TQM TOOLS

Pareto Principle

The Pareto principle suggests that most effects come from relatively few causes. In quantitative terms: 80% of the problems come from 20% of the causes (machines, raw materials, operators etc.); 80% of the wealth is owned by 20% of the people etc. Therefore effort aimed at the right 20% can solve 80% of the problems. Double (back to back) Pareto charts can be used to compare 'before and after' situations. General use, to decide where to apply initial effort for maximum effect.

Scatter Plots

 

A scatter plot is effectively a line graph with no line - i.e. the point intersections between the two data sets are plotted but no attempt is made to physically draw a line. The Y axis is conventionally used for the characteristic whose behavior we would like to predict. Use, to define the area of relationship between two variables.

Flow Charts

 

Pictures, symbols or text coupled with lines, arrows on lines show direction of flow. Enables modelling of processes; problems/opportunities and decision points etc. Develops a common understanding of a process by those involved.

Cause and Effect , Fishbone, Ishikawa Diagram

The cause-and-effect diagram is a method for analysing process dispersion. The diagram's purpose is to relate causes and effects. Three basic types: Dispersion analysis, Process classification and cause enumeration. Effect = problem to be resolved, opportunity to be grasped, result to be achieved. Excellent for capturing team brainstorming output and for filling in from the 'wide picture'. Helps organise and relate factors, providing a sequential view. Deals with time direction but not quantity. Can become very complex. Can be difficult to identify or demonstrate interrelationships.

Histogram or Bar Graph

 

 

  

A Histogram is a graphic summary of variation in a set of data. It enables us to see patterns that are difficult to see in a simple table of numbers. Can be analysed to draw conclusions about the data set.

A histogram is a graph in which the continuous variable is clustered into categories and the value of each cluster is plotted to give a series of bars as above. The above example reveals the skewed distribution of a set of product measurements that remain nevertheless within specified limits. Without using some form of graphic this kind of problem can be difficult to analyse, recognise or identify.

  

Check Sheets

 

A Check Sheet is a data recording form that has been designed to readily interpret results from the form itself. It needs to be designed for the specific data it is to gather. Used for the collection of quantitative or qualitative repetitive data. Adaptable to different data gathering situations. Minimal interpretation of results required. Easy and quick to use. No control for various forms of bias - exclusion, interaction, perception, operational, non-response, estimation.