Technique Objective Where Applicable When Application
Multi-van Chart 1.Discover family of red X-position, cycle or time
2.Detect non random patterns Wherever stratified samples of consecutive units, at different times, can be measured 1.At engineering pilot run; production pilot run;
2.During production problem-solving
Components Search Home in on a few key variables from a wide variety of variable to capture family of red X or red X itself Where there are good and bad assemblies, capable of being disassembled and reassembled 1.When only a few units are available (prototype stage or in pilot runs)
2.During production problem-solving
Paired Comparisons Same as components search Wherever there are good and bad assemblies or components that cannot be disassembled Mainly in production field returns or in failure analysis
Variables Search 1.Pinpoint red X, pink X, etc
2.To separate and quantify the main effects and interaction effects of variables Homing in on the culprit variable following mutli-vari, components search or paired comparison experiments 1.When more than 4 variables are investigated
2.At prototype, pilot run stages
3.In production problem solving
Full Factorials Same as variable search Same as variables search 1.When 4 or less variables are investigated
2.Same as variables search
3.Same as variables search
B vs C 1.Evaluate a minimum amount of superiority of method over method C or vice-versa
2.Select B or C if there is no performance difference, but a definite cost difference 1.As a validation of red X, pink X factors selected in variables search or full factorial experiments
2.When problem is easy to solve, us as the only experiment, by passing other techniques 1. In prototype or pilot stages
2.In productions as a capping run
3.In nontechnical fields sales, advertising, human relations etc, a universal tool
Scatter Plots 1.Determine optimum values of important variables
2.Cost reduction of unimportant variables Following previous 6 techniques In pilot run of production stage