Our assignment was to find a way to predict a tree's mass by it's girth
because, obviously, you can't go out and cut down every three just because
you want to know how much wood it would yield. Because we couldn't go out and
cut down a few trees on which to base our equation, we used carrots instead.
Jeff went out and bought a bag og carrots, of which we used most, and numbered
them one through 24.
To get some measurements so we xould know something about the carrots, we
used a caliper, a triple beam balance, and a calculator. First, we measured the diameter
of the carrots with the caliper, which is an istrument used to measure that is used to
measure the diameter or height or anything else, really, of objects. Then we got the mass
with the triple beam balance. And finally we multiplied the diameter time's pi to find the girth.
After we had gotten our measurements we put the girth and the mass into the computer
software JMP, and graphed it. We had two points that were suspected as outliers and were
messing up all of our data; out leat squares regression line only predicted 74.7% of our variations,
a very low score. So we deleted points five and seven and that brought our R squared up to 92%
accuracy, which was ahuge improvement.
After removing points 5 and 7 were-graphed the data and tested to see wether a curve
would fit it better than a line. We graphed the natural log of the mass (a curve) on top of the linear
fit of Mass by Girth. We found that the In (mass) fit the data better than the linear option
by approxiamately 2%. We finished our work by plotting In (mass) by girth on a separate set of
axis.
In conclusion, the equation that we found to fit out data best was In(mass)=2.55+.02 girth.
It is important to remeber, however, that because we eliminated two outliers during our
process, trees of unusual height, girth, or with certain anomalies should be excluded and may
not be correctly predicted by this model.
