# Lab 5 Ryan Connaghan

Lab 5

After plotting the normal probability plot, I can see that the graph is not normal. I cannot seem to get the graphs out of PAST, but will try to do so in class.  For the most part, graphs of the Normal Probably Plot are showing a log scale relationship, with C having the inverse of this.  Taking the log of shape length and plotting it the same way gives me an s shaped graph with the left side lower and the right side slightly raised.  So this is much closer to a normal distribution, but from what I am reading, the S shape means that there is more variability than a normal distribution.

Below are the univariate statistics for shape length and shape area.  You can see in the length chart the minimum value of 30, a max of 3734, and a mean of 184, that the values are skewed toward the lower end.  The standard deviation of 80, tells me that the values are not that widely off, and that there are most likely outliers toward the upper limits.  The graphs shows that there is at least one value that is significantly larger than the rest, it does however line up along a curve with the other values, suggesting that a log scale might work on this data to better conform to a normal distribution.

0 Shape_Leng
N 11737
Min 30.75888
Max 3734.881
Sum 2169967
Mean 184.8826
Std. error 0.7394023
Variance 6416.804
Stand. dev 80.10495
Median 169.0816
25 prcntil 138.6115
75 prcntil 211.7965
Skewness 9.702852
Kurtosis 500035.8
Geom. mean 173.7321

 Shape_Leng N 11737 Min 30.75888 Max 3734.881 Sum 2169967 Mean 184.8826 Std. error 0.7394023 Variance 6416.804 Stand. dev 80.10495 Median 169.0816 25 prcntil 138.6115 75 prcntil 211.7965 Skewness 9.702852 Kurtosis 500035.8 Geom. mean 173.7321
 Shape_Area N 11737 Min 67.93181 Max 682473.5 Sum 3.35E+07 Mean 2850.52 Std. error 63.74081 Variance 4.77E+07 Stand. dev 6905.516 Median 2095.673 25 prcntil 1406.141 75 prcntil 3276.017 Skewness 82.01563 Kurtosis 1.15E+07 Geom. mean 2186.083
• No labels