Factorial Designed Experiment Aim

Multivariate experiments are a very powerful management tool to learn and improve performance. Experiments in general, and designed factorial experiments in particular, are dramatically underused by managers. A question on LinkedIn asks?

When doing a DOE we select factors with levels to induce purposely changes in the response variable. Do we want the response variable to move within the specs of the customers? Or it doesn’t matter since we are learning about the process?

The aim needs to consider what you are trying to learn, costs and potential rewards. Weighing the various factors will determine if you want to aim to keep results within specification or can try options that are likely to return results that are outside of specs.

If the effort was looking for breakthrough improvement and costs of running experiments that might produce results outside of spec were low then specs wouldn’t matter much. If the costs of running experiments are very high (compared with expectations of results) then you may well want to try designed experiment values that you anticipate will still produce results within specs.

There are various ways costs come into play. Here I am mainly looking at the costs as (costs – revenue). For example the case where if the results are withing spec and can be used the costs (net costs, including revenue) of the experiment run are substantially lower.

It is possible you have a case where just the cost of the experiment is very high but the loss of an unusable end result is low (this case is probably not very likely but is possible). In such a case then within spec targets would not matter much – the cost of being outside of the spec is low. It is also possible to have another situation a bit like this, if the end result was very costly but the loss function of being a certain amount outside of spec was low. So, for example, while your current customer would not accept the result, you could easily sell the product at 95% of the cost as long as it was somewhat near the spec.

Limiting your experiments to values that you felt confident would return results within the spec however, may well limit what can be learned. So the proper aim really depends on the hopes for the designed experiment and the associated costs and risks.

From my previous post on the use of factorial designed experiments:

Statistics for Experimenters: Design, Innovation, and Discovery shows that the goal of design of experiments is to learn and refine your experiment based on the knowledge you gain and experiment again. It is a process of discovery. If done properly it is very similar to the PDSA cycle with the application of statistical tools to aid in determining the impact of various factors under study.

This process of discovery is the key. If the costs of experiments (that produce out of spec results) are very high, then you likely need to attempt to experiments within bounds that are likely to have results remain in spec. Of course, if the rewards for a breakthrough are high enough then even high cost experiments can be wise.

Related: Articles on Multivariate Designed ExperimentsUsing Design of Experiments101 Ways to Design an Experiment, or Some Ideas About Teaching Design of ExperimentsYouTube Uses Multivariate Experiment To Improve Sign-ups 15%

3 thoughts on “Factorial Designed Experiment Aim

  1. Pingback: One factor at a time (OFAT) Versus Factorial Designs » Curious Cat Management Improvement Blog

  2. brad is absolutely correct.
    designing and running a statisticaly designed experiment requires effort and skill- what are we getting in return?
    the argument in favor of interaction discovery minimizes the issue.
    any statistical effort can be assessed with two measures:
    1. PSE- practical statiqstical efficirncy
    this measures assess the impact of the work. can the effort be reused, is the result sustainable over time,….
    2. InfoQ – what is the qualitybof the information provided by the effort.
    here we are assessing the way out work has adresses specific goals and added informwtion based on data and its anakysis.
    using DOE has an effect on both measures which is why it should be taught and applied.

    for more on this google my paper with galit shmueli in SSRN or my paper with silvia salini in ASMBI

    Reply
  3. Pingback: Statistical Techniques Allow Management to do a Better Job « The W. Edwards Deming Institute Blog

Leave a Reply

Your email address will not be published. Required fields are marked *