ProjectAssistance

Part 2

Analyze:

1. Choose one continuously variable factor to experiment with. For example, you might choose “length of airplane” as a factor to experiment with. I will use this example factor in the remaining description, but you simply substitute the experimental factor that you choose. Collect data for flight times versus airplane length. Create a scatter plot (X-Y plot) to visually detect if there is a relationship between length of plan and time of flight. In Minitab you can create this scatter plot as follows: Graph>scatter plot>simple>ok> click on time of flight as the Y variable and length of plane as the X variable > click on the column with your time of flight data as the Y variable and the column with the length of plan as the X variable.

Click ok and you will get a plot that looks like this:

You can visually assess if there is a relationship. In this example, it appears that shorter planes fly longer.

2. You can fit a regression line by choosing the option “with regression” instead of “simple”.

In order to determine if the apparent regression relationship is statistically significant, you need to run a regression analysis as follows: Stat>Regression>Fitted line plot> click on the column with your time of flight data as the Response (Y) variable and the column with the length of plan as the Predictor (X) variable. You will get the scatter plot with the regression line as follows:

You also need to go to the session window to check the p-value:

In this example the p-value is 0.007. For regression analysis, the p-value determines the significance of the regression relationship. If the p-value is 0.05 or less, then you conclude that the relationship is significant and is not simply a coincidence. If the p-value is greater than 0.05, then you conclude that whatever relationship that may seem apparent in the scatter plot might well be simply a coincidence and that the relationship is not statistically significant.

In Excel, you can create a scatter chart by highlighting the two columns of data and then clicking insert>scatter chart. If you then click on the data dots, you can right click and add trendline, which will be the regression line (usually linear).

To get the regression analysis, you click data>data analysis>regression> highlighting the Time flight data for Y and the length of plane data for X. The results will be as follows:

The p-value is the Significance F value.

Improve:

Based your Analyze phase work, choose and implement an improvement to your airplane. Fly the improved plane ten times. Compare the improved plane results with the baseline results determined in the Define phase. Is the improved phase average flight time longer than the baseline average flight time? This can be statistically assessed with a two sample t-test. In Minitab: Stat>basic stat> 2-sample t>click samples in different columns>put baseline data in first cell and new data in second cell>click ok. You get something like the following in the session window:

Two-Sample T-Test and CI: time of flight, new