For my project, I will be assessing the question: what better predicts the salary of a current NBA player, total points collected while in the NBA, or total minutes played while in the NBA? Points collected and time played will be my two explanatory variables, while salary will be my response. My hypothesis is that the total points collected will better predict the salary of a current NBA player. To collect my random sample of current NBA players, I gave each player a number 1-568 alphabetically by using: http://stats.nba.com/#!?GameScope=Season&SeasonType=Pre%20Season. I then used a random numbers generator, which was supplied by my calculator, to collect 30 random numbers, skipping repeats. Each number correlated to a specific player, and I then collected each chosen players salary from the website: http://espn.go.com/nba/salaries. After that, I collected …show more content…
R in this data set is 0.043, this means that the 4.3% of the variation in salary is accounted for by the linear model relating salary to points scored by an NBA player. The equation for the least squared linear regression line is ŷ=3,239,630+131.887·x, meaning that for every average point made, salary should increase by $3,239,630. The slope for this set of data is 275.175, meaning that an NBA player's salary is predicted to go up by $275.18 as points increase. The standard deviation/average distance between the residuals and the predictions is about $4,738,380 meaning this scatterplot is somewhat accurate. The y-intercept of the data is $3,362,050, meaning if a player scored 0 points for a season, he would still receive $3,362,050 for the season; this is not very accurate because although players are sometimes guaranteed a certain amount money when drafted, the value varies for each