Design of experiment 👏

For this entry, we were tasked to do data analysis on either of the Case Study given and find out which factor has the most impact on the desired result accordingly.

Below is the Case Study that I have chosen.

The excel file which I have done can be retrieved from below here:

Case Study 1 Excel File

To begin off, I tabulated all the data into the excel file already prepared and as shown below is what I have tabulated according to the result given in Case Study 1.

Full Factorial Design Analysis

Moving on I did the full factorial data analysis first using the data tabulated, the results below are what I have gotten. After tabulating the data shown on the left, I was able to plot 3 different graphs which showed me how each factor has a different significance in terms of affecting the mass of unpopped yield.

It can be seen from the graph that power has the steepest gradient followed by microwaving time and lastly the diameter of bowl. The graph with the steepest gradient would have the highest effect on the mass of unpopped yield and the one with the least steep gradient would have the lowest effect. Thus, the factor with the highest to lowest effect would be Power, Microwaving time, and lastly Diameter of the bowl.

After determining the single factor effect, I then proceed to determine whether there is an interaction effect between the 3 different factors. As shown below is the results.

For the interaction of AB, the gradient of both lines are different (one is + and the other is -). Therefore there’s a significant interaction between A and B.

For the interaction of AC, the gradient of both lines are different by a little margin. Therefore there’s an interaction between A and B, but the interaction is small. If both lines are parallel, then there’s NO interaction

For the interaction of BC, the gradient of both lines are negative and has different values. Therefore there’s a significant interaction between B and C.

Thus to conclude full factorial design analysis, power has the most effect followed by microwaving time and lastly the diameter of the bowl in terms of yield of the unpopped kernel. Hence in order to minimize the yield of the unpopped kernel, power and microwave time will be set to high configuration and the diameter of the bowl can be either low or high configuration as the line plotted is close to horizontal which meant not much difference in low or high configuration.

Fractional Factorial Design Analysis

Next, I moved on to doing fractional factorial design analysis based on the result I tabulated on excel, the runs I chose are #2, 3, 5 & 8. Highlighted in red are the runs I will be doing fractional factorial design on.

With this data, I was able to plot another graph similar to the one I have shown on top, but this time with half the data points.

The results I have gotten this time doing fractional factorial differs slightly from the previous result I have gotten using full factorial design. It is clearly seen that the diameter of the bowl has an effect on the yield of unpopped kernel this time round. Thus in terms of effects of the single factor starting from highest to lowest, it will be as followed, Power, Microwave time, and lastly the Diameter of the bowl.

Thus to conclude fractional factorial design analysis, power has the most effect followed by microwaving time and lastly the diameter of the bowl in terms of yield of the unpopped kernel. Hence in order to minimize the yield of the unpopped kernel, power and microwave time will be set to high configuration, and the diameter of the bowl will be set to low configuration.