Study guide
Correlation & Linear Regression
ASQ Certified Six Sigma Green Belt CSSGB
Track your progress as you read
Finished sections get a check mark.
1 Introduction & Exam Overview
CSSGB Exam Facts & Logistics How to Study for an Open-Book ASQ Exam
2 Six Sigma Foundations
Value of Six Sigma & Quality Leaders Organizational Goals & Six Sigma Projects Organizational Drivers, KPIs & Scorecards
3 Lean Principles in the Organization
Lean Concepts Value Stream Mapping & Waste
4 Design for Six Sigma
DFSS Road Maps: DMADV & IDOV FMEA Basics & Risk Priority Number Design FMEA vs Process FMEA
5 Define: Project Identification
Project Selection & DMAIC Fit Process Elements & Boundaries Benchmarking Types SIPOC & Process Inputs/Outputs
6 Stakeholders & Voice of the Customer
Owners & Stakeholders Customer Identification Customer Data Collection CTQ, QFD & Kano
7 Project Management Core
Agile & Top-Down Project Methods Project Charter & Problem Statements Project Scope Tools Primary & Consequential Metrics
8 Planning, Risk & Closure
WBS, Gantt, CPM, PERT & Tollgates Project Documentation Project Risk Analysis & Management Project Closure & Lessons Learned
9 Management Tools & Business Results
Eight Management & Planning Tools DPU, RTY, CoPQ, DPMO & Sigma Level Organizational Communication
10 Team Dynamics & Performance
Team Stages & Negative Dynamics Six Sigma Roles & RACI Brainstorming, NGT & Multivoting Team Communication
11 Measure: Process & Probability
Process Maps & Documentation Basic Probability Concepts Central Limit Theorem
12 Distributions & Data Collection
Statistical Distributions Data Types & Measurement Scales Sampling & Data Collection Plans Descriptive Statistics
13 Graphical Methods & MSA
Graphical Methods Measurement System Analysis & Gage R&R Bias, Linearity, Stability & P/T
14 Process & Performance Capability
Process Performance vs Specifications Process Capability Studies Cp, Cpk, Pp, Ppk & Cpm Short-Term vs Long-Term Capability & Sigma Shift
15 Analyze: Exploratory Data Analysis
Multi-Vari Studies Correlation & Linear Regression
16 Hypothesis Testing
Hypothesis Testing Basics Tests for Means, Variances & Proportions
17 Gap & Root Cause Analysis
Gap Analysis Root Cause Analysis Tools
18 Design of Experiments
DoE Basic Terms Main Effects & Interaction Plots
19 Implementation Planning & Lean Tools
Implementation Planning & Pilots Waste Elimination Tools Cycle-Time Reduction & SMED Kaizen & Kaizen Blitz
20 Statistical Process Control
SPC Basics: Common vs Special Cause Rational Subgrouping Control Chart Selection & Construction
21 Sustaining Improvements & Lean Control
Control Plans Document Control, Training & Audits Plan-Do-Check-Act (PDCA) TPM & Visual Factory
Correlation & Linear Regression
Key Takeaways Explain This Topic AI
Correlation measures the strength and direction of a linear association between two continuous variables; it does not prove that X causes Y.
Pearson’s correlation coefficient r ranges from −1 to +1; |r| near 1 means strong linear association, r near 0 means little linear association (nonlinear patterns can still exist).
Simple linear regression fits ŷ = b₀ + b₁x; b₁ is the estimated slope (change in Y per unit X) and b₀ is the intercept (fitted Y when x = 0, which may be extrapolation).
Use the fitted line for estimation within the observed X range; treat prediction outside that range as extrapolation and treat causal claims as separate from the fit.
A small p-value for the slope (or for r) indicates the linear association is unlikely under a null of no linear relationship—not that the model is perfect or that X is the true root cause alone.
Last updated: July 2026
Correlation & Linear Regression
Quick Answer: Correlation quantifies how strongly two continuous variables move together in a linear pattern (Pearson’s r from −1 to +1). Simple linear regression fits ŷ = b₀ + b₁x for estimation and prediction. A significant p-value supports a real linear association in the data; it does not by itself prove causation .
Correlation vs. Causation
In Analyze you often plot a candidate X against Y. A strong r is evidence of association , not proof that changing X will change Y.
