I have decided to compare the weight of a book to the number of pages in the book. I will only take samples from 30 paperback novels, no textbooks, picture books, encyclopedias, etc. I am expecting that books with more pages will weigh more. I am expecting a positive association between the variables, and a strong linear relationship.

Photo of data being collect

PAGES OF A BOOK VS WEIGHT OF A BOOK GRAPH

Coefficient of Determination:
r= 0.63199
r^2= 0.399 x 100= 39.9%
39.9%

Least Square Regression Equation:
weight= 6.0608015 + 0.014083323(pages)

Transformed plot (reciprocal)

Analysis:

For the most part, my hypothesis was correct. Books with more pages generally weigh more. The graph has a positive association and is linear, but not as linear as I would have expected. It is moderately strong, with the exception of a few outliers. The correlation pf my graph is .632, which shows that it is moderately strong. The residual graph is not as spread out as I would have liked, and is still quite clustered around the (0,400) region. The transformation, which is the reciprocal of the residual graph, is a little more spread out, but, again, not as spread out as I would I have liked. However, the y ranges from 0 to only .01, which is very nice. The lower left region of the graph is also not as clustered in this transformation.
The scale that I used was not completely accurate, which could have caused slight changes in the data. I also sampled conveniently, picking books that were easiest for me to get a hold of, which could have caused the data to be somewhat biased.

Photo of data being collect

PAGES OF A BOOK VS WEIGHT OF A BOOK GRAPH

Coefficient of Determination:

r= 0.63199

r^2= 0.399 x 100= 39.9%

39.9%

Least Square Regression Equation:

weight= 6.0608015 + 0.014083323(pages)

Transformed plot (reciprocal)

Analysis:

For the most part, my hypothesis was correct. Books with more pages generally weigh more. The graph has a positive association and is linear, but not as linear as I would have expected. It is moderately strong, with the exception of a few outliers. The correlation pf my graph is .632, which shows that it is moderately strong. The residual graph is not as spread out as I would have liked, and is still quite clustered around the (0,400) region. The transformation, which is the reciprocal of the residual graph, is a little more spread out, but, again, not as spread out as I would I have liked. However, the y ranges from 0 to only .01, which is very nice. The lower left region of the graph is also not as clustered in this transformation.

The scale that I used was not completely accurate, which could have caused slight changes in the data. I also sampled conveniently, picking books that were easiest for me to get a hold of, which could have caused the data to be somewhat biased.