A designed experiment is a type of scientific research where researchers control variables (factors) and observe their effect on the outcome variable (dependent variable).

The phases of experimental design include:

## 1. Planning

Careful planning and attention to detail can help you avoid any pitfalls along your path. In most situations, you would have limited resources to conduct experiments. You would want to get the best results by conducting the minimum number of runs.

You start with a clear understanding of the problem and a well-defined purpose of the experiment.

At this stage, you would identify the potential factors (independent variables) that could be significantly affecting the response. Here you can use your past experience and subject matter expert knowledge to define relevant factors and their levels for the experiment.

In addition, you would need to ensure that the process being analyzed is under statistical control (Statistical Process Control) and that the measurement system variation is acceptable.

## 2. Screening

If the number of factors to be studied is large (typically more than 5), then as the first step, you would conduct screening experiments to reduce them.

The number of factors to be studied significantly impacts the number of runs. For example, if you decided to study 10 factors that could impact the response in the planning, then for a full factorial design, you would need to conduct 2^10 or 1024 experimental treatments (runs). In almost all situations conducting an experiment with these numbers of runs is impossible. In these situations, you would want to do some screening experiments to reduce the number of factors to be studied in the next step.

The following designs are typically used during the screening phase:

a) Fractional Factorial Design

b) Plackett-Burman Design

c) Definitive Screening Designs

## 3. Modelling

Once you have identified the significant factors using screening experiments, you will model the relationship between significant factors and the response. This is done using regression analysis.

The following designs are typically used during the modelling phase:

a) Fractional Factorial Design

b) Full Factorial Design

## 4. Optimizing

After identifying the significant factors and modelling the relationship between factors and response, you would optimize the process conditions to achieve the desired result. This phase involves finding the best combination of factors and levels to produce the optimum output.

The following designs may be used during the optimization phase:

a) Central Composite Design

b) Box-Behnken Design

## 5. Verifying

Verification is the final phase of the experiment. It is conducted after the optimized condition has been achieved. The verification helps you confirm whether the optimized condition was indeed optimal. If not, then you would modify the experimental plan accordingly.

## Conclusion

A good experiment should always begin with a clearly defined objective. A well-designed experiment will ensure that the experiment meets its objectives.