Matthew Walls
Pre-cal Per 5
Regression Report

Weather; is it really that important that we know how it’s going to feel outside
before we “suit up” for the day? Maybe. My two variables that I chose for this project
were time in months and average weather in Austin, Texas, USA. I chose this data
because I believe that it is interesting to try to guess the weather for the next day or
month. To do that you have to understand the physics of the earth’s tilt of 23.5 degrees
that makes the seasons of the earth, and you have to apply that to a certain latitude and
longitude and take in to consideration the high and low pressures of the air in the area and
see if there might be a storm brewing. There are a lot of things that you will have to take
into account when you try and calculate the weather and that is why I find it a very
interesting profession that I might be interested in sometime in the future. This analysis is
relevant because it shows just how much a city’s weather changes that is about 30
degrees north of the equator. This data is relevant to anybody that is in the business of
taking weather data for a living, or just to somebody who wants to know what to wear for
the day.

The regression equation is Quartic and it looks like, .0044x4 - .0747x3 - .6239x2 +
12.7631x + 38.4705. The r2 value is .8690, which is pretty good. The r2 value is a
number that represents how fitting the regression equation is to all of the data points that I
provided on a scale of zero to one. The r2 value of .8690 that I received when I put my
data into the Quartic regression shows me that 86.9% of my best fitting line goes straight



through my data points. I predict that the next three points on my graph will be as
follows, (19, 86.1), (20, 79.7), (21, 69.6). I guessed on those three points using the
previous data I recovered. Those are the months of July, August and September in the
year of 2013. I took what I the averages were from previous years and looked at the
other months in 2013 and how they compare to their corresponding months in 2012 and
made an educated guess. A career that could use this data would be a weatherman
because he would have to interpret the data and present it in a neat and understanding
way to his viewers on TV. Another Career that could possibly use this data would be a
traveling salesperson that is planning a trip to Austin Texas and needs to know what type
of clothes to pack for his trip. The y-intercept of my data stands for the average
temperature in Austin Texas.

Average weather per month will always go up and down no matter where on earth
you are. Before I started this project, Mr. Kiker told our class that out data could not be
linear, and since I know that the weather in Austin fluctuates a lot I knew that it would
not be linear, so I decided to pick data points from the average temperature in Austin over
the past two years starting with January 2012. While I was trying to find the average
temperature per month in Austin for the past two years it was extremely hard to do so
because all the weather sites would give you average high and low for a month, but never
just average temperature. In this project I have analyzed why a Quartic regression is the
best type of regression for my data, predicted the next three months average weather in
Austin Texas, and finally found several careers that could use this data.

Works Cited
Past Monthly Weather Data for Austin Texas, Weather Warehouse, 2013, web, 9-4-13
http://www.srh.noaa.gov/images/ewx/aus/attmontemp.pdf 2013, web, 9-4-13