For my mini propsal I plan of taking 30 to 40 leaves and measuring the length and width. From that data I will be able to figure the average length and width of a leaf and will be able to compare the two. Then I will be able to see the association between all the different types of leaves in my back yard.

I suspect that the width and the length of ht leaf will be a positive, strong, and linear association as well. I will collect all the leaf from my yard not in a certain area specifically but they will not be spread out away from the tree they fell from if that makes sense.

Data collection-
Screen Shot 2017-11-23 at 11.27.34 AM.pngIMG_4872-2.jpgIMG_4876-2.jpg
Screen Shot 2017-11-23 at 11.43.47 AM.pngScreen Shot 2017-11-23 at 11.45.57 AM.png
Screen Shot 2017-11-23 at 11.53.29 AM.pngScreen Shot 2017-11-23 at 12.20.33 PM.png
X-Values vs Residuals

My data doesn't have a very strong association as predicted in my mini proposal as well as not having a very strong association my data isn't very linear. There is a little bit if a pattern in the data. The residuals compared to my regression result had very different data from the collation coefficient to equations that are displayed in the data. In the scatterplots of the data you can still see the curve in the data between 3 and 4 but the lower points in the residual vs Width scatter plot the lower 8 points are arranged differently.

Screen Shot 2017-11-23 at 12.30.33 PM.png
I think that if we were to add points in the places of the red dots that it would pull the data up and make the graph more linear.

Confounding influences could have been that the data wasn't collected from trees that were all the same but were scattered around my yard. Perhaps collecting the data from a set of trees that were all the same would have created a more linear model and you wouldn't need an transformation points.