liveBook · Manning

Standard deviation

As explained in Timeless Algorithms, The Foundational Ideas

The square root of variance, measuring dispersion in the same units as the quantity described.

Standard deviation measures spread in the same units as the quantity being described. In regression, the error standard deviation estimates the typical vertical spread of observed scores around the fitted line under a normal-error model. section 3.4.3

Two measures of variability

For study hours and exam scores, maximum likelihood estimates an error standard deviation of 0.33 score points. In a beverage-fill example, the sample standard deviation is 0.30 ounces; with population standard deviation unknown, a one-sample t-test is used. section 3.4.3 section 4.2.1

σ^MLE=0.1098=0.33

The regression estimate is the square root of the maximum-likelihood error-variance estimate. section 3.4.3

How Timeless Algorithms, The Foundational Ideas covers it

How deep it goes
Works through maximum-likelihood error-variance and standard-deviation estimates, then interprets sample standard deviation in a t-test.
Examples
Uses study-hour and exam-score regression with estimated error standard deviation 0.33, and beverage fills with sample standard deviation 0.30 ounces.
What it assumes
Assumes a random sample and approximately normal population for the t-test application.
What you can do after
Evaluate/understand typical residual spread in a fitted line and the role of sample variability in selecting a t-test.
How it connects
Links variance and normal errors to regression, and sample standard deviation to the one-sample t-test.

Also explained in

Compare how 18 books explain standard deviation →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Timeless Algorithms, The Foundational Ideas. The section links point to the parts of the book it draws on.