top of page

Design of Experiment (DOE) ☆~(ゝ。∂)

  • jiamin20
  • Jan 24, 2022
  • 3 min read

Updated: Jan 25, 2022

To start off, DOE is a statistics-based approach to designing experiment. In clearer context, experimentation is required in order to obtain statistical data for DOE. Experimentation is conducted to study the effects of factors as they are set at various levels.


We were tasked to perform the FULL factorial design data analysis as well as the FRACTIONAL factorial design data analysis based on the case study we selected. I decided to choose case study 1 and here is how I performed the data analysis.




FULL factorial design data analysis



With the values of total effect calculated for me, I am able to plot a graph to determine the ranking of the most significant factor to the least significant factor affecting the amount of bullets formed. The factor with the steepest gradient would be the most significant factor.


Based on the line graph plotted, Factor C have the steepest positive gradient, it is the most significant factor affecting the amount of bullets formed. When power increases from 75% to 100%, the amount of bullets formed increases from 1.15g to 1.75g (+0.6g).


The second in rank would be Factor A, with a less steep negative gradient. When the diameter of bowls increases from 10cm to 15cm, the amount of bullets formed decreases from 1.55g to 1.35g (-0.2g).


Factor B, with the least steep positive gradient is ranked last. When the microwaving time increases from 4 mins to 6 mins, the amount of bullets formed increases from 1.375g to 1.53g (+0.155g).


Ranking of factors:

1. Factor C = Power

2. Factor A = Diameter

3. Factor B = Microwaving time


Next, after determining the effect of single factors and their rankings, I have to determine the interaction effect of the three factors.








AxB as well as BxC has significant interactions whereas AxC has a small interaction. I am able to conclude that changing Factor B has the most impact.


Conclusion:

The outcome of the effect of single factors (Part 1) differs from the outcome of interaction effect (Part 2). In part 1, the result shows that Factor C is the most significant factor due to it having the steepest gradient. However, in part 2, the interaction effect shows that Factor B have the most impact on the amount of bullets formed.


The outcome in part 1 may be less accurate as it is only based on one factor alone. The factors will be more effective when they interact with each other. Hence, the outcome in part 2 would be more accurate and reliable. Therefore, Factor B is the most significant factor affecting the amount of bullets formed.




FRACTIONAL factorial design data analysis


Run 1,2,5 & 7 is selected to conduct a Fractional Factorial Design Data Analysis. This is to ensure a balanced design by ensuring that all factors occur (high level and low level) the same number of times. This helps the design to obtain a good statistical property.




Based on the line graph plotted, Factor B have the steepest positive gradient, it is the most significant factor affecting the amount of bullets formed. When the microwaving time increases from 4 mins to 6 mins, the amount of bullets formed increases from 0.125g to 1.18g (+1.325g).


The second in rank would be Factor C, with a less steep positive gradient. When power increases from 75% to 100%, the amount of bullets formed increases from 0.3g to 1.28g (+0.98g).


Factor A, with the least steep positive gradient is ranked last. When the diameter of bowls increases from 10cm to 15cm, the amount of bullets formed increases from 0.4g to 1.18g (+1.58g)


Ranking of factors:

1. Factor B = Microwaving time

2. Factor C = Power

3. Factor A = Diameter


Interactions for Fractional Factorial cannot be found as there is insufficient data points to plot a interaction graph.

Conclusion:

Since it is not possible to find the interaction for fractional factorial, I am unable to make any comparison between the interactions. Hence, using FULL factorial would be the better option as I am able to identify all the significant interactions of the factors.




That's all for today, thanks for reading! (っ^▿^)💨


Comments


 
 

- THE END -
by Yeo Jia Min

bottom of page