Effect size
An effect size is a quantitative measure of the magnitude of a difference, association, relationship, or other observed phenomenon, independent of statistical significance.
Statistics Slam Dunk
An effect size is a quantitative measure of the magnitude of a difference or relationship between groups or variables.
- How it is used
- Quantifies practical magnitude alongside significance tests when comparing NBA groups or relationships.
- How deep it goes
- Defines standardized mean differences, qualitative ratings, test selection, and record-count limits without deriving inferential formulas.
- Examples
- Uses home-versus-road fouls and free throws plus Gini coefficients for winning and championship teams.
- What you can do after
- Apply it via R packages to quantify and classify differences while retaining companion significance tests.
Effective Data Analysis
An effect size is a quantitative measure of the magnitude of a difference or relationship, independently of sample size.
- How it is used
- Separates the magnitude of a group difference from statistical significance when assessing whether an analytics finding merits action.
- How deep it goes
- Names Cohen’s d and explains its interpretive role, then uses assumed effects in power analysis without deriving an effect-size formula.
- Examples
- Uses student test scores and conversion rates to illustrate meaningful differences, plus simulated normal samples to explore power-analysis inputs.
- What you can do after
- Apply it via statsmodels power analysis to plan sample size, and evaluate significant results against a domain-specific minimum difference.
Related concepts
- Cohen's dmeasured by · Cohen's d is a standardized effect-size measure expressing the difference between two group means in standard-deviation units, commonly using the pooled standard deviation.
- causationequivalent to · Causation is a relationship in which one event, condition, or factor produces or influences a change in another.
- hypothesis testingcontrasts with · Hypothesis testing is a formal statistical procedure that uses observed data to evaluate evidence about a specified hypothesis, often a null hypothesis.
- t-testcontrasts with · A t-test is a statistical hypothesis test that uses a t-distributed statistic to assess a mean or difference in means under specified assumptions.
- Cramer's Vincludes · Cramer's V is a normalized, directionless measure of association strength between categorical variables in a contingency table, ranging from 0 to 1.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.