AMDM-DB3

DB#3
Conduct an experiment and record the data
Table SEQ Table * ARABIC 1. Frequency table for data distribution
Distance in meters that tennis ball is thrown by kindergarten pupils (m) Number of pupils (frequency)
1 2
2 4
3 7
4 5
5 7
6 3
7 15
8 13
9 10
10 6
Create a histogram

Figure SEQ Figure * ARABIC 1. Histogram of distance in meters that a tennis ball is thrown by kindergarten pupils against the number of pupils that throw the distanceWhat is the shape of the histogram
The histogram shape is random with no apparent pattern having many peaks.
How would you expect the data distribution to look?
I would expect the data distribution to be random with no distinct pattern and with many peaks and modes.
Why?
This random pattern with many peaks has been introduced by variations noted for the kindergarten pupils’ ages, sizes, gender, and race. In essence, the demographic peculiarities noted for the pupils have introduced many classes that have in turn caused the random distribution.
Does it look like a normal curve? Why?
No, it does not look like a normal curve.
That is because it is asymmetric with a notable skew towards the right. Additionally, the mean is not at the center, there are more than one peak, the standard deviation predictions do not follow the normal distribution rules (68-95-99.7 rule). In essence, it does not present Gaussian distribution thereby making it clear that it is not a normal curve.
How do the results differ from your ‘expected distribution shape’?
I expected the distribution to adopt a random pattern.

As such, my expectations match the results. I believe that the two match because a large number of kindergarten pupils (72) was evaluated for the study. The large number of pupils ensured that there was a wide distribution of classes so that many peaks were presented to result in the random distribution.
How does this relate to our understanding of the normal distribution (bell curve) and the use of sampling for measurement of population?
The realized distribution shows that presence of classes has an influence on the shape of the curve. Having fewer classes increases the possibility of realizing normal distribution with a single peak whereas having a more classes increases the possibility of having a random distribution with many peaks. That is because each class would present its own peak although there are some peaks that could overlap. As such, presenting normal distribution for a population is best achieved by sampling from a single class and introducing exclusion criteria for the other classes.
What do you think this assignment was made part of the course?
This assignment was made part of the course to improve understanding of sampling strategies and how they influence data distribution. To have more reliable results, it is important to have a normal distribution since it is evidence that the sample was from a single class and the results can then be inferred for that class. Having a distribution that differs from the bell curve is evidence that the sample presented many classes thereby making it difficult to infer the results for any single class and thus reducing the research value. In this respect, it is important to ensure that the population sample is restricted to a single class so that the results distribution present a bell curve and the results become easier to interpret. As a result, the assignment was made part of the course to improve the understanding on importance of sampling and how it affects the results and their interpretation.