Six Sigma Green Belt Practice Test

Six Sigma Green Belt Analyze Phase Practice Questions

Analyze Phase makes up 9 of every 50 questions on a Tigertest Six Sigma Green Belt practice test. Here are 10 of the 36 questions on it, each with the answer and the reason.

Updated 2026-11-10 by the Tigertest team. Based on the ASQ CSSGB Body of Knowledge (2022). Original questions; 50 per test, weighted like the real exam, with a 70% readiness line. Not affiliated with ASQ.

1. 1. A multi-vari chart shows large thickness differences across the width of each sheet. Which family of variation is this?

   • A.Measurement error
   • B.Positional (within-piece)
   • C.Cyclical (piece-to-piece)
   • D.Temporal (time-to-time)

   Answer: B. Positional variation occurs within a single unit or across locations, such as cavities or points across a sheet. Cyclical is between consecutive units; temporal is over time.

2. 2. Consecutive parts from one fixture differ a lot, but each part is uniform. Which variation family dominates?

   • A.Temporal (time-to-time)
   • B.Gauge-to-gauge error
   • C.Positional (within-piece)
   • D.Cyclical (piece-to-piece)

   Answer: D. Cyclical variation is the change from one unit to the next in a run. A multi-vari chart shows it as long jumps between consecutive pieces.

3. 3. A multi-vari study shows the averages drift steadily from morning to afternoon shifts. What family of variation is this?

   • A.Temporal (time-to-time)
   • B.Random sampling error
   • C.Positional (within-piece)
   • D.Cyclical (piece-to-piece)

   Answer: A. Temporal variation appears between time periods such as hours, shifts or days, and often points to tool wear, temperature or material lots.

4. 4. What is the main purpose of a multi-vari study early in Analyze?

   • A.To set specification limits
   • B.To replace a designed experiment
   • C.To locate the biggest variation
   • D.To prove cause and effect

   Answer: C. A multi-vari study is passive: it samples the running process to show which family of variation is biggest, so later experiments can focus there.

5. 5. Ice cream sales and drowning incidents are strongly correlated. What is the best conclusion?

   • A.Drowning drives up ice cream sales
   • B.The data must be wrong
   • C.A third factor may drive both
   • D.Ice cream causes drowning

   Answer: C. Correlation does not prove causation. Hot weather increases both swimming and ice cream sales, so it is a lurking variable.

6. 6. A correlation coefficient of r = -0.91 indicates:

   • A.A strong negative relationship
   • B.No relationship at all
   • C.A strong positive relationship
   • D.A weak negative relationship

   Answer: A. r ranges from -1 to +1. A value near -1 means that as x rises, y falls in a nearly straight line.

7. 7. Two variables have r = 0.82. What share of the variation in y does a linear fit with x explain?

   • A.About 18%
   • B.About 82%
   • C.About 91%
   • D.About 67%

   Answer: D. The coefficient of determination is r squared: 0.82 x 0.82 = 0.67, so about 67% of the variation in y is explained.

8. 8. A regression gives cycle time = 12.4 + 3.1 x (number of line items). What cycle time is predicted for 5 line items?

   • A.20.5 minutes
   • B.27.9 minutes
   • C.15.5 minutes
   • D.77.5 minutes

   Answer: B. Substitute x = 5: 12.4 + 3.1 x 5 = 12.4 + 15.5 = 27.9 minutes. 15.5 minutes is only the slope term.

9. 9. In the model y = 12.4 + 3.1x, what does 3.1 mean?

   • A.The correlation is 3.1
   • B.y rises 3.1 for each unit of x
   • C.y equals 3.1 when x is zero
   • D.x explains 3.1% of y's variation

   Answer: B. The slope is the change in y for a one-unit increase in x. The intercept, 12.4, is the predicted y when x is zero.

10. 10. Why is it risky to use a regression model to predict y far outside the range of x in the data?

    • A.r squared always doubles
    • B.The residuals all vanish
    • C.The slope becomes zero
    • D.The trend may not hold

    Answer: D. Extrapolation assumes the same linear relationship continues beyond the data, which may not be true. Predictions are safest inside the observed range.