Hypothesis Testing
Eva Chan 
Analyze_Phase_Intro_HT_Jul19_Exercise.pptExercise
- Develop Hypothesis Statement(s)
- Determine applicable Hypothesis Test
- Gather and format appropriate data
- Perform Hypothesis Test and check test assumptions
- Make a decision
Lean Six Sigma Green Belt
Running a hypothesis test is rather straightforward. Care should be taken to translate your statistical questions into “Practical” solutions.
Lean Six Sigma Green Belt
Advanced Analysis
IIL-LSSGB
5_AA-*
Advanced Analysis
Develop Hypothesis Statements
Complete, concise description of what you think is happening
Hypothesis: Enter hypothesis here
Null hypothesis: enter null hypothesis statement here
Alternative hypothesis: enter alternative hypothesis statement here
Lean Six Sigma Green Belt
P525-13C05
In everyday, non-statistical use, Hypothesis Statements are used to explain what you believe is happening: for example, what’s causing a problem or variation in a process. Remember that a good Hypothesis Statement should give a clear explanation that would probably be understood by anyone not familiar with the problem.
When we elect to do a statistical test of hypotheses, these “what we think” hypotheses have to be revised into TWO hypothesis statements: the Null and the Alternative. (The examples above are really Alternative hypotheses).
Advanced Analysis
Chi-Square Exercise
Identify the strata (e.g.: post with hashtags vs. post with no-hashtags
Gather observed frequencies data and populate matrix:
Count number of posts with number of likes less than customer specification, enter on defect row for each strata.
Count number of posts with number of likes greater or equal than customer specifications, enter on Good row for each strata.
| Strata 1 | Strata 2 | ||
| Defect | fo = | fo = | Total = |
| Good | fo = | fo = | Total = |
| Total = | Total = | N = |
Lean Six Sigma Green Belt
Lean Six Sigma Green Belt
Advanced Analysis
IIL-LSSGB
5_AA-*
Advanced Analysis
Chi-Square Example
Calculate the “expected” frequencies from the observed data:
fe = f row x f col
N
Enter expected frequency
Enter expected frequency
Enter expected frequency
Enter expected frequency
Calculate degrees of freedom
Degrees of freedom = (# rows – 1) x (# cols – 1) = enter value
| Strata 1 | Strata 2 | |
| Defect | fe = | fe = |
| Good | fe = | fe = |
Lean Six Sigma Green Belt
Lean Six Sigma Green Belt
Advanced Analysis
IIL-LSSGB
5_AA-*
Advanced Analysis
Chi-Square Example
Calculate chi-square (2) value for each cell and then sum together to calculate total 2:
Total X2 is the sum of the values for each cell:
X2 = enter value here
Lean Six Sigma Green Belt
Lean Six Sigma Green Belt
Advanced Analysis
IIL-LSSGB
5_AA-*
Advanced Analysis
Chi-Square Example
Step 5: Compare to a 2 critical (from chi-square table)
If 2calc > 2crit,
there is a significant difference
2critical =
Write your conclusion here. Is there a statistical significant difference or not?
Enter value
2calc =
Enter value
Lean Six Sigma Green Belt
Lean Six Sigma Green Belt
Advanced Analysis
IIL-LSSGB
5_AA-*
Advanced Analysis
