Minitab 16 Workshop

Download Report

Transcript Minitab 16 Workshop

Minitab® 16
Workshop
Presented by Arved Harding
Your friendly, neighborhood statistician

Outline
 A few new features
 Minitab® ’s Assistant
 Minitab® Tutorials
 Quality Trainer
 Quality Companion
2
Credit where it is due.
 Many slides were created by Christy Bower, a statistician
at Eastman in our Applied Statistics Group.
3
A few new features “Quality People” might like






Gage R&R Study (Expanded) – Include additional factors,
analyze studies that are balanced or unbalanced, specify if
factors are fixed or random, and more.
Compare two capability analyses to see if the process has
improved.
General Regression – Easily specify interaction and
polynomial terms, compute confidence intervals for regression
coefficients, and more.
Export to PowerPoint and Word – Directly export your graphs
and Session window output for easy reporting.
Nonlinear Regression – Model non-linear relationships
between variables.
Power and Sample Size – Use seven new tools to calculate
power and sample size.
4
More power and sample size options
Minitab® provides power, sample size, and difference (effect)
calculations for the following procedures:














·
·
·
·
·
·
·
·
·
·
·
·
·
·
one-sample Z
one-sample t
two-sample t
paired t
one variance test
two variances test
one-sample proportion
two-sample proportion
one-sample Poisson rate
two-sample Poisson rate
one-way analysis of variance
two-level factorial designs
Plackett-Burman designs
General Full Factorial designs
You can also estimate the sample size you need to obtain a specific
margin of error using Sample Size for Estimation.
5
Sample size for estimation
Parameters include:
•Mean
•Standard deviation
•Variance
•Proportion
•Rate
6
7 Languages
In addition to all the stuff we know about you can have Minitab® in a language other
than English if you prefer.
Under Tools>Options you can change the language
Korean English French
German Chinese Spanish Japanese
7
Resources for Minitab® 16
 Minitab® Tutorials (45 topics)
• Provides detailed descriptions and examples
 Uses for the given tool or analysis
 Required data for the analysis
 How to complete the analysis in Minitab®
 Minitab® Assistant
• Interactive step-by-step guide to completing common data analyses in
•
Minitab®
Provides ready-made reports
 Report card of analysis assumptions
 Diagnostic reports for evaluating outliers and stability
 Summary report of the analysis
 Quality Trainer
• E-learning course with hands-on exercises and quizzes
8
Minitab® tutorials
9
45 Tutorials on many topics of
interest
10
2-Sample t Minitab® tutorial
11
What kind of data is required and
what the worksheet should look like.
12
As well as an example and step-by-step
instructions on how to use the tool
13
Minitab® ’s Assistant
When you feel you’re beyond help, consider
Minitab® Assistant
Minitab® has chosen a specific set of platforms to
add in Assistant targeted to help Six Sigma belts
get their analyses done more efficiently.
15
16
Check this out
17
18
19
20
Output is much nicer 
Typical output is a 3-4 page report. Each page can be copied
as a graphic to Word or PowerPoint with a right click.
Such as this 3 page report from a 2-sample t-test.
21
2-Sample t Test for the Mean of A and B
Report Card
Check
Status
Description
Unusual
Data
There are no unusual data points. Unusual data can have a strong influence on the results.
Normality
Because the sample size of A is less than 15, normality can be an issue. If the data are not normally
distributed, the p-value may be inaccurate with small samples. Because normality cannot be reliably
checked with small samples, you should use caution when interpreting the test results.
Sample
Size
!
The sample is sufficient to detect a difference between the means.
22
2-Sample t Test for the Mean of A and B
Diagnostic Report
Data in Worksheet Order
Investigate outliers (marked in red).
A
B
120
80
40
1
2
3
4
5
6
7
8
9
10
11
12
60%
90%
3
5
7
9
11
13
15
17
What difference can you detect with your
sample sizes of 12 and 17?
Power
What is the chance of detecting a difference?
< 40%
1
100%
11.652
17.074
For alpha = 0.05 and sample sizes = 12, 17:
If the true means differed by 11.652, you would have a 60%
chance of detecting the difference. If they differed by 17.074,
you would have a 90% chance.
Difference
Power
11.652
13.081
14.754
17.074
60.0
70.0
80.0
90.0
Power is a function of the sample sizes and the standard deviations. To detect a difference smaller than 14.754, consider increasing the
sample sizes.
23
2-Sample t Test for the Mean of A and B
Summary Report
Statistics
Do the means differ?
0
0.05 0.1
> 0.5
Yes
No
P = 0.004
The mean of A is significantly different from the mean of B
(p < 0.05).
Sample size
Mean
95% CI
Standard deviation
A
B
12
78.333
(71.27, 85.40)
11.122
17
62.294
(53.958, 70.630)
16.213
Difference between means*
95% CI
16.039
(5.6046, 26.474)
* The difference is defined as A - B.
95% CI for the Difference
Does the interval include zero?
Comments
0
10
20
30
Distribution of Data
Compare the data and means of the samples.
A
-- Test: You can conclude that the means differ at the 0.05
level of significance.
-- CI: Quantifies the uncertainty associated with estimating
the difference from sample data. You can be 95% confident
that the true difference is between 5.6046 and 26.474.
-- Distribution of Data: Compare the location and means of
samples. Look for unusual data before interpreting the
results of the test.
B
40
60
80
100
24
A warning about using Minitab® Assistant
Minitab® has tried to make these platforms as simple and
easy to use as they could. In doing so, sometimes they
took a slightly different statistical approach than the
traditional one. So the p-values may not always match up
exactly.
Minitab® Assistant
Minitab® Menu
(assuming equal variances)
25
2-Sample t using Minitab® Assistant
26
Demo using Hospital Rating Data
27
Minitab® ’s Quality Trainer
29
Quality Trainer
 Interactive guide to review specific statistical topics
 Learn at your own pace
 Test your knowledge with exercises and quizzes
 Certificate of completion for certification credits is
available from Minitab® upon completion of all material
and quizzes (approximately 20 hours of training)
30
Quality Trainer contents
31
Links to more information
 For a complete description of the new features see:
http://www.Minitab® .com/en-US/products/Minitab® /whats-new.aspx

 See the link below for access to a Minitab® 16 tour, webinars
and more. http://www.Minitab® .com/en-US/products/Minitab®
/launch.aspx
 For prerecorded webinars see http://www.Minitab® .com/enUS/theater/default.aspx

 To learn what statistical techniques Minitab® assistant uses in each
platform and why a detailed white paper has been written, go to
http://www.Minitab® .com/enUS/support/answers/answer.aspx?id=2613
32
Pricing info
33
34
Organize
35
Execute
36
Customize and The Dashboard
37
Questions?
38