Section A.3: Bivariate analysis

Question 1

What is bivariate analysis?

Answer 1

Bivariate analysis is the form of data analysis which compares two variables at a time. It is useful for identifying patterns, trends, correlation/relationships, between two variables. It is a form of inferential statistics.

Question 2

Bivariate analysis techniques (4)

Answer 2

Graphs,
Correlation analysis, regression analysis, contingency table analysis

Question 3

Analysis for numerical v. numerical data (3)

Answer 3

Line graph, Scatterplot, Correlogram

Question 4

Analysis for categorical v. categorical data (3)

Answer 4

Stacked column chart, chi squared test, mixed graphs

Question 5

Analysis for numerical v. categorical data (6)

Answer 5

Side by side boxplots, violin plots, grouped bar plots, line graph, z-test, t-test

Question 6

What is correlation analysis?

Answer 6

Correlation analysis is a bivariate analysis technique which determines the existence of and measures the strength of a relationship between two continuous variables. Measured using a correlation coefficient, ranging from -1 to 1.

Question 7

What is regression analysis?

Answer 7

Regression analysis is a bivariate analysis technique which involves generating a predictive model, allowing you to predict the value of a dependent (response) variable based on the values of one (predictor) variable.

NOTE: Simple linear regression is bivariate in nature, multiple linear regression is multivariate.

Question 8

What is a correlogram?

Answer 8

A correlogram is a visualisation used in bivariate analysis. It is a matrix of correlation coefficients that show the strength of the relationship between numerical variables

Question 9

Scatterplot

Answer 9

A scatterplot is a graph in which the values of two continuous variables are plotted along the x-axis and y-axis, useful for identifying and demonstrating a relationship between two variables.