Association can appear because X truly drives Y, Y drives X, a lurking Z drives both, the sample is coincidence, or both variables share a time trend. Green Belt rule: use correlation and regression to prioritize and quantify ; confirm mechanism with process knowledge, DOE, or controlled interventions before locking Improve actions.
Scatter Plots First
Before computing r or fitting a line: plot Y vs X; look for linearity, outliers, clusters, and curvature. A curved cloud can yield r near zero despite a strong nonlinear link, or a moderate r with a poor linear fit. Scatter plots prevent misleading coefficients.
Pearson Correlation Coefficient
For paired observations ( x i , y i ) (x_i, y_i) ( x i , y i ) , i = 1 , … , n i = 1, \ldots, n i = 1 , … , n : r = ∑ i = 1 n ( x i − x ˉ ) ( y i − y ˉ ) ∑ i = 1 n ( x i − x ˉ ) 2 ∑ i = 1 n ( y i − y ˉ ) 2 r = \frac{\sum_{i=1}^{n}(x_i - \bar{x})(y_i - \bar{y})}{\sqrt{\sum_{i=1}^{n}(x_i - \bar{x})^2 \, \sum_{i=1}^{n}(y_i - \bar{y})^2}} r = ∑ i = 1 n ( x i − x ˉ ) 2 ∑ i = 1 n ( y i − y ˉ ) 2 ∑ i = 1 n ( x i − x ˉ ) ( y i − y ˉ )
Computational form: r = n ∑ x i y i − ( ∑ x i ) ( ∑ y i ) [ n ∑ x i 2 − ( ∑ x i ) 2 ] [ n ∑ y i 2 − ( ∑ y i ) 2 ] r = \frac{n\sum x_i y_i - (\sum x_i)(\sum y_i)}{\sqrt{\bigl[n\sum x_i^2 - (\sum x_i)^2\bigr]\bigl[n\sum y_i^2 - (\sum y_i)^2\bigr]}} r = [ n ∑ x i 2 − ( ∑ x i ) 2 ] [ n ∑ y i 2 − ( ∑ y i ) 2 ] n ∑ x i y i − ( ∑ x i ) ( ∑ y i )
- r (approx.) Linear association
- ±1 Perfect linear
- ±0.7 to ±0.9 Strong
- ±0.4 to ±0.6 Moderate
- near 0 Little linear association
Thresholds are guidelines. Pair r with the plot and with practical significance.
Properties: r is unitless and invariant to linear rescaling of X or Y; it measures linear association only; strength is about the pair, while regression slope Y on X is not the same as X on Y .
Simple Linear Regression Y = β 0 + β 1 X + ε , y ^ = b 0 + b 1 x Y = \beta_0 + \beta_1 X + \varepsilon, \quad \hat{y} = b_0 + b_1 x Y = β 0 + β 1 X + ε , y ^ = b 0 + b 1 x b 1 = ∑ ( x i − x ˉ ) ( y i − y ˉ ) ∑ ( x i − x ˉ ) 2 = r s y s x , b 0 = y ˉ − b 1 x ˉ b_1 = \frac{\sum (x_i - \bar{x})(y_i - \bar{y})}{\sum (x_i - \bar{x})^2} = r \, \frac{s_y}{s_x}, \quad b_0 = \bar{y} - b_1 \bar{x} b 1 = ∑ ( x i − x ˉ ) 2 ∑ ( x i − x ˉ ) ( y i − y ˉ ) = r s x s y , b 0 = y ˉ − b 1 x ˉ
b₁: estimated average change in Y per one-unit increase in X.
b₀: fitted Y when x = 0 (often extrapolation if 0 is outside the operating range).
For one predictor, R² = r² — the fraction of sample Y variation explained by the linear fit. R² = 0.64 means 64% of observed Y variation in the sample is accounted for by the line, not that the project is 64% finished.
Estimation, Prediction, and Significance
Estimation: for x in the observed range, y ^ = b 0 + b 1 x \hat{y} = b_0 + b_1 x y ^ = b 0 + b 1 x estimates the mean response at that x.
Individual prediction: same point estimate is often used, but uncertainty for one future point is larger than for the mean.
Extrapolation: x far outside the data range is risky.
