A colleague show me some data that he collect from 3 different machines and ask for help to perform some statistical analysis.
He would like to assess the consistency of the part dimension that produce from 3 different machines, to make decision whether to mix into 1 bin or pack it separately according to machine number.
I am trying to use ANOVA for the data analysis
The sample data given to me as below
Control dimension = 15.000
|
M/C1 |
M/C2 |
M/C3 |
|
13.548 |
13.655 |
14.469 |
|
16.353 |
13.793 |
15.409 |
|
13.456 |
15.263 |
14.006 |
|
14.246 |
16.279 |
13.935 |
|
15.091 |
14.716 |
13.264 |
|
15.329 |
15.770 |
13.986 |
|
12.926 |
15.154 |
13.206 |
|
12.705 |
14.339 |
13.972 |
|
14.581 |
15.204 |
15.913 |
|
13.111 |
15.611 |
15.731 |
Below are the steps I perform the data analysis
Step 1: Normality Test
Minitab procedure: Stat >> Basic Statistic >> Normality test
Note: ANOVA only can apply on Normal Data
Step 2: Stack up the data into single column
Minitab procedure: Data >> Stack >> Columns
Step 3: Check for process stability
Minitab procedure: Stat >> Control Charts >> Variable Charts For Individuals >> I-MR
Step 4: Check for process capability
Minitab procedure: Stat >> Quality Tool >> Capability Analysis >> Normal
Step 5: Test for equal variance
Minitab procedure: Stat >> ANOVA >> Test for equal variances
Note: To assess whether the machine running under same std dev condition or not.
Step 6: ANOVA
Minitab procedure: Stat >> ANOVA >> One Way
Done….
I will discuss the analysis result on my next post…
