Statistics & R Progress Blog

 The main goal of this project was to observe price changes of certain videogame titles released for the Nintendo GameCube starting from 2007 and up to 2022, and try to predict how pricing trends will continue through the future. To start off, I first made a list of eleven games to study closely, which in the order examined are: Sonic Mega Collection Need for Speed: Most Wanted Shrek: Extra Large SpongeBob SquarePants: Battle for Bikini Bottom LEGO Star Wars: The Video Game Super Monkey Ball TimeSplitters 2 Turok Evolution Geist Batman: Dark Tomorrow Donkey Kong Jungle Beat Then, I used a price-recording website known as PriceCharting in order to get the average value of loose copies of each of these games for every year starting from 2007 and ending at 2022. Each price was picked from every December of each year. Here are the results I gathered in the CSV file used for the coding portion of this project: Once the data was collected, it was time to code. I used the ggplot2 library ...

 With R, we can construct time plots that give us a visual glimpse at data that changes over time. In this particular example, we can see how a student's credit card bills differ in value from month to month for both 2012 and 2013: However, this graph doesn't tell us a whole lot aside from showing the variation between months. With the Exponential Smoothing Model, however, we can take a closer look at each year's variations separately, and get both an average measure of change through the year as well as the mean of all the bill payments per month. Boy, according to these average rates of change, looks like this student's got quite the spending record, huh? Jokes aside, both the ESM and time plotting in general are both very useful tools for visualizing data over a span of time, and can be applied in potentially supporting plans for future financial decisions by estimating a proper budget.

 This week's assignment mainly deals with practicing working with more advanced forms of managing and working with data. 1.  The varying sizes of each group causes the period and treatment tests order-dependent. In more layman's terms, the results of the tests change depending on how the values for both period and treat are arranged in the equation. The ANOVA results are somewhat inaccurate as a result of this variance. 2. The only singularity present in these model tests occurs in the last test (z ~ b * (x+y)). In this model, both x and y are proportional with each instance of B. R is incapable of recognizing this, resulting in coefficient values of N/A. The main issue with this is that R is not able to detect a singularity if a main effect (b in this case) is affecting the categorical variables present in the equation.