Overview of the Course (Lecture 1-2) Sumanta Pasari Assistant Professor, BITS Pilani Department of Mathematics, Pilani Campus BITS Pilani, Pilani Campus Books
T1: David R Anderson, Dennis J Sweeney, Thomas A Williams,
Jeffrey D. Camm and James J. Cochran, Statistics for Business and Economics. Twelfth edition. Cengage Learning. 2014
Reference Books: R1: Deepak Chawla and Neena Sondhi, Research Methodology. Second Edition, Vikas. 2016 R2: Richard Johnson and Dean W Wichern, Applied Multivariate Statistical Analysis, Pearson, 2007 (online pdf available) 3 BITS Pilani, Pilani Campus Evaluation Scheme
Component Duration Weightage
Mid-sem 90 min 30% Quizzes (two) 15-20 min 10% Experiential Learnings (two) About a week 20% Comprehensive Exam 120 min 40%
4 BITS Pilani, Pilani Campus
Course Objectives 1. To provide exposure to various theoretical univariate/multivariate methods and their practical applications. • Probability distributions • Statistical Inference (one population and two populations) • ANOVA (Analysis of Variance) • Regression and correlation • Discriminant analysis; factor and cluster analysis • Time series analysis and forecasting • Nonparametric methods and statistical quality control
2. To be familiar with MS Excel to solve a variety of problems (Students
are encouraged to learn ‘R’ by themselves)
5 BITS Pilani, Pilani Campus
Examples
6 BITS Pilani, Pilani Campus
Examples
7 BITS Pilani, Pilani Campus
Examples Dullus County Schools is renewing its school bus service contract for the coming year, and must select one of the two bus companies, the Milbank company or the Gulf Park Company. The variance of the arrival of pick-up/delivery times will be used as a measure of quality service. Consider sample data collection: n1 26, s12 48 and n2 16, s22 20. Formulate a hypothesis test H 0 : 12 22 against H1: 12 22 . Test at 0.10.
For a public opinion survey, a sample of 41 women and 31 men is used to study the attitudes about current political issues. The researcher conducted this study to see whether the sample data indicate that women show a greater variation in attitude of political issues. Consider sample data collection: n1 41, s12 120 and n2 31, s22 80. Formulate a hypothesis test H 0 : women 2 men 2 against H1: women 2 men 2 . Test at 0.05
8 BITS Pilani, Pilani Campus
Examples J.D. Power and Associates use the proportion of owners likely to re-purchase a particular automobile as an indication of customer loyality for the automobile. (a) p1 proportion likely to repurchase Impala for the population of Chevrolet Impala (b) p2 proportion likely to repurchase Fusion for the population of Ford Fusion (c) p3 proportion likely to repurchase Accord for the population of Honda Accord Ho : p1 p2 p3 Ha : Not all population proportions are equal Automobile Owners Likely to Impala Fusion Accord Total repurchase YES 69 120 123 312 NO 56 80 52 188 Total 125 200 175 500
9 BITS Pilani, Pilani Campus
Examples A beer industry association conducted a survey to determine the preferences of beer drinkers for light, regular, and dark beers. A sample of 200 is taken. The survey also considered the gender: male or female drinkers. A research question arises, whether the beer preference and gender are independent. Ho : Beer preference is independent of gender Ha : Beer preference is not independent of gender
Gender Beer Preferences Male Female Total Light 51 39 90 Regular 56 21 77 Dark 25 8 33 Total 132 68 200
10 BITS Pilani, Pilani Campus
Examples
11 BITS Pilani, Pilani Campus
Examples (ANOVA)
12 BITS Pilani, Pilani Campus
Examples (ANOVA)
13 BITS Pilani, Pilani Campus
Examples (ANOVA)
14 BITS Pilani, Pilani Campus
Examples (Regression Analysis) Mr. Agrawal, the Marketing Manager of the S P Pickles Pvt. Ltd. was wondering about the reasons for the decline in the sale of the company’s pickle for the last two years. He called a meeting of his team to discuss the possible reasons. The members suggested that it may be worthwhile to list the variables that influence the sale of the pickles. They listed the average price of the pickles sold by them, the competitor's average price, consumer’s income, taste and preference, and the amount spent on advertisement. Having done so, they were wondering what to do next. How can they determine the important variables influencing the sale of their pickles? What is the relative contribution of different variables?
