of Andrew Hayes
B) correlation, as per the work of Francis Galton
C) correlation, as per the work of Karl Pearson
D) processR is not a statistical package