STAT 3113 Final Project Report

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

1

Aidan Brooks
Ryan Brooks
Lucas Scott
April 29, 2021
Statistics 3113, Final Project Report

Project Objective:

The objective of the experiment is to assess the effect of drop distance and inner-ear
perturbance on the number of ping-pong balls, out of three, successfully dropped into an empty
standard-sized Solo cup.

Experimental Procedure:
To perform the experiment, the ping-pong ball dropper was asked to attempt a treatment
combination according to the randomization performed beforehand. The ping-pong ball dropper
then attempted to drop a ping-pong ball into a Solo cup from the specified drop distance three
times after completing the specified number of blindfolded/eyes-closed clockwise spins. The
number of successful drops was recorded. The dropper took five-minute breaks between each
treatment to ensure that no residual dizziness impacted the next trial. After each break, the
dropper then repeated this procedure with the next treatment combination until all 27 treatment
combinations were completed. Replications were completed by the experimenters individually to
minimize bias due to using a single test subject, and to minimize the risk of contracting COVID-
19 (an impartial observer was located for each experimenter and asked to supervise as a means of
preventing injury while spinning).

Response Variable:
Number of successful drops out of three.

Factors and their Levels:


Factor 1: Drop Height
Levels:
a) 1 ft
b) 3ft
c) 5ft

Factor 2: Inner-Ear Perturbance


Levels:
a) Spin 0 times with eyes closed or blindfolded.
b) Spin 5 times with eyes closed or blindfolded.
c) Spin 10 times with eyes closed or blindfolded.
2

Randomization:
Randomization was obtained within each replication by using a random number
generator (randperm(n)) in MATLAB. Replications were performed independently and
potentially concurrently so that no bias was introduced.

Replication Number:
33=27 replications
Project Design:
The experiment that was performed was a test of how a series of two factors would affect
the accuracy of a ping pong ball dropped into a cup. The factors tested were the drop distance
and inner ear perturbance (dizziness). After each combination the results, whether or not the ball
landed in the cup, were recorded. The experiment was replicated three times to insure a wider
sample of data.

Data Collection:
Each of the test combinations were performed in a random order with the drop height and
spin count being set each time (with rest time in between), after which the statistic of whether the
ball landed in the cup was recorded in a data table made in Microsoft Word. The process was
then repeated for all of the combinations.

Check Assumptions:
The results from the series of experiments were inputted into the statistical software JMP
in order to create an in depth way to analyze the data. JMP was used to create the residual plots
for the data (Figure 1) and the residual normal plot (Figure 2).

Figure 1: Residual Plots


3

Since an identifiable pattern is clearly visible within the Residual by Predicted Plot, the
data does not satisfy the constant variance assumption of ANOVA. Regardless, ANOVA was
used to determine the significance of the factors to the experimental model.

Figure 2: Residual Normal Quantile Plot

From the residual normal quantile plot in Figure 2 above, we conclude that the data fit a
normal distribution well enough to satisfy the normality assumption of ANOVA.

ANOVA:
With the residual plots created and the assumptions checked, JMP was used to create the
corresponding ANOVA table for the experimental data (Figure 3).

Figure 3: ANOVA Table

Overall Model:
H0: Overall model is not significant.
H1: Overall model is significant.
F-value: 3.3958
P-value: 0.0149
Based on the values from the effects table it is clear that because the P-
value of 0.0149 is less than the significance level of 0.05 (P-value = 0.0149 < ⍺
= 0.05), the null hypothesis can be rejected. Resulting in the conclusion that
4

the overall model does have a significant effect on the number of successful
drops out of three.

Effects Table:
The effects were reviewed using Figure 4 below and a significance level of ⍺ = 0.05.
It was assumed that the ANOVA process would fit the experimental data.

