The exam is divided in three parts: 1) Analysis of variance and design of experiments; 2) quality through experimental design; 3)multi-attribute valuation methods: conjoint analysis and choice experiments and modelling (logit model)
Lectur notes available on the teacher's web-site: http://local.disia.unifi.it/berni/
Levi R., Lombardo A., Vicario G., 2006, progettazione e analisi degli esperimenti, McGraw-Hill libri capp.: 4-6-7-8.
In addition, for the multi-attribute valuation methods, lectur notes will be supplied by the teacher.
Learning Objectives
The aim of the teaching course is to learn how to improve the quality of a product or a production process in a company through design of experiments, ANOVA and Taguchi's Philosophy (robust design and two-step procedure); the second aim is to give the ability for evaluating a new product or service by a consumer's point of view (Consumers' preferences) through the multi-attribute valuation methods.
Prerequisites
Statistical inference
Teaching Methods
Theoretical lecturers and laboratories, for helping the students to be able to apply the studied statistical methods
Further information
Two locations:
PIN- Prato and
Dept. of Statistics, Computer Science, Applications- DISIA (Florence) for laboratories
Type of Assessment
Oral examination.
More precisely, questions will cover all the specific topics of the course.
Particular attention is payed to the critical and constructive student's abilities; the student should show his/her ability to apply the main concept of statistical quality control methods in order to improve a product/process in a firm. Moreover, he/she will establish to be able to apply the multi-attribute valuation methods for a new product/service.
Course program
1-One-way and two-way analysis of Variance; introduction to the experimental design: trial, replication, randomization, source of variability, factor role (experimental factor and sub-experimental factor); factor at fixed levels; experimental planning.
2-Complete factorial design; interaction; fractional factorial design at two levels: Resolution, confounding effects, alias pattern, building of the fractional factorials. Fractional factorial design and robust design, also considering the Taguchi’s philosophy.
3- Customer satisfaction and multi-attribute valuation methods: conjoint analysis, choice experiments and modelling (logit model).