Running Head Case: Select Collections Case: Select Collections Webster University Prediction to collect money from delinquent accounts David Cathalina Business Statistics 1 Running Head Case: Select Collections Introduction: My research question will be whether the model is predicting accurately the variable, so that the company might run into profits? Select Collection, Inc. purchased distressed consumer debt from major credit card companies like Chase, Bank of America and so on. It used data-driven decision making and dynamic value approach to optimize the collection processes for each purchased accounts. For each account, the first decision to make whether to resell or attempt to collect. For that purpose, it wants to build a model to predict how much money it will collect from such delinquent accounts. The model will help them to decide which accounts to purchase and how much to pay. For model building purpose, the linear regression technique and ANOVA test have been followed. Data description: There are two datasets: training and test. Both of them consist of 3570 delinquent accounts. The model will be developed on training dataset and validated on test dataset. The dependent variable of interest is totalpay i.e., the total amount of payments received from the accountholder (in US dollars). Independent variables of interest are as follows: cobal : Balance of the account at the point of \"charge-off\Structure of Final Paper Please note that these are just general guidelines; you do not have to follow everything religiously. If you feel that some issues could better fit in other sections than the one specified below, you are free to reshuffle things. You are also encouraged to include additional issues not mentioned below. (1) Introduction: Introduce and formulate clearly a central research question. You can also have several (sub)questions, if appropriate. Explain the relevance and importance of the research question. Include any other issues that could help us understand the issue in a broader context and make your paper more appealing. Explain briefly which approach you will be following in the paper and why. (2) Literature review: Cite properly a few most relevant articles on the subject. No Wikipedia cites please. You might want to do a search in databases, such as Ebscohost and/or ABI/Inform. When reviewing the literature, try to find papers with opposing points of view or contradictory findings. This will make the paper more interesting. Cite in APA-style. (3) Data description: Introduce your dataset. Identify clearly your dependent variable and explain the choice of independent variables. Present basic descriptive statistics of the main variables accompanied by appropriate graphs. Form confidence intervals around the means (proportions) of the main variables of interest. Present some preliminary evidence on relationships between variables such as crosstabulations, correlation coefficients and/or scatter diagrams. Identify and explain any outliers. Identify and interpret any missing values of the variables. (4) Hypotheses and methods of testing: Clearly formulate your hypotheses. Explain how you are going to test them and why you have chosen that test or method. (5) Data analysis (this is the crucial section): Start by performing some basic statistical tests of your hypotheses: t-tests and/or ANOVA. Perform regression analysis. Present and interpret the regression coefficients both statistically and substantively. Compare your results to other findings in the literature reviewed in (2). If you use more than one regression model, for each model report the quality of the model fit (R2) and interpret the differences between models. When presenting results of regression (and other statistical) analysis, you should not copy the tables directly from Excel; preferably, the results should be presented in APA style. (6) Write the discussion and conclusion section Summarize the main findings. Draw conclusions and implications from your findings. Describe the limitations of your analysis and explain what could be done to improve it. (7) Put APA bibliography followed by appendices at the end of the paper