NBA TEAM PERFORMANCE AND SALARY

Name:

Course:

Instructor:

Instructions:

DateTitle: NBA TEAM PERFORMANCE AND SALARY

Introduction

This chapter outlines the research design which offers an explanation into what type of research this study is. It also defines the population of the study and the specific sampling technique used. The method of data collection is carefully detailed and the method of data analysis explained.

Research design

Research design is considered as a “blueprint” for research, dealing with at least four problems: which questions to study, which data are relevant, what data to collect, and how to analyze the results. The best design depends on the research question as well as the orientation of the researcher. The study will adopt descriptive survey as a method of collecting information by using a standardized form of interview schedule and administering a questionnaire to a sample of individuals. Both the secondary and primary data will be used in the analysis.

The design is also suitable because it gives an in-depth description of the phenomena in their existing setting. This fits well when undertaking a game level analysis of team total salary dispersion and team performance in the National Basketball Association. Descriptive survey is also preferred because it is economical in collecting data from over a large sample with high data turn over.

The model

In order to undertake a game level analysis of team total salary dispersion and team performance in the National Basketball Association, the study use the following model:

Yij=β0+ β1coaeij + β2coawij+ β3salij + Eij

Where: Yij=yi/yj is the ratio of the point scored by team I to that by team j in the game where j is at home against I.

Salij=sali/salj= is the ratio of the total team salary of team I and total salary of team j.

Coaeij=coaei/coaej is the coach experience ratio by team I and tam j

Coawij=coawi/cowj is the coach winning ratio by team I and team j

Coawi=thw/thp for team I coach winning ratio is the total game he won divided by the total game he played.

Estimation method

Estimation of the model above involves several econometric analyses. Yij is clearly endogenous variable because it depends on the number of game-minute played and hence is determined jointly with the game outcome. The measures of salary dispersion employed to the study are strictly exogenous. When coaching variables are included in the estimation, they are treated as exogenous and entered into the model as conventional instruments. Using these instruments sets, we can efficiently estimate the coefficients β1, β2, β3.

Data

The data collected covers five NBA seasons from 2006 to 2011. The set of data comprise of information on player who participated in games for those seasons. During the study, the variables are calculated from a reduced dataset that comprise the first game of each season for match-up of teams. The game level panel is a combination of 4176 unique games which is a reflection of 69% of total games played throughout the season. The sources of data for the salaries are: basketball-reference.com, USAToday.com and the independent statistician Patricia Bender. All the players had at least one salary figure quoted for each season. The box scores of the regular NBA seasons provided the game-level statistics for the period 2006-2011 sourced from basketball-reference.com.

Data Analysis Methods

The data collected will be analyzed using both descriptive and inferential statistics. Qualitative data will be analyzed using grounded theory methods. Open coding was used to initially name and categorize the data, and selective coding was used to develop a more general framework. The study will use frequency count, charts, bar graphs and percentage in data analysis. Because the study is quantitative in nature, it will be appropriate to use frequency count, mean, standard deviation, minimum and maximum values of variables

Quantitative data analysis will be done to generate frequencies percentages. The resulting quantitative data will be then interpreted using simple statistical method. The questionnaire will be coded and analyzed using Statistical Package for Social Sciences (SPSS) version 21. The data collected through questionnaire will be classified on the basis of common attributes then tallied to obtain statistical frequencies, tabulated and finally analyzed using descriptive statistics. This helps to collapse large volume of quantitative data in numerical form for ease of statistical interpretation. Frequency and distribution tables, graphs and diagrams will be integral during the process of data analysis. Comparative analyses will be used to outline the characteristics of the leading attributes within the variables under discussion.

The hypotheses formulated for this study will be tested using logistic regression model. This has been used to examine the relationship between dependent and independent variables. Logistic regression is multiple regression but with an outcome variable that is a categorical dichotomy and predictor variables that are continuous or categorical.

References

Dobson, Stephen, and John A. Goddard. The Economics of Football. Cambridge: Cambridge University Press, 2011. Print.

Fort, Rodney D. International Sports Economics Comparisons. Westport, Conn. [u.a.: Praeger, 2004. Print.

Rosner, Scott, and Kenneth L. Shropshire. The Business of Sports. Sudbury, Mass: Jones and Bartlett Publishers, 2004. Print.

Wise, Aaron N, and Bruce S. Meyer. International Sports Law and Business. The Hague: Kluwer Law International, 1997. Print.

Bonner, John T. The Social Amoebae: The Biology of Cellular Slime Molds. Princeton: Princeton University Press, 2009. Print.