Which data relationship algorithm compares two numerical values?

Linear regression is the statistical algorithm used within TIBCO Spotfire's Data Relationships tool to model and evaluate the relationship between two continuous, numerical variables. It fits a straight line to the data points to quantify the strength and direction of the association between an independent variable and a dependent variable. The tool calculates an R-squared value to indicate how well the model explains the variability of the response data around its mean, which is a key measure for comparing two numerical columns.

A. Chi-square: This test is used to determine if there is a significant association between two categorical variables, not numerical ones.

B. Anova: Analysis of Variance (ANOVA) is used to compare the means of a numerical variable across different groups defined by a categorical variable.

D. Kruskal-Wallis: This is a non-parametric alternative to ANOVA, used for comparing a numerical variable across groups when the assumptions for ANOVA are not met.

1. TIBCO Spotfire® Analyst User's Guide

Version 12.0 LTS. In the section "Details on Data Relationships

" the guide explicitly states the statistical methods used for different column types:

"Comparison of two continuous columns: A linear regression model is fitted to the data. The R-squared value is used as a measure of how well the model fits the data."

"Comparison of two categorical columns: A chi-square test is performed..."

"Comparison of a continuous and a categorical column: An ANOVA test is performed..."

"If the assumptions for the ANOVA test are not met

a Kruskal-Wallis test is performed instead."

(This information is consistently found in user guides for various Spotfire versions

including 10.x

11.x

and 12.x).