Making Sense of Data: A Practical Guide to Exploratory Data Analysis and Data Mining

5.5 SUMMARY

Central Tendency

Mode: Most common value

Median: Middle value

Variation

Range: high–low

Quartiles: Q1 (25%), Q2 (50%), Q3 (75%)

Confidence Levels

Hypothesis Test

Specify null (e.g. H₀: μ = μ₀) and alternative hypothesis (e.g. H_a: μ > μ₀)

Select significance level (e.g. α = 0.05)

Compute test statistics (t- or z-)

Determine critical value for t or z using α/2 for two sided tests

Reject the null hypothesis if test statistic fall in the “reject H₀” region

Comparing Groups

When comparing more than two groups, use:

Chi-square test for categorical data

One-way analysis of variance test for continuous.

Comparing Variables