… deals with multiple input variables, complex sets of results, and unstable processes through a practical, hands on, combination of proven techniques.
…describes a range of problems encountered by a typical business and introduces the techniques as they address each issue. Analytical tools are explained using a graphical, pictorial, approach.
A series of case studies – one for each method – explains how the tools are chosen, used, and interpreted. Spread sheets minimise calculation and improve understanding.
At the end of the course, students tackle their own issues by identifying appropriate tools and making progress with help and guidance.
- Find influential factors from historical data
- Discover root causes through regression and correlation
- Reduce queues, inventory and bottlenecks through simulation
- Rescue designed experiments which don’t work by using multiple regression
- Analyse using Excel spreadsheets provided with
You will learn to:
- Predict future results
- Establish relationships between input and output
- Discover the different clusters hidden within your data
- Identify the onset of problems with a cusum chart
- Find the clues in your routine records to uncover root causes
- Discover which potential factors really do have an effect.
The Need for Advanced SPC for Problem Analysis
Problem situations which need advanced statistical methods. Visual outline of each method, explaining basic idea and how and when it is useful.
Establishing relationships and predicting future results
- Line of best fit, regression and correlation
- Deciding whether a relationship is real
- Danger of assuming cause from correlation
- Prediction and prediction errors.
Finding influential inputs
- Multiple regression as an extension of line of best fit
- Pictorial explanation
- Spread sheet functions
- Determining size and significance of inputs
- Application to business data.
Rescuing designed experiments
- How to analyse DOE and Taguchi experiments using spread sheet regression functions
- How to modify the array and continue to analyse when data is missing because results were lost or a condition was forgotten.
- Analysis of variance for differences between dimensional results.
- Contingency tables and the chi square test for differences between distributions of categorical results.
- Spread sheet exercises.
Finding clusters in complex data
- Visual approach to find groupings of products, plants, locations.
- Cluster analysis using statistical computer packages.
Discovering underlying factors
- Correlations between multiple measurements
- The correlation matrix
- Factor analysis to find the basic few drivers which generate the data
- Visual methods to discover the underling factors
- Factor analysis using statistical computer packages.
Understanding the flow of information and material
- Queues, inventory and bottlenecks.
- Histograms for describing flow.
- How to build a simulation model
- Testing the model with a desk simulation.
- Transferring to a spread sheet
- Danger of drawing conclusions from a simulation model, and need for comparative experiments with alternative models.
Finding when a problem started
- Cusum charts using a spread sheet template.
Combining estimates of PPM (DPMO)
- Parts per million (defects per million opportunities) and probability
- Combining independent probabilities, and translating back to PPM.
SPC for non standard processes
- The four major types of non standard process
- How to use SPC methods to reduce variation and improve control.
Alan Long is a Chartered Statistician and is an acknowledged expert in the application of statistical methods to a wide range of business and operational environments. He has been applying statistical methods to problems in Australian businesses since 1982. Alan enjoys an excellent reputation as an instructor and wrote the STATMAN series of training books.
$795 excluding GST. Fee includes Training Manual, Excel spreadsheet files for data analysis, lunch & refreshments. Please see refer to our standard terms and conditions for further registration details.