15 BITS Pilani, Pilani Campus
Examples (Discriminant Analysis) Mr. Ghosh owns a restaurant named Rasoi, which serves Indian and Chinese cuisine. The restaurant is more than 20 years old, located in a posh locality of Delhi and caters to upscale consumers. About three years back, another restaurant came up in the vicinity of Rasoi. In the beginning, Mr. Ghosh didn’t observe any significant impact of the competition. However, with the passage of time, the clientage of Rasoi declined sharply. Mr. Ghosh wondered the possible reasons for this. He wanted to figure out the variables that differentiate between the choice of Rasoi to that of the competition. He also wanted to know the relative importance of variables in discriminating between the choice of Rasoi to that of the competition. He was wondering if it was possible to understand (predict) whether a prospective customer would choose Rasoi or not.
16 BITS Pilani, Pilani Campus
Examples (Factor Analysis) In a PGDM programme, a business school usually receives 5000-6000 applications per year for their intake of 120 students. However, this year, it received only 1200 applications. The admission manager noticed that there is a 20% decline in CAT registration this year, however, the decline to their PGDM course was much more.
A faculty meeting is called to discuss the possible cause of sharp decline
in applications. After several sessions, it was decided to conduct a survey of prospective students to find out what makes them choose a business school for PGDM programme. A random sample of 200 respondents was chosen to fill up a questionnaire. About 70 variables were found. But how to draw inferences from the same, as many variables seem to be interrelated? A factor analysis may be worthy to detect the factors that could be extracted from these variables. 17 BITS Pilani, Pilani Campus Examples (Cluster Analysis)
18 BITS Pilani, Pilani Campus
Time Series Analysis – The Art of Forecasting
19 BITS Pilani, Pilani Campus
Example: Nevada Occupational Health Care NOHC is a privately owned medical clinic in Nevada. The clinic specializes in industrial medicines. Operating more than 20 years, the clinic had been in a rapid growth phase. Monthly billings increased from $57000 to more than $300000 in just 26 months, when the main building burned to the ground. Seven months it took to rebuild. The clinic’s insurance policy covered physical property and equipment as well as loss of income due to the interruption of regular business operations. However, there was no established rules to explain “what would have happened” to the clinic’s buildings if the fire had not occurred.
BITS Pilani, Pilani Campus
Examples (Non-Parametric Tests) John and Neha were working in a research firm as management trainees after completing their MBA from a top business school in India. Their first assignment was a perception study of a high-class restaurant. As part of this study, a questionnaire was designed. Some of the questions were on nominal scale like gender, marital status, profession, age group and income group. There was an ordinal scale question where the respondents were asked to rank various attributes like food quality, food variety, ambience, price and location of the restaurants. John and Neha found that the data on these variables did not follow a normal distribution. They also realized that similar situation may arise from other qualitative research problems. They had learnt that it was either necessary for the population to follow normal distribution or the sample size had to be large enough to perform any standard test of significance. However, in the case of nominal or even ordinal scale data, sometimes normality assumption does not hold true. They wonder how they can relate the perception about the various attributes of the restaurants with the variables? 21 BITS Pilani, Pilani Campus Statistical Quality Control
• Statistical process control (SPC) is a method of quality
control which employs statistical methods to monitor and control a process. This helps to ensure that the process operates efficiently, producing more specification-conforming products with less waste. SPC can be applied to any process where the "conforming product" (product meeting specifications) output can be measured. Key tools used in SPC include run charts, control charts, a focus on continuous improvement, and the design of experiments. An example of a process where SPC is applied is manufacturing lines.
BITS Pilani, Pilani Campus
Statistical Quality Control
• In a manufacturing context, no matter how carefully machines are
calibrated, environmental factors are controlled, materials and other inputs are monitored, and workers are trained, diameter will vary from bolt to bolt, some plastic sheets will be stronger than others, some circuit boards will be defective whereas others are not, and so on. • We might think of such natural random variation as uncontrollable background noise. • There are, however, other sources of variation that may be attributable to contaminated material, incorrect machine settings, unusual tool wear, and the like. • These sources of variation have been termed assignable causes in the quality control literature. • Control charts provide a mechanism for recognizing situations where assignable causes may be adversely affecting product quality. BITS Pilani, Pilani Campus
