Links

F. Husson website





Book Exploratory Multivariate Analysis Using R

Outline

Introduction

Principal Component Analysis

Correspondence Analysis

Multiple Correspondence Analysis

Clustering

Multiple Factor Analysis

To conclude

Forum

PCA quiz : studying individuals and variables

For each question, tick the correct answer or answers.

Q1) In principal component analysis,
variables with a large mean have more importance in the analysis
variables should always be standardized if they come in different units
variables should never be standardized if they have the same units
standardization gives the same importance to all variables

Q2) The individuals plot:
two individuals quite far apart on the first axis have very different values for many variables
individuals close the center of gravity of the cloud have average values for many variables
two individuals superposed on the principal plane (1-2 axes) have the same values for all variables
two individuals superposed on the principal plane (1-2 axes) have similar values for the variables that are well-projected onto this plane.

Q3) For the variables,
the graph with the correlation circle allows us to visualize the correlation matrix between variables
the graph with the correlation circle helps us to interpret the individuals plot
the cos of the angle between the arrows representing variables on the projection plane is equal to the correlation coefficient between the variables
the cos of the angle between the arrows representing variables in the global space is equal to the correlation coefficient between the variables

Q4) The first factor axis (the first dimension)
is the one that best-separates the points of the individuals
represents the variable that is most related to the set of variables (in the R2 sense)
is the one that contains the most information

Q5) Interpreting the variables plot

Using the following representation of the variables, examine the correlations between pairs of variables.
Graphe des variables

Q5.1) The correlation between V1 and V5
is close to -1
is close to -0.5
is close to 0
is close to 0.5
is close to 1
cannot be discovered by looking at the plot - it can vary between -1 and 1

Q5.2) The correlation between V1 and V7
is close to -1
is close to -0.5
is close to 0
is close to 0.5
is close to 1
cannot be discovered by looking at the plot - it can vary between -1 and 1

Q5.3) The correlation between V2 and V4
is close to -1
is close to -0.5
is close to 0
is close to 0.5
is close to 1
cannot be discovered by looking at the plot - it can vary between -1 and 1

Q5.4) The correlation between V1 and V3
is close to -1
is close to -0.5
is close to 0
is close to 0.5
is close to 1
cannot be discovered by looking at the plot - it can vary between -1 and 1

Q5.5) The correlation between V7 and V8
is close to -1
is close to -0.5
is close to 0
is close to 0.5
is close to 1
cannot be discovered by looking at the plot - it can vary between -1 and 1

Q5.6) The correlation between V5 and V6
is close to -1
is close to -0.5
is close to 0
is close to 0.5
is close to 1
cannot be discovered by looking at the plot - it can vary between -1 and 1

Score =
Correct answers: