Study guide
9.2 One-Sample and Two-Sample t-Tests for Means
CSSC Certified Six Sigma Green Belt Exam
Track your progress as you read
Finished sections get a check mark.
1 Chapter 1: Six Sigma Overview & Exam Fundamentals
1.1 CSSC Certification Overview, Exam Structure & Body of Knowledge 1.2 History & Evolution of Six Sigma 1.3 Six Sigma Organizational Roles & Belts Hierarchy 1.4 The DMAIC Methodology Architecture vs. DMADV / DFSS 1.5 Other Process Improvement & Quality Methods, and When to Choose Six Sigma
2 Chapter 2: Quality Fundamentals & Defect Metrics
2.1 Critical to Quality (CTQ) Characteristics & VOC Translation 2.2 Defect Metrics: DPU, DPO, DPMO, & the 1.5-Sigma Shift 2.3 First Time Yield (FTY) & Rolled Throughput Yield (RTY) 2.4 Cost of Quality (COQ) & Cost of Poor Quality (COPQ)
3 Chapter 3: Lean Principles & Waste Elimination
3.1 Foundations of Lean & The Toyota Production System (TPS) 3.2 The Eight Wastes: CSSC's Seven Muda, TIMWOODS, and Muri/Mura 3.3 The 5S Visual Workplace Methodology 3.4 Value Stream Mapping (VSM), Cycle Time & Takt Time
4 Chapter 4: Project Selection & Team Management
4.1 Six Sigma Project Selection & Business Case Development 4.2 Project Charter Architecture 4.3 Stakeholder Management & Communication Planning 4.4 Team Dynamics & Tuckman's Team Development Stages 4.5 The 15-Point Project Viability Model & Process-Level Selection
5 Chapter 5: Define Phase & Process Mapping
5.1 High-Level Process Mapping with SIPOC 5.2 Flowcharts, Swimlane Diagrams & Process Value Analysis 5.3 Quality Function Deployment (QFD) & The House of Quality 5.4 Critical to Quality (CTQ) Flowdown & Operational Definitions 5.5 The CSSC Define Toolset: Ground Rules, In/Out of the Box, Is–Is Not & Tollgate Checklists
6 Chapter 6: Measure Phase & Measurement Systems
6.1 Data Types: Continuous (Variable) vs. Discrete (Attribute) Data 6.2 Data Collection Planning & Sampling Strategies 6.3 Measurement System Analysis (MSA) Foundations & Calibration 6.4 Variable & Attribute Gage R&R Studies
7 Chapter 7: Analyze Phase & Root Cause Analysis
7.1 Qualitative Root Cause Tools: 5 Whys & Cause-and-Effect (Fishbone) Diagrams 7.2 Failure Mode and Effects Analysis (FMEA) & RPN Calculation 7.3 Pareto Analysis & The 80/20 Rule 7.4 Graphical Data Exploration: Histograms, Box Plots, & Multi-Vari Charts 7.5 The Root Cause Verification Matrix
8 Chapter 8: Descriptive Statistics & Distributions
8.1 Measures of Central Tendency & Dispersion 8.2 The Normal Distribution & The Empirical Rule (68-95-99.7) 8.3 Standard Normal Distribution, Z-Scores & Probability Calculations 8.4 Common Non-Normal Distributions: Binomial, Poisson & Chi-Square
9 Chapter 9: Inferential Statistics & Hypothesis Testing
9.1 Fundamentals of Hypothesis Testing: Null vs. Alternative, Alpha & Beta Risks 9.2 One-Sample and Two-Sample t-Tests for Means 9.3 Paired t-Tests & Non-Parametric Alternatives 9.4 Analysis of Variance (One-Way & Two-Way ANOVA) & Chi-Square Contingency Tests 9.5 Sample Size Determination for Hypothesis Tests
10 Chapter 10: Correlation, Regression, & Experimentation
10.1 Linear Correlation & Pearson's Correlation Coefficient (r) 10.2 Simple Linear Regression & Coefficient of Determination (R^2) 10.3 Introduction to Design of Experiments (DOE) & Factorial Designs 10.4 Analyzing Experimental Results: Main Effects & Interaction Plots
11 Chapter 11: Improve Phase & Solution Implementation
11.1 Brainstorming, Creative Thinking & Solution Selection Matrices 11.2 Piloting Solutions & Conducting Risk Assessments 11.3 Implementation Planning & Change Management (ADKAR) 11.4 Mistake-Proofing (Poka-Yoke) & Standard Work
12 Chapter 12: Control Phase, SPC, & Process Capability
12.1 Statistical Process Control (SPC) & Shewhart Control Rules 12.2 Variable Control Charts: I-MR, X-bar & R, X-bar & S 12.3 Attribute Control Charts: p, np, c, and u Charts 12.4 Process Capability & Performance Indices: Cp, Cpk, Pp, Ppk 12.5 Communicating Statistical Results to Business Audiences
9.2 One-Sample and Two-Sample t-Tests for Means
Key Takeaways Explain This Topic AI
Student's t-tests are parametric inferential tests for continuous data when the population standard deviation σ is unknown and estimated by the sample standard deviation s, introducing heavier tails and degrees of freedom.
The One-Sample t-test compares a single sample mean (x̄) against a known target or historical specification (μ0) with degrees of freedom df = n - 1.
The Two-Sample Independent t-test compares the means of two mutually exclusive groups; an F-test or Levene's test must be conducted first to determine whether to use pooled variance (df = n1 + n2 - 2) or Welch's unpooled variance.
Pooled variance combines the sample variances weighted by their degrees of freedom when population variances are equal, whereas Welch's test adjusts degrees of freedom downward to protect against unequal variances.
If the 95% Confidence Interval for the difference between two means encompasses zero, fail to reject H0 (p > 0.05); if the entire interval excludes zero, reject H0 (p ≤ 0.05).
Last updated: September 2026
9.2 One-Sample and Two-Sample t-Tests for Means
Quick Summary: In the Analyze and Improve phases of DMAIC, continuous metrics—such as component dimensions, cycle times, pressure, fluid temperature, and chemical concentrations—must be statistically evaluated to validate root causes or verify that solutions shifted process averages. When the true population standard deviation σ \sigma σ is unknown, practitioners utilize Student's t t t -tests . The One-Sample t t t -test compares a single sample mean ( x ˉ \bar{x} x ˉ ) to a hypothesized target or customer specification ( μ 0 \mu_0 μ 0 ). The Two-Sample Independent t t t -test compares means from two mutually exclusive groups ( μ 1 \mu_1 μ 1 vs. μ 2 \mu_2 μ 2 ), requiring an preliminary F F F -test to determine whether to pool variances or deploy Welch's unpooled formulation.
Foundations of Parametric Means Testing
A parametric test makes specific mathematical assumptions regarding the parameters of the underlying probability distribution from which sample data is drawn. The three core parametric assumptions for t t t -tests are:
Continuous Metric Data : The dependent response variable ( Y Y Y ) is measured on a continuous interval or ratio scale (e.g., millimeters, seconds, degrees Celsius, foot-pounds).
Random and Independent Sampling : Data observations are collected randomly, and the measurement of one unit does not influence or depend upon another.
Normality : The data within each group is approximately normally distributed. By virtue of the Central Limit Theorem, t t t -tests are robust against moderate violations of normality when subgroup sample sizes satisfy n ≥ 30 n \ge 30 n ≥ 30 .
When these assumptions hold, the t t t -test family provides the most statistically powerful framework available for evaluating continuous means.
Why Student's t Instead of Z?
When evaluating a sample mean, if the true population standard deviation σ \sigma σ were known from exhaustive historical records, the test statistic would follow the Standard Normal Z Z Z -distribution : Z = x ˉ − μ 0 σ / n Z = \frac{\bar{x} - \mu_0}{\sigma / \sqrt{n}} Z = σ / n x ˉ − μ 0
However, in real-world industrial environments, the population standard deviation σ \sigma σ is virtually never known. Practitioners must estimate σ \sigma σ using the sample standard deviation ( s s s ) : s = ∑ i = 1 n ( x i − x ˉ ) 2 n − 1 s = \sqrt{\frac{\sum_{i=1}^n (x_i - \bar{x})^2}{n - 1}} s = n − 1 ∑ i = 1 n ( x i − x ˉ ) 2
Because s s s varies from sample to sample, substituting s s s for σ \sigma σ introduces an additional source of random sampling error. In 1908, William Sealy Gosset (publishing under the pen name "Student" while working at the Guinness Brewery in Dublin) derived the Student's t t t -distribution to account for this extra variability. Comparison of Z vs. Student's t-Distribution
- ┌───┐ ◀── Z Distribution (Taller peak, thin tails)
- ┌──┘ └──┐
- ┌─┘ ┌───┐ └─┐
- ┌─┘ ┌──┘ └──┐ └─┐ ◀── t Distribution (Lower peak, fat tails)
- ───┴───┴─────────┴───┴───
- 3 -2 -1 0 1 2 3
The t t t -distribution is symmetric and bell-shaped with a mean of 0, but features heavier (fatter) tails and a lower center peak than the standard normal curve.
The heavier tails mean that extreme values occur more frequently, requiring larger critical values to achieve statistical significance at a given α \alpha α .
Governed by degrees of freedom ( d f df df ) ; as sample size increases ( d f → ∞ df \to \infty df → ∞ ), the t t t -distribution converges toward the standard normal Z Z Z -distribution.
- The One-Sample t-Test
- Purpose & Operational Context
The One-Sample t t t -test evaluates whether the arithmetic mean of a single continuous sample ( x ˉ \bar{x} x ˉ ) is statistically significantly different from a known benchmark, engineering specification, historical standard, or contractual target ( μ 0 \mu_0 μ 0 ), when σ \sigma σ is unknown.
Test Statistic & Formula t calc = x ˉ − μ 0 s / n t_{\text{calc}} = \frac{\bar{x} - \mu_0}{s / \sqrt{n}} t calc = s / n x ˉ − μ 0
Where:
- x ˉ = \bar{x} = x ˉ = Sample mean
- μ 0 = \mu_0 = μ 0 = Hypothesized population target
- s = s = s = Sample standard deviation
- n = n = n = Sample size
- Standard Error: S E = s n \text{Standard Error: } SE = \frac{s}{\sqrt{n}} Standard Error: SE = n s
D e g r e e s o f F r e e d o m : d f = n − 1 \mathbf{Degrees\ of\ Freedom:}\ df = n - 1 Degrees of Freedom : df = n − 1
Step-by-Step Worked Example: One-Sample t t t -Test
Scenario: A precision CNC milling center produces stainless steel shafts with an engineering target diameter of μ 0 = 50.00 mm \mu_0 = 50.00\text{ mm} μ 0 = 50.00 mm . A quality engineer samples n = 16 n = 16 n = 16 consecutive shafts to verify machine calibration. The sample yields:
- Sample mean: x ˉ = 50.08 mm \bar{x} = 50.08\text{ mm} x ˉ = 50.08 mm
- Sample standard deviation: s = 0.12 mm s = 0.12\text{ mm} s = 0.12 mm
- Significance level: α = 0.05 \alpha = 0.05 α = 0.05 (two-tailed)
- Step 1: State the Hypotheses
- H 0 : μ = 50.00 mm H_0: \mu = 50.00\text{ mm} H 0 : μ = 50.00 mm (Machine is on target)
- H a : μ ≠ 50.00 mm H_a: \mu \ne 50.00\text{ mm} H a : μ = 50.00 mm (Machine has drifted off target)
- Step 2: Determine Degrees of Freedom and Critical t t t -Value
- d f = n − 1 = 16 − 1 = 15 df = n - 1 = 16 - 1 = 15 df = n − 1 = 16 − 1 = 15
Looking up a two-tailed t t t -distribution table at α = 0.05 \alpha = 0.05 α = 0.05 ( 0.025 0.025 0.025 per tail) with d f = 15 df = 15 df = 15 : t crit = ± 2.131 t_{\text{crit}} = \pm 2.131 t crit = ± 2.131
Step 3: Calculate the Test Statistic ( t calc t_{\text{calc}} t calc )
Standard Error: S E = s n = 0.12 16 = 0.12 4 = 0.030 mm SE = \frac{s}{\sqrt{n}} = \frac{0.12}{\sqrt{16}} = \frac{0.12}{4} = 0.030\text{ mm} SE = n s = 16 0.12 = 4 0.12 = 0.030 mm
t calc = 50.08 − 50.00 0.030 = + 0.08 0.030 = + 2.667 t_{\text{calc}} = \frac{50.08 - 50.00}{0.030} = \frac{+0.08}{0.030} = +2.667 t calc = 0.030 50.08 − 50.00 = 0.030 + 0.08 = + 2.667
Step 4: Formulate the Decision and Confidence Interval
Decision Rule: Reject H 0 H_0 H 0 if ∣ t calc ∣ ≥ 2.131 |t_{\text{calc}}| \ge 2.131 ∣ t calc ∣ ≥ 2.131 .
Since ∣ + 2.667 ∣ > 2.131 |+2.667| > 2.131 ∣ + 2.667∣ > 2.131 (corresponding to software p = 0.0177 < 0.05 p = 0.0177 < 0.05 p = 0.0177 < 0.05 ), we Reject H 0 H_0 H 0 .
95% Confidence Interval for μ \mu μ : C I = x ˉ ± t α / 2 , d f × S E = 50.08 ± ( 2.131 × 0.030 ) = 50.08 ± 0.0639 = [ 50.016 mm , 50.144 mm ] CI = \bar{x} \pm t_{\alpha/2, df} \times SE = 50.08 \pm (2.131 \times 0.030) = 50.08 \pm 0.0639 = [50.016\text{ mm}, 50.144\text{ mm}] C I = x ˉ ± t α /2 , df × SE = 50.08 ± ( 2.131 × 0.030 ) = 50.08 ± 0.0639 = [ 50.016 mm , 50.144 mm ]
Notice that the 95% CI [ 50.016 , 50.144 ] [50.016, 50.144] [ 50.016 , 50.144 ] excludes the target value of 50.00 mm 50.00\text{ mm} 50.00 mm , confirming rejection of H 0 H_0 H 0 .
Conclusion: There is statistically significant evidence at α = 0.05 \alpha = 0.05 α = 0.05 that the CNC machine has drifted off target; the mean shaft diameter ( 50.08 mm 50.08\text{ mm} 50.08 mm ) is significantly greater than 50.00 mm 50.00\text{ mm} 50.00 mm .
- The Two-Sample Independent t-Test
- Purpose & Architecture
The Two-Sample Independent t t t -test compares the means of two mutually exclusive, unrelated groups to determine whether they originate from populations with identical means:
Comparing yield strength between Supplier A and Supplier B.
Comparing cycle time between Day Shift and Night Shift.
Comparing wear rates between Machine 1 and Machine 2. Two-Sample Independent t-Test Architecture
- GROUP 1 (e.g., Supplier A) GROUP 2 (e.g., Supplier B)
- Independent Random Sample Independent Random Sample
- n1, x̄1, s1 n2, x̄2, s2
- │ │
- └─────────────────────┬──────────────────────┘
- ▼
- Compare Means: μ1 vs. μ2
- H0: μ1 - μ2 = 0 (Equal)
- Ha: μ1 - μ2 ≠ 0 (Different)
- Critical Pre-requisite: The F-Test for Equal Variances
Before running an independent two-sample t t t -test, the Green Belt must verify whether the two population variances are equal ( σ 1 2 = σ 2 2 \sigma_1^2 = \sigma_2^2 σ 1 2 = σ 2 2 ). This is tested using an F F F -test of Variances (or Levene's test): F calc = s 1 2 s 2 2 ( Placing the larger sample variance in the numerator ) F_{\text{calc}} = \frac{s_1^2}{s_2^2} \quad (\text{Placing the larger sample variance in the numerator}) F calc = s 2 2 s 1 2 ( Placing the larger sample variance in the numerator )
If Variances are Equal ( p ≥ 0.05 p \ge 0.05 p ≥ 0.05 ): Use the Pooled Two-Sample t t t -test .
If Variances are Unequal ( p < 0.05 p < 0.05 p < 0.05 ): Use Welch's Unpooled t t t -test .
Case A: Pooled Two-Sample t t t -Test (Equal Variances Assumed)
When variances are homogeneous, the two sample variances are weighted and combined into a single pooled variance estimate ( s p 2 s_p^2 s p 2 ): s p 2 = ( n 1 − 1 ) s 1 2 + ( n 2 − 1 ) s 2 2 n 1 + n 2 − 2 s_p^2 = \frac{(n_1 - 1)s_1^2 + (n_2 - 1)s_2^2}{n_1 + n_2 - 2} s p 2 = n 1 + n 2 − 2 ( n 1 − 1 ) s 1 2 + ( n 2 − 1 ) s 2 2 S E pooled = s p 1 n 1 + 1 n 2 SE_{\text{pooled}} = s_p \sqrt{\frac{1}{n_1} + \frac{1}{n_2}} S E pooled = s p n 1 1 + n 2 1 t calc = ( x ˉ 1 − x ˉ 2 ) − ( μ 1 − μ 2 ) 0 s p 1 n 1 + 1 n 2 t_{\text{calc}} = \frac{(\bar{x}_1 - \bar{x}_2) - (\mu_1 - \mu_2)_0}{s_p \sqrt{\frac{1}{n_1} + \frac{1}{n_2}}} t calc = s p n 1 1 + n 2 1 ( x ˉ 1 − x ˉ 2 ) − ( μ 1 − μ 2 ) 0 D e g r e e s o f F r e e d o m : d f = n 1 + n 2 − 2 \mathbf{Degrees\ of\ Freedom:}\ df = n_1 + n_2 - 2 Degrees of Freedom : df = n 1 + n 2 − 2
Exam Trap: Notice why d f = n 1 + n 2 − 2 df = n_1 + n_2 - 2 df = n 1 + n 2 − 2 . Two degrees of freedom are lost because two independent sample statistics ( x ˉ 1 \bar{x}_1 x ˉ 1 and x ˉ 2 \bar{x}_2 x ˉ 2 ) are estimated from the data.
Case B: Welch's Unpooled t t t -Test (Unequal Variances)
When variances are unequal, pooling is mathematically invalid. Welch's test uses individual sample standard deviations: S E unpooled = s 1 2 n 1 + s 2 2 n 2 SE_{\text{unpooled}} = \sqrt{\frac{s_1^2}{n_1} + \frac{s_2^2}{n_2}} S E unpooled = n 1 s 1 2 + n 2 s 2 2 t calc = x ˉ 1 − x ˉ 2 s 1 2 n 1 + s 2 2 n 2 t_{\text{calc}} = \frac{\bar{x}_1 - \bar{x}_2}{\sqrt{\frac{s_1^2}{n_1} + \frac{s_2^2}{n_2}}} t calc = n 1 s 1 2 + n 2 s 2 2 x ˉ 1 − x ˉ 2
Degrees of freedom are calculated using the Welch-Satterthwaite approximation : d f = ( s 1 2 n 1 + s 2 2 n 2 ) 2 ( s 1 2 / n 1 ) 2 n 1 − 1 + ( s 2 2 / n 2 ) 2 n 2 − 1 df = \frac{\left( \frac{s_1^2}{n_1} + \frac{s_2^2}{n_2} \right)^2}{\frac{(s_1^2 / n_1)^2}{n_1 - 1} + \frac{(s_2^2 / n_2)^2}{n_2 - 1}} df = n 1 − 1 ( s 1 2 / n 1 ) 2 + n 2 − 1 ( s 2 2 / n 2 ) 2 ( n 1 s 1 2 + n 2 s 2 2 ) 2
This formula adjusts d f df df downward into a non-integer value to compensate for variance disparity, protecting against inflated Type I error.
Step-by-Step Worked Example: Pooled Two-Sample t t t -Test
Scenario: A Green Belt investigates whether a new quenching oil formulation improves the hardness of carbon steel gears. Two independent production batches are quenched using Oil A (standard) and Oil B (new formulation). Hardness is measured in Rockwell C (HRC):
Group Sample Size ( n n n ) Sample Mean ( x ˉ \bar{x} x ˉ ) Sample Std Dev ( s s s )
Oil A (Standard) n 1 = 12 n_1 = 12 n 1 = 12 x ˉ 1 = 58.2 HRC \bar{x}_1 = 58.2\text{ HRC} x ˉ 1 = 58.2 HRC s 1 = 1.4 HRC s_1 = 1.4\text{ HRC} s 1 = 1.4 HRC
Oil B (New) n 2 = 14 n_2 = 14 n 2 = 14 x ˉ 2 = 60.1 HRC \bar{x}_2 = 60.1\text{ HRC} x ˉ 2 = 60.1 HRC s 2 = 1.6 HRC s_2 = 1.6\text{ HRC} s 2 = 1.6 HRC
Test whether the new quenching oil significantly increases gear hardness at α = 0.05 \alpha = 0.05 α = 0.05 .
- Step 1: Formulate Hypotheses
- H 0 : μ 2 − μ 1 ≤ 0 H_0: \mu_2 - \mu_1 \le 0 H 0 : μ 2 − μ 1 ≤ 0 (Oil B does not increase hardness)
H a : μ 2 − μ 1 > 0 H_a: \mu_2 - \mu_1 > 0 H a : μ 2 − μ 1 > 0 (Oil B significantly increases hardness; one-tailed upper test)
Step 2: Verify Equality of Variances
F calc = s 2 2 s 1 2 = ( 1.6 ) 2 ( 1.4 ) 2 = 2.56 1.96 = 1.306 F_{\text{calc}} = \frac{s_2^2}{s_1^2} = \frac{(1.6)^2}{(1.4)^2} = \frac{2.56}{1.96} = 1.306 F calc = s 1 2 s 2 2 = ( 1.4 ) 2 ( 1.6 ) 2 = 1.96 2.56 = 1.306
Degrees of freedom: d f 1 = 14 − 1 = 13 df_1 = 14 - 1 = 13 d f 1 = 14 − 1 = 13 , d f 2 = 12 − 1 = 11 df_2 = 12 - 1 = 11 d f 2 = 12 − 1 = 11 .
Critical F 0.05 , 13 , 11 = 2.79 F_{0.05, 13, 11} = 2.79 F 0.05 , 13 , 11 = 2.79 . Since 1.306 < 2.79 1.306 < 2.79 1.306 < 2.79 ( p > 0.10 p > 0.10 p > 0.10 ), variances are homogeneous. Proceed with pooled t t t -test.
Step 3: Compute Pooled Variance ( s p 2 s_p^2 s p 2 ) and Standard Error s p 2 = ( 12 − 1 ) ( 1.4 ) 2 + ( 14 − 1 ) ( 1.6 ) 2 12 + 14 − 2 = ( 11 ) ( 1.96 ) + ( 13 ) ( 2.56 ) 24 = 21.56 + 33.28 24 = 54.84 24 = 2.285 s_p^2 = \frac{(12 - 1)(1.4)^2 + (14 - 1)(1.6)^2}{12 + 14 - 2} = \frac{(11)(1.96) + (13)(2.56)}{24} = \frac{21.56 + 33.28}{24} = \frac{54.84}{24} = 2.285 s p 2 = 12 + 14 − 2 ( 12 − 1 ) ( 1.4 ) 2 + ( 14 − 1 ) ( 1.6 ) 2 = 24 ( 11 ) ( 1.96 ) + ( 13 ) ( 2.56 ) = 24 21.56 + 33.28 = 24 54.84 = 2.285 s p = 2.285 = 1.5116 HRC s_p = \sqrt{2.285} = 1.5116\text{ HRC} s p = 2.285 = 1.5116 HRC S E diff = s p 1 n 1 + 1 n 2 = 1.5116 1 12 + 1 14 = 1.5116 0.0833 + 0.0714 = 1.5116 0.1548 = 1.5116 × 0.3934 = 0.5947 HRC SE_{\text{diff}} = s_p \sqrt{\frac{1}{n_1} + \frac{1}{n_2}} = 1.5116 \sqrt{\frac{1}{12} + \frac{1}{14}} = 1.5116 \sqrt{0.0833 + 0.0714} = 1.5116 \sqrt{0.1548} = 1.5116 \times 0.3934 = 0.5947\text{ HRC} S E diff = s p n 1 1 + n 2 1 = 1.5116 12 1 + 14 1 = 1.5116 0.0833 + 0.0714 = 1.5116 0.1548 = 1.5116 × 0.3934 = 0.5947 HRC
Step 4: Calculate Test Statistic ( t calc t_{\text{calc}} t calc ) t calc = ( x ˉ 2 − x ˉ 1 ) − 0 S E diff = 60.1 − 58.2 0.5947 = 1.90 0.5947 = + 3.195 t_{\text{calc}} = \frac{(\bar{x}_2 - \bar{x}_1) - 0}{SE_{\text{diff}}} = \frac{60.1 - 58.2}{0.5947} = \frac{1.90}{0.5947} = +3.195 t calc = S E diff ( x ˉ 2 − x ˉ 1 ) − 0 = 0.5947 60.1 − 58.2 = 0.5947 1.90 = + 3.195
Step 5: Make Statistical Decision
Degrees of freedom: d f = n 1 + n 2 − 2 = 12 + 14 − 2 = 24 df = n_1 + n_2 - 2 = 12 + 14 - 2 = 24 df = n 1 + n 2 − 2 = 12 + 14 − 2 = 24
Critical value for one-tailed test at α = 0.05 \alpha = 0.05 α = 0.05 with d f = 24 df = 24 df = 24 : t crit = + 1.711 t_{\text{crit}} = +1.711 t crit = + 1.711
Since t calc = + 3.195 > + 1.711 t_{\text{calc}} = +3.195 > +1.711 t calc = + 3.195 > + 1.711 ( p = 0.0019 < 0.05 p = 0.0019 < 0.05 p = 0.0019 < 0.05 ), Reject H 0 H_0 H 0 .
Conclusion: Quenching Oil B delivers a statistically significant increase in gear hardness of approximately 1.90 HRC 1.90\text{ HRC} 1.90 HRC ( p = 0.0019 p = 0.0019 p = 0.0019 ).
