february 18, 2004
Prof. Luigi Fabbris
Dep. of Statistics, University of Padua

The lecture in Video Streaming (in Italian): 
Low connection speed (56KB/s) | Intermediate connection speed(100KB/s)| High connection speed (ADSL - 300KB/s)

You will need QuickTime plugin  

Power Point slides of lecture (in Italian): download Zip file (1074KB)

Abstract:
More and more often the public opinion comes interrogated daily, through an opinion poll, about numerous questions: the use of a certain product, the approval of positions assumed by political parts in territorial or national level, the attitude regarding serious problems like social integration or the respect of the weakest. If the results of the survey give a cognitive contribution on topics where the scientific debate is still open, it can be assigned an informative meaning to them; when the results can be used for strengthening, through the manifestation of an explicit consent, guidelines or decisions assumed by a company or a government, it will be assigned a confermative meaning to them. But whichever survey deserves the same credibility?  Or better, how can we distinguish between a survey offering guarantees of scientific correctness and a survey rigged up in improper way?  This lecture wants to give a contribution on these nodal aspects addressing the attention on essential requirement that a survey should have so its outcomes could be generalize, either in reading or in interpretating them.

Prof. Luigi  Fabbris (Dept. of Statistical Sciences - University of Padova)
Ordinary university professor of Social Statistics in the Faculty of Statistical Sciences of the University of Padova.  He is Director of the Department of Statistical Sciences of the same Athenaeum.  He is ordinary member of the International Statistical Institute and associate of the Italian Society of Statistics, and also member of several national and international scientific Associations.  He is the author of books and articles about the sampling from ended populations, the analysis of multidimensional data, problems of didactical evaluation with computer-assisted systems.