For the slope (or testing population correlation ρ = 0 under standard assumptions):
H₀: β₁ = 0 (no linear relationship)
H₁: β₁ ≠ 0 (two-sided), or one-sided when the project cares about one direction t = b 1 S E ( b 1 ) ( df = n − 2 ) t = \frac{b_1}{SE(b_1)} \quad (\text{df} = n - 2) t = SE ( b 1 ) b 1 ( df = n − 2 )
The p-value is the probability under H₀ of a result at least as extreme as observed. At α = 0.05, p < 0.05 usually means the linear association is statistically significant . A tiny significant slope can still be practically useless; p ≥ 0.05 means insufficient evidence of a linear link—not proof that X never matters nonlinearly.
Worked Numerical Example
Oven temperature offset X (°C) vs bond strength Y (MPa), n = 5 (small for arithmetic; real projects use larger n):
- Run x y
- 1 0 10
- 2 2 12
- 3 4 13
- 4 6 15
- 5 8 16
- x ˉ = 4 \bar{x} = 4 x ˉ = 4 , y ˉ = 13.2 \bar{y} = 13.2 y ˉ = 13.2
- ∑ ( x i − x ˉ ) 2 = 40 \sum (x_i - \bar{x})^2 = 40 ∑ ( x i − x ˉ ) 2 = 40
- ∑ ( y i − y ˉ ) 2 = 22.80 \sum (y_i - \bar{y})^2 = 22.80 ∑ ( y i − y ˉ ) 2 = 22.80
∑ ( x i − x ˉ ) ( y i − y ˉ ) = 30.0 \sum (x_i - \bar{x})(y_i - \bar{y}) = 30.0 ∑ ( x i − x ˉ ) ( y i − y ˉ ) = 30.0 r = 30.0 40 × 22.80 = 30.0 912 ≈ 0.993 r = \frac{30.0}{\sqrt{40 \times 22.80}} = \frac{30.0}{\sqrt{912}} \approx 0.993 r = 40 × 22.80 30.0 = 912 30.0 ≈ 0.993 b 1 = 30.0 40 = 0.75 , b 0 = 13.2 − 0.75 × 4 = 10.2 b_1 = \frac{30.0}{40} = 0.75, \quad b_0 = 13.2 - 0.75 \times 4 = 10.2 b 1 = 40 30.0 = 0.75 , b 0 = 13.2 − 0.75 × 4 = 10.2 y ^ = 10.2 + 0.75 x \hat{y} = 10.2 + 0.75x y ^ = 10.2 + 0.75 x
Each 1 °C increase in offset is associated with about 0.75 MPa higher strength in the fitted range. At x = 5 (in range): y ^ = 10.2 + 3.75 = 13.95 \hat{y} = 10.2 + 3.75 = 13.95 y ^ = 10.2 + 3.75 = 13.95 MPa. R² ≈ 0.987 in this nearly linear toy series—real process R² is usually lower. With r ≈ 0.993 and n = 5, the t-test of H₀: β₁ = 0 would give a small p-value under standard assumptions; the Green Belt still asks whether the link is causal, practical, and stable on the full line.
Reading p-Values Without Overclaiming
- Result Legitimate conclusion Overclaim to avoid
- p < 0.05 for slope Evidence of a nonzero linear slope (given assumptions) “X is the only root cause”
Large |r| and small p Strong, significant sample linear association “Improving X will definitely improve Y by b₁ forever”
- p ≥ 0.05 Not enough evidence of a linear link “X has no effect of any kind”
- High R² Line explains much of sample Y variation “The process is solved”
Always check residual plots when software is available: curved residuals mean the linear model is the wrong shape; fan-shaped residuals mean nonconstant variance.
Assumptions and DMAIC Fit
Least squares and usual p-values assume roughly linear mean structure, independent errors, roughly constant variance, and approximately normal errors for exact t procedures. Outliers can dominate b₁ and r. Multiple predictors need multiple regression (often deeper than core Green Belt practice); omitted variables still bias causal stories even when r looks clean.
Multi-vari and stratification locate major noise families.
Correlation/regression quantify continuous X–Y links.
Hypothesis tests and DOE confirm effects before Improve.
Expect CSSGB items that distinguish correlation from causation; interpret r, b₁, b₀, and R²; compute a simple slope or fitted value; and read a slope p-value without overclaiming process truth.