Statistics with Minitab

Course details

IT IS NOT POSSIBLE TO UTILISE SAMPLE DATA WITHOUT STATISTICS. Much time and effort is devoted to the collection of data in industry, for example; quality control measurements, data collected for validation of manufacturing processes, incoming and outgoing inspection data, data produced in the development of products in R&D, etc. It is not possible to get value from this data without using statistics. Many people who do actually use statistical tools such as Statistical Process Control, Design of Experiments, sampling standards, gauge R&R, and other applications don’t understand the underlying statistics. This course is intended to provide that essential understanding so that people will choose the appropriate statistical tools for data analysis and understand the outcome of the analysis.

There are several brands of reasonably priced computer statistical software packages available to assist in the application of statistics, and most people with a reasonable background in maths (example, pass leaving certificate level) can be readily trained to use this software so as to utilise data for continual process improvement, and better decision making.

Minitab software will be used throughout the training course. Delegates will be trained to use both the main menus and the Assistant in Minitab to undertake the analysis that will be met in the Programme set out below.  Where the course is presented in-company the programme can be modified to include specific statistical applications.

Learning Outcomes

On successful completion of this training course, delegates should be able to:

• Undertake statistical analysis using Minitab software
• Select appropriate statistical tests such as two-sample t, F-test, ANOVA, etc. for comparing data means and variances
• Calculate and interpret confidence intervals on population parameters
• Determine sample sizes for statistical tests
• Model data using regression analysis

Course Programme

Day 1

• Outline of the applications of statistics such as Statistical Process Control, Design of Experiments, Sampling, and the relationship with the underlying statistics.
• Explanation of how statistics are used to obtain valuable information on processes from sample data
• Description of statistical terms including population, parameter, random sample, expected value
• Types of data – continuous (variables) and discrete (attributes) data
• Construction of a histogram and explanation of the meaning of frequency distributions, cumulative frequency distributions, measures of dispersion and central tendency
• Graphical methods – box-and-whisker plots, scatter plots
• The normal distribution – testing for normality – Anderson Darling and Ryan Joiner tests
• Normal and Weibull probability plots
• Dealing with non-normal data – Box-Cox and Johnson transformation, distribution fitting using Weibull, Smallest Extreme Value, Largest Extreme Value, etc.

Day 2

• Central limit theorem and sampling distribution of the mean
• Calculation of the confidence interval for the mean in variables and attribute data.
• Hypothesis testing – tests for means, variances and proportions – Z-test, t-test, 2-sample t-test, F-test, meaning of significance level
• Meaning of the P-value in hypothesis testing and how the rules for assessing P are derived
• Type I and Type II error – difference between statistical and practical significance
• Sample sizes for hypothesis testing – the effect on Power
• Goodness-of-fit tests

Day 3

• Analysis of variance (ANOVA) – analysis of a designed experiment illustrating the ANOVA – using Tukey’s multiple-sample comparison to compare population means
• Simple and multiple linear regression and correlation. Calculation of the regression equation. Hypothesis testing of the regression statistics. Using the regression model for estimation and prediction. R-squared and R-squared adjusted – the difference between these two statistics.