Figure 4: Effects Table

Two-way interaction:
H0: The two-way interaction between drop distance and spin count is
insignificant.
H1: The two-way interaction between drop distance and spin count is significant.
F-value: 0.9583
P-value: 0.4540
Based on the values from the effects table it is clear that because the P-
value of 0.4540 is greater than the significance level of 0.05 (P-value = 0.4540
> ⍺ = 0.05) the null hypothesis cannot be rejected. Resulting in the conclusion
that the two-way interaction between drop distance and spin count does not
have a significant effect on the accuracy of a dropped ping pong ball.

Main effects:
Drop Distance:
H0: The effect caused by the drop distance of the ping pong ball is not significant.
H1: The effect caused by the drop distance of the ping pong ball is significant.
F-value: 9.0833
P-value: 0.0019
Based on the values from the effects table it is clear that because the P-
value of 0.0019 is less than the significance level of 0.05 (P-value = 0.0019 < ⍺
= 0.05) the null hypothesis can be rejected. Therefore, the drop distance has a
significant effect on the accuracy of a dropped ping pong ball.
5

Main effects:
Spin Count:
H0: The effect caused by the number of spins is not significant.
H1: The effect caused by the number of spins is significant.
F-value: 2.5833
P-value: 0.1032
Based on the values from the effects table it is clear that because the P-
value of 0.1032 is greater than the significance level of 0.05 (P-value = 0.1032
> ⍺ = 0.05) the null hypothesis cannot be rejected. Therefore, it can be
concluded that the number of spins is not a significant factor in the accuracy
of a dropped ping pong ball.

Conclusion:
In this experiment we observed and analyzed the effect of drop distance and inner ear
perturbance on successfully dropping a ping pong ball into a Solo cup. Our experiments were
designed to eliminate as many outside interferences as possible with measures like waiting five
minutes between drops and wearing a blindfold while spinning. These measures, however, could
not eliminate all outside interferences such as external perturbation and air movement. These
interferences were negligible and did not impact our results. Through ANOVA, we concluded
that our model was significant. Through the analysis of the effects, we came to the conclusion
that the two-way effect was not significant (P-value=0.454>𝛼=0.05). Interestingly, when
analyzing the effect of each factor individually, we found that only drop distance was significant
(P-value=0.0019<𝛼=0.05) as spin count, like the two-way effect, was not significant (P-
value=0.1032>𝛼=0.05). This means that as the drop distance increases the chance of a successful
drop decreases, regardless of the number of spins before the drop. So we can conclude that inner-
ear perturbance, in relation to drop distance, has no significant effect on successfully dropping
the ball into the Solo cup. It is believed that the reason this experiment did not satisfy the
constant variance assumption is because of a lack of test data with a larger spread. In order to
correct this, data would have to be obtained from a test with a greater range of drop distances and
spins, such as one to ten feet or 0-20 spins, each done with a larger number of replications.
6

Appendix:
Table 1: Lucas’s Raw Data
Lucas Scott Zero spins Five spins Ten Spins

1 foot
1 2 3 4 5 6 7 8 9

yes yes yes yes yes yes yes no yes

3 feet
10 11 12 13 14 15 16 17 18

yes yes no Yes no yes No yes No

5 feet
19 20 21 22 23 24 25 26 27

Yes no yes yes No No no no no

Table 2: Aidan’s Raw Data


Aidan Brooks Zero spins Five spins Ten Spins

1 foot
1 2 3 4 5 6 7 8 9

Y Y Y Y Y Y Y Y Y

3 feet
10 11 12 13 14 15 16 17 18

N Y Y N Y Y Y Y Y

5 feet
19 20 21 22 23 24 25 26 27

Y N Y Y Y Y N Y Y

Table 3: Ryan’s Raw Data


7

Ryan Brooks Zero spins Five spins Ten Spins

1 foot
1 2 3 4 5 6 7 8 9

Y Y Y Y Y Y Y Y Y

3 feet
10 11 12 13 14 15 16 17 18

Y Y Y N Y Y N Y Y

5 feet
19 20 21 22 23 24 25 26 27

Y N Y Y N Y N N N
8

You might also like