03. Relationships

Lesson 4 03 Relationships V2

Definition

When your problem statement requires understanding the relationship between 1+features

_ Example_

Problem statement reads: How do the demographic characteristics of a FIFA player affect his market value?

Which test will give the appropriate visualization?

  1. When both your variables are numeric, Pearson correlation or Regressions are a good choice

  2. When both your variables are categorical, Chi-square test or any other non-parametric test can be used.

    Here are some resources on non-parametric tests:

  1. When one variable is categorical variable and the other a numerical variable, ANOVA test is a good choice

    Here are some resources on ANOVA tests:
  • Wikipedia
  • Blog from minitab.com
    • When you need to run an ANOVA test on multiple dependent variables, statisticians use MANOVAs, but outside academic research, these are rarely used in the industry. Nonetheless, if you are curious, here is a link to explore this topic more in-depth:

Next Concept