27598 - Statistical Methods for Business Analysis

Learning outcomes

This course provide a practical introduction to the issues involved in the design and analysis of sample surveys planned to support firm decisions.
The main learning goals are: to familiarize students with several of the well known statistical sampling methods for finite populations; to familiarize students with the planning of quality control systems.

Course contents

PART A
Inference in finite population sampling: design based inference; simple random sampling with and without replacement; unequal probability sampling; the Horvitz Thompson estimator. Stratified random sampling. The estimation of the sample size. Cluster sampling. Quotient  estimator.

PART B
Statistical process control: Shewhart control charts; control charts for variable; control charts for attributes; the role of the power function and the Average Run Length (ARL) function on the design of the control charts; CUSUM and EWMA charts.
Acceptance sampling: acceptance sampling for attributes; the Operative Curve (OC) of a sampling plan; simple sampling; rectification schemes.

Readings/Bibliography

PART A
Sharon Lohr, “Sampling: design and analysis”, Pacific Grove, Duxbury press, 1999.

PART B and PART C
D.C Montgomery, Statistical quality control, McGraw-Hill, 2007.

Some additional readings will be found at sito http://www2.stat.unibo.it/trivisano/.

Teaching methods

Each topic covered in the lectures will be followed by exercises in practical classes.

Assessment methods

Written test and oral examination

Office hours

See the website of Carlo Trivisano