Interpreting Statistical Output & Confidence Intervals for Differences
Statistical packages output both test statistics ( t t t , p p p ) and an associated Confidence Interval for the Difference : Interpreting the Confidence Interval for (μ1 - μ2)
- INTERVAL CONTAINS ZERO: [-2.5, +1.8] INTERVAL EXCLUDES ZERO: [+1.4, +5.2]
- 2.5 0 +1.8 0 +1.4 +5.2
- ├───┼──────┼────────┼──┤ ├───┼──────┼───────┼──┤
- ▲ ▲ ▲ ▲ ▲ ▲
- LCL Zero UCL Zero LCL UCL
- ───────────────────────────────────── ─────────────────────────────────────
- FAIL TO REJECT H0 (p > 0.05) REJECT H0 (p ≤ 0.05)
Difference may be zero (No effect) Statistically significant difference!
If the 95% CI contains 0 (e.g., [ − 1.40 , + 2.80 ] [-1.40, +2.80] [ − 1.40 , + 2.80 ] ): Zero is a plausible value for the true difference. Therefore, p > 0.05 p > 0.05 p > 0.05 ; Fail to Reject H 0 H_0 H 0 .
If the 95% CI excludes 0 (e.g., [ + 1.20 , + 4.50 ] [+1.20, +4.50] [ + 1.20 , + 4.50 ] or [ − 6.10 , − 1.80 ] [-6.10, -1.80] [ − 6.10 , − 1.80 ] ): Zero is ruled out as a plausible value. Therefore, p ≤ 0.05 p \le 0.05 p ≤ 0.05 ; Reject H 0 H_0 H 0 .
Summary Selection Matrix for Continuous Means
Test Name Subgroups Data Relationship Population σ \sigma σ Status Degrees of Freedom ( d f df df ) Common Six Sigma Use Case
One-Sample Z Z Z -test 1 sample Single sample vs specification Known (Rare) ∞ \infty ∞ ( Z Z Z -table) Baseline standard deviation known from long history.
One-Sample t t t -test 1 sample Single sample vs specification Unknown (Estimated by s s s ) d f = n − 1 df = n - 1 df = n − 1 Verifying if current production meets specification target.
Two-Sample Pooled t t t -test 2 samples Independent / Unrelated Unknown (Equal variances, p ≥ 0.05 p \ge 0.05 p ≥ 0.05 ) d f = n 1 + n 2 − 2 df = n_1 + n_2 - 2 df = n 1 + n 2 − 2 Comparing mean cycle times of Line 1 vs Line 2.
Welch's Two-Sample t t t -test 2 samples Independent / Unrelated Unknown (Unequal variances, p < 0.05 p < 0.05 p < 0.05 ) Welch-Satterthwaite approx Comparing suppliers with vastly different process variability.
Critical CSSC Exam Traps
Trap 1: Skipping the Equal Variance Check — Running a pooled two-sample t t t -test without verifying variance equality via an F F F -test risks invalid conclusions if variances differ significantly.
Trap 2: Degrees of Freedom Confusion — For a pooled two-sample t t t -test, degrees of freedom is d f = n 1 + n 2 − 2 df = n_1 + n_2 - 2 df = n 1 + n 2 − 2 , NOT n 1 + n 2 − 1 n_1 + n_2 - 1 n 1 + n 2 − 1 .
Trap 3: One-Sided vs. Two-Sided Output — Most statistical software packages report two-tailed p p p -values by default. If your hypothesis is directional (one-tailed) and the sample shifted in the hypothesized direction, the one-tailed p p p -value is p two-tailed / 2 p_{\text{two-tailed}} / 2 p two-tailed /2 .
Trap 4: Misinterpreting Difference CIs — If a 95% CI for ( μ 1 − μ 2 ) (\mu_1 - \mu_2) ( μ 1 − μ 2 ) is [ − 0.4 , + 3.2 ] [-0.4, +3.2] [ − 0.4 , + 3.2 ] , the test cannot reject H 0 H_0 H 0 at α = 0.05 \alpha = 0.05 α = 0.05 because zero is contained in the interval.