Research Methods for Managers

Résumé du document

Since the mid 1990s, research in business and commerce areas became more qualitative: instead of using strict guidelines, researchers decided to use new techniques such as participant observation and individual interview. Since then, research has been divided into two different methods: the qualitative (the interpretivist paradigm)one and the quantitative (positivist paradigm) one. Now, we consider that both are essential to lead to a good business analysis. Some tools are provided in order to interpret this analysis. The Linear Programming, which has been developed by the Russian economist Leonid Kantorovich and the US economist C. Koopmans is a quantitative technique for optimizing an objective function, given a set of constraints. It is a mathematical technique used in computer modeling to find the best possible solution in allocating limited resources to achieve maximum profit or minimum cost. Once this data is input, we have to use the Sensitivity Analysis which provides insight into how the LP's optimal solution is affected by changes in the LP's data. In fact, the LP's data only estimates and it is essential to understand how changes in the LP data affects LP's optimal solution.

Extraits

[...] Assigning planes This exercise deals with an airline company, Air Iseg, which needs to minimize its costs. To reach this goal we had a lot of information: - Cities where incoming flights started - Cities where outgoing flights will go - Number of passengers that come from a city and go to the corresponding city - Number of passengers that start their travel at the hub and will board outgoing flights - Capacity of planes arriving to the hub Thanks to all these elements, we had to determine what the best solution is in order to save resources by filling the planes, in order to minimize the number of passengers that should change of plane, thanks to an efficient flight dispatcher. [...]

[...] Some, however, represent changes that the manager can initiate, such as enlarging capacities or adding new activities”. So, a parameter will be sensitive, if a small parameter change results in large changes in the outcomes. Moreover, the linear programming consists in three main elements: - The decision variables: the variables in a LP that represent the decisions that must be made - The objective function: the portion of a LP that specifies the linear function to be maximized or minimized - The constraints: the portion of a LP that specifies the linear equations and linear inequalities that any assignment of values to the decision variables must satisfy The objective function and the constraints are both linear. [...]

[...] So, the linear programming can also be used to solve an investment planning issue even if there are too many constraints. This scenario is a way to understand that given the limited availabilities of cash for each of the next 12 months and given government-imposed of self-imposed regulations that must be complied with, how much money should a portfolio manager invest each month in each of a variety of stocks and bonds to maximize the portfolio's annual return on investment. [...]

[...] The main lesson of this exercise is that it is sometimes interesting and profitable to combine two solutions. We understood what the Production planning scenario means: given limited availabilities of resources such as labor and machinery, how much of each of a firm's products should be produced this month to maximize total profit? Reallocating vehicles This exercise deals with a car-renting company which needs to reduce its costs per car and to decrease the costs per cars between each agency. [...]

[...] To begin we used the linear programming. Then we used the Solver but the result that we found didn't match with what we should have found: the slack (which will be defined in the 2nd question). Thanks to the solver program, we succeed and knew that the solution were to produce 11,3 computers S and 4,3 computers After having solved the problem thanks to the Solver tool, we knew that in order to satisfy all our constraint, the solution was to produce 11 ( 11.3 ) computers and 4 ( 4.3 ) computers It was a way for us to understand what the sensitivity analysis is. [...]