Gostovanje izr. prof. Daniele Toninellija iz Univerze v Bergamu

V tednu od 3. do 10. decembra 2018 je na Fakulteti za družbene vede, na programih družboslovne informatike, v okviru programa Erasmus +, gostoval izr. prof. dr. Daniele Toninelli iz Univerze v Bergamu. Dr. Toninelli sodeluje s pedagogi s Katedre za družboslovno informatiko in metodologijo že vrsto let, predvsem na področju metodologije spletnih anket, in v okviru tega so sodelovali tudi že v nekaj raziskovalnih projektih. Tokrat je na obisku na FDV za študente in širšo javnost izvedel dve predavanji: »Veliki podatki v anketah ali ankete za velike podatke?« in »Mobilne spletne ankete«.


Abstrakti predavanj (v angleščini)

“Big Data in Surveys or Surveys for Big Data?”

This is without any doubt the era of “Big Data”. On a daily basis a huge amount of data are (directly or indirectly) produced, collected and (sometimes) studied and used for various purposes (to drive policies or decisions, to optimize production processes and services supply, to evaluate products, to study customers satisfaction, and so forth). Nevertheless, so far we still do not have an official and general definition of “Big Data”, seen that the concept is also strictly linked to and varies in relation to specific contexts and fieldworks.

Starting from the paper of Callegaro & Yang (2010), this guest lecture will try to synthesize the main feature and will attempt to merge the main definitions of the phenomenon, classifying as well the main types of data (e.g. according to their source and taking into account if they are structured or unstructured).
Big data also affected the survey methodology field; survey methodologists will play a more and more important role within this scenario. This because on the one hand surveys can contribute in producing “Big data” using new advanced tools and trying to optimize the product’s quality (in terms of reliability, accuracy of the estimates, coverage and representativeness, and so forth), attempting, at the same time, to reduce collection costs. On the other hand, survey methodologists can now count on readily available sources of data (such as social networks or web scraping). These potentially interesting sources can be used not only in order to integrate data directly collected, but also in order to help planning and developing a survey project (for example for the sample selection, for a prior check of the estimates, for testing the survey tools, and so on).

These partially new roles within the research field highlight the importance of being prepared: survey methodologists should update and/or enhance their skills in order to face the new challenges of the “Big data” era.


“Mobile Web Surveys: a new (un)explored frontier”

The increasing use of mobile devices in order to complete web surveys started posing, in the last years, a lot of unexplored methodological questions for survey designers and researchers.

Nowadays they have to face with a different kind of participation, involving potentially different categories of respondents, they have to develop or update data collection tools in order to take the mobile participation into account, they have to face with new data quality challenges, and so forth. However, the mobile devices inclusion has also positive consequences for survey projects: for example, it allows to significantly reduce the costs and the length of the data collection phase, to reach respondents immediately and everywhere, to develop and use new interesting data collection features (such as sharing pictures, geo-coordinates, and so on).
If the main advantages of a mobile participation are mostly overlapping with the ones of web surveys, this new kind of involvement also caused new technical and methodological questions and issues that have necessarily to be kept into account and dealt with.

This guest lecture will start with an initial overview of the mobile web participation phenomenon and of the main consequences it includes (e.g. in terms of data quality, representativeness, confidentiality issues). Moreover, it will highlight the main drawbacks of such kind of survey participation, as well as some practical suggestions usable in order to overtake the limits and to reduce the impact of the issues linked to it.


Nazaj na seznam vseh obvestilObjavljeno: 17. december 2018 | v kategoriji: Obvestilo