Hypothesis Identification

¶ … Identification Weeks, Matthew. (May 5, 1998) "A Regression Analysis of NBA Player's Salaries." Triangle Net Journal. Retrieved on July 27, 2004 at http://www.trianglenet.net/~weeks/nba/nbasalary.htm One common assumption in economics is that salary is directly related to quantitatively established aspects of job performance and the quality of one's education and occupation.

But is this true of all professions? To take one extreme case to make a larger point, the example of NBA players can be used to see that player's statistics are not necessarily a reflection on what he is accomplishing on the job.

One might assume just as in any other business field, a worker's salary is dependent on his performance but the author of this article "used basic statistics such as minutes per game and points per game to determine" if a player's wages could be explained by actual talents, or if the salaries are more due to non-related job performance aspects such as agent's strengths and weaknesses. (Weeks, 1998) The "hypothesis going into the project was that a player's statistics have a positive affect on his salary.

We decided that minutes per game, assists per game, rebounds per game, and points per game were variables that would affect salary in a positive manner. We felt that minutes per game was important in our analysis because we felt the more minutes the athlete played, the more he should earn. Points per game were included as a variable because we thought it to be a measure of the worth of that player to the team. Rebounds and assists are also important in judging the athlete's worth to the team.

Our final analysis was that points per game would be the most significant variable." Yet, "through the use of hypothesis testing, we discovered that the model was significant at an alpha of.05. By performing individual tests, we found that the only variable that has a significant.