Reach Methods For Managerial Decisions

Simulation: Research Methods for Managerial Decisions

Introduction

Over a year has past since CoffeeTime entered into the coffee bar market in Mumbai, India. CoffeeTime's entrance has proven successful with the Mumbai, India outlet reporting profits. As the company moves forward with their business plans in the Indian market, management wants to make sure they are making informed decisions that will be in the best interested of the company such as predicting revenues, introducing new products and maximizing growth opportunities, profits and minimizing risk and losses. Therefore, CoffeeTime's management wants to apply different statistical procedures such as multiple regression, Z test for proportions, the chi-square test in order to make informed and appropriate business decisions to meet two key goals (1) strengthen presence in India vis-Ðo-vis competition from Quick Brew (a local coffee bar), and (2) introduce a new snack to suit the taste of Indian customers identified in this year's business plan.

Statement of the Problem

Over the past six months CoffeeTime in continuing their efforts to gain market share and set itself up firmly in the Indian coffee bar market and Quick Brew attempting to not lose ground has created advertising and promotions battle between the two companies. Additionally, CoffeeTime conducted a survey about sandwiches that created concerns about the opportunities and risks associated with the introduction of a new a product such as customer preferences and profitability. Therefore in order to help CoffeeTime stay focused on and achieve their organizational goals, statistical research models using multiple regression, Z test for proportions, and the chi-square test will be utilized to formulate strategies, make decisions and address the major research concerns in the simulation of (1) predicting weekly revenues (2) determining potential opportunities, losses and risks associated with the introduction of sandwiches,(3) determining if our customer's preference for sandwiches depends on their gender, (4) Explain the differences in Laura Jones selections for multiple regression first using all normal values and then using all lagged values and how CoffeeTime could further optimize this model., and (5) using a 0.05 significance level (alpha) test Laura's claim of 10% of tourist will include a visit to a cafй.

Research Methods, Analysis and Results

Through a series of three tasks in the research methods for managerial decisions simulation I made selection based past data on CoffeeTime's weekly revenues, weekly advertising expenditure, price index (last 24 week period), and estimate on Quick Brew's weekly advertising expenditure.

1. Using past data on CoffeeTime's weekly revenues, weekly advertising expenditure, price index (last 24 week period), estimate on Quick Brew's weekly advertising expenditure, and input for Laura Jones, build an optimal multiple regression model.

Regression Selection and Results:

Run 1

Build a multiple regression model for predicting weekly revenues

CoffeeTime's weekly advertising expenditure _x_ Normal (X1) _x_ Lagged (X4)

CoffeeTime's price index _x_ Normal (X2) _x_ Lagged (X5)

Estimates on Quick Brew's weekly advertising expenditure _x_ Normal (X3) _x_ Lagged (X6)

Regression Statistics

Multiple R 0.869 Regression Coefficients

R Square 0.756 X1 4.525 X4 5.729

Adjusted Square 0.670 X2 695.443 X5 621.636

Standard Error 26,483.26 X3 0.039 X6 1.742

Observations 24 Intercept 245,632.957

Multiple Regression Equation:

^

Y = 245,632.957 + 4.525X1 - 695.443X2 + 0.039X3 + 5.729X4 - 621.636X5 - 1.742X6

The results showed the decisions made to be good resulting in an optimal RІ value based on my multiple regression model is 0.765 which means that 76.5% of the variation in predicted weekly revenue is explained by CoffeeTime's weekly advertising expenditure, CoffeeTime's price index, estimate of Quick Brew's weekly advertising expenditure, CoffeeTime's weekly advertising expenditure (Lagged), CoffeeTime's price index (Lagged), estimate on Quick Brew's weekly, and advertising expenditure (Lagged). It was also offered that when there is a high degree of correlation between two variables if is good the remove one, therefore the value X3: 0.657 should be removed since it is the one that is correlated with a greater number of independent variables in the model.

2. Using financial information in the trade off matrix to conduct a hypothesis testing and the Z-statistic as applied to proportions in order to determine whether 30% of the population would like the sandwiches. Based on the analysis decide whether CoffeeTime should introduce the new spicy sandwiches.

Selection and Results:

Hypotheses & Test

Null hypothesis Ho: p < 0.30

Alternative hypothesis H1: p > 0.30

Type of Z test one-tailed test

z-score = - 1.656

Region of rejection

Region of non-rejection

Optimal Choice

Null hypothesis Ho: p < 0.30 Critical Value Z 0.840

Alternative hypothesis H1: p > 0.30

Type of Z test one-tailed test