EXAMPLE DISCUSSION OF RESULTS
· Data is highly linear (1)
· Voltage increases linearly as displacement increases (1)
· Mention Zero Offset (1) What does it mean? (2)
· One outlying point (1)
o Impact on Linearity (2)
o Should point be kept? (2)
o Do Stats with and without point
· Hysteresis is not significant (include quantitative evidence - slope out equals slope in) (2)
· Precision error
The precision error was _____. (Observation)
What does it mean? (Interpretation)
TOOLS FOR DISCUSSION
o State the Obvious
o Describe Trends
o Put down what you see
o What does it mean?
o What difference does it make?
o Who cares?
Discussion of Results paragraph written by Dr. Ralph Budwig
The following is provided to give you an example for your future report writing endeavors.
The data shown in figure 1 clearly reveals a linear trend with voltage increasing as the LMP position increases. The standard deviation result indicates that a typical point is about 0.05 inches from the best-fit line. The third data point is about 10 times further away from the best-fit line than all the other data. We have retained this point in our performance specification calculations, but we recommend that a second set of calculations be performed with this point excluded.
Table 1. Performance specifications calculated for the LMP.
The linearity of the LMP is 3.9%. This value is higher than anticipated because it is calculated with the residual of the worst point (the third data point). The standard deviation of the LMP is 0.051 inches. This value also seems large since the caliper that was used for the calibration has a resolution of 0.001 inches. A LMP with this standard deviation could not be used as a milling machine readout transducer because the scatter would be too large (a typical milling machine has readout resolution of 0.0005 inches).
Figure 1. Calibration results for the linear motion potentiometer. The diamond symbols represent data taken on LMP closure, and the square symbols when the LMP was being opened. The solid line is the best fit linear approximation to all of the data.