# Statistics in Validating Root Causes Analysis

**Topics:**Statistics, Causality, Statistical significance

**Pages:**5 (574 words)

**Published:**May 13, 2013

Silvia Pederzolli

Milan, the 15th of april 2013

attivaRes

Define Opportunities

Measure Performance

Analyze Opportunity

Improve Performance

Control Performance

CCR’S

Objective

• • • • • Identify problem statement: what is wrong and why. Deviation from what is expected (targeted performance). How much/how often Effects on Customers. Find and validate the root causes that assure the elimination of “real” root causes.

Actions

•Identify Root Causes •Design Root Cause

Verification Analysis •Validate Root Causes

Deliverables

•Data Analysis •Process Maps •Validated Root Causes

attivaRes

1.

The major objective is to find the root causes

1. 2.

Source of Customer dissatisfaction Poor business result

2.

Through root causes analysis the team leads to the underlying source of defect and can change the process to permanently eliminate the problem

market

Process input

process

process

output

CCR’s

Output Variation = defect

attivaRes

Not easy to find Preconceived ideas for causes Pressure for quick solution Difficulty in collecting more data Wrong or apparent root causes will lead to wrong solutions

attivaRes

The fishbone diagram is useful to reach a common understanding of the problem. It is a strong tool to establish the relation between an effect and its main causes. It starts from the statement in the «head of the fish» Then the team determines major categories (the «bone») and brainstoms all possible causes

attivaRes

attivaRes

It turns brainstorming into practical solutions

brainstorming

Statistical results

Practical solutions

Verified root causes will lead to right solutions in the IMPROVE phase

Define Opportunities

Measure Performance

Analyze Opportunity

Improve Performance

Control Performance

CCR’S

attivaRes

«Root causes» mean «hypothesis» of relationship between two variables and they have to be tested. Not every relationship between two variables confirms that one causes the other It is possible that two or more variables are strongly related one to the other but none of them is caused by the other Validation of root causes is made only when: o There is a statistically relevant relationship between the root cause and its effect o Knowledge of the process confirms this causal relationship The more common statistical tools are the regression analysis and the chi-square test. But they are not the only tools the team can use…. attivaRes

Regression analysis (scatter plot)

It is a mathematical diagram using Cartesian coordinates. The scatter plot would give a visual comparison of the two variables in the data set, and would help to determine what kind of relationship there might be between the two variables. o It is, as Ishikawa diagram, one of the seven basic tools for quality (Ishikawa, Deming) o It is a simple statistic tool o

Dependent variable on Vertical axis

Independent variable or control parameter On horizontal axis

attivaRes

Chi-square test

Chi-square is a statistical test commonly used to compare observed data with data we would expect to obtain according to a specific hypothesis. o The chi-square test is always testing null hypothesis, which states that there is no significant difference between the expected and observed result. o That is, chi-square is the sum of the squared difference between observed (o) and the expected (e) data (or the deviation, d), divided by the expected data in all possible categories. o

Let’s try it!

attivaRes

5 W’s + 1 H approach

how

Workshop technique

Play Role! attivaRes

What is in it for me? What do we learned as a team?

Questions? Doubts? Suggestions?

Thank you for your active participation!

attivaRes

Please join StudyMode to read the full document