1. Course Description
The behavior of variables relating to product, service and financial markets is determined by the joint and simultaneous operation of a number of economic relations. The statistical/mathematical models summarize such relationships amongst the variables in the aforesaid markets. They are used to explain the behavior of such variables, estimate and forecast the future movements of the variables under study. In such circumstances the manager is confronted with the task of decision making by using several econometric/statistical models which capture the crucial features of the economic sector or system. Moreover, for a decision making process a manager takes a holistic view of the economy/market. This leads to the understanding of simultaneous operation/interaction of the variables which form the basis of multivariate analysis.
2. Course Objective
The Course viz., Advanced Methods of Data Analysis (AMDA) seeks to sensitize the students on various methods of data analysis especially the multivariate data analysis methods and the uses of the integrated approach of these methods in the analysis of market data. It focuses on the use of multivariate analysis and their applicability in the economy/market with respect to both dependent and interdependent techniques. A few econometric techniques with respect to the single equation model for data analysis have also been included in the course.
The basic econometric problems in estimation and their possible remedies, use of qualitative variables in the single equation methods under dependence methods along with the application of a few multivariate techniques from the interdependence methods form the contents of a course.
The statistical package viz. SPSS will be used in this course while fitting the models to the data sets. For regression in panel data E-Views will be used. Similarly, PLS/AMOS will be used for SEM. Emphasis will be more on developing skill to interpret the statistics obtained from the use of the packages relating to the different multivariate methods for decision making. The students may also use a package like STATA for arriving at the statistics relating to the multivariate tools.
A brief review of the forecasting methods will be discussed. Analysis of time series data with the single equation models will focus on estimation and forecasting along with analysis of forecast errors.
The course will be a project-oriented course so as to give the participants an opportunity to use the different techniques to analyze market data and interpret the results, acquaint themselves with the related problems in estimation and their applicability in the realistic situation.
3. Student Learning Outcomes
After going through this course the students will be able to
· Expose the students to understand the theory and application of a few multivariate statistical techniques from both dependence and interdependence methods for data analysis. · Acquaint the students with the problems in estimation.their remedial measures and the applicability in the realistic situation particularly in the single equation model using both cross section and time series data. · Select suitable multivariate techniques for data analysis in several situations and for different types of data relating to the markets. · Use a few forecasting technique and compare these with respect to forecasting efficiency. · Learn the skill to use statistical packages like SPSS, PLS/AMOS and E-View packages and interpret the output of the packagess for decision making.
There are a few books on the techniques of data analysis. The following books will be very useful for understanding the concepts, fitting the data in the models and interpreting the statistics.
5. Tentative Session Plan
I. Dummy dependent and independent variables and their uses in cross section and time series data II. Discriminant Analysis III. Binary Choice Models (Logistic Regression)
Evaluating Forecasting Efficiency using time series data.
Integration of dependence and interdependence multivariate methods for data analysis and market/buyer segmentation
6. Pedagogy:
The course is divided into two parts i.e. (i) the theoretical concept of the multivariate techniques (both dependence and interdependence) and (ii) the application of the techniques and interpretation of the statistics in the realistic situations. The participants will make use of SPSS and other statistical packages as mentioned above to analyze the data. The theoretical concept relating to the techniques will be dealt in the lecture sessions and the participants shall have to apply the techniques by using data collected by them.
The presentation and discussion sessions will be used to acquaint the participants with the interpretation of different statistics of the multivariate tools for drawing inferences.
The exercises will focus on group work during the term. The groups will collect data for the group exercises and use those in the multivariate techniques and present it in the class for discussion( Secondary Data set in a few cases will be supplied to the groups and / or the groups will be advised to consult relevant sources of secondary data).
The groups will submit a final project report within two weeks from the date of completion of the term. The final report will incorporate all the exercises and the participants will take note of the points made and feedbacks given during the presentation.
5. Evaluation
Class Assignments & Presentation 40 %
End Term 40 %
Final Project (A Summary of the Exercises after the feedback) 20 %
6. Academic Integrity: It is expected from all the students to maintain decorum and discipline in the class and the examination hall. Any kind of academic dishonesty and indiscipline viz. exchange of remarks in the class; exchange of notes, gestures or glances at another student’s paper in the examination hall will lead to disciplinary action against the person involved in such dishonest activities. The volume of disciplinary action depends on the gravity of the offence.