evaluating information on the web Tefko Saracevic School of Communication, Information and Library Studies Rutgers University http://www.scils.rutgers.edu/~tefko © Tefko Saracevic, Rutgers University 1 evaluating internet resources • impossible? not really • hard? very • help? exists if you persist • LECTURE TOPICS: – how to go about it? – what are the main criteria? – where to verify? © Tefko Saracevic, Rutgers University 2 the web • fastest growing technology in history • explosive growth of WWW provided – ubiquity of information and access – but also information chaos & anarchy • growing difficulty in identifying, searching, retrieving and EVALUATING • metaphors: ‘lost in an ocean’ ‘finding pearls in garbage dumps’ ‘needle in haystack’ © Tefko Saracevic, Rutgers University 3 web is value neutral • all kinds of information can be found – misinformation • deliberate, just plain wrong or plain stupid – disinformation, censored – hate information – propaganda, spin doctored information – questionable, inaccurate, – harmful, objectionable, insulting © Tefko Saracevic, Rutgers University 4 value neutral ... • but information that is also – valid, reliable, useful, relevant, accurate, factual, timely, credible … • to a high degree • appropriate to many problems & tasks – otherwise hard or impossible to find, retrieve & access – from sources that are trustworthy © Tefko Saracevic, Rutgers University 5 prerequisite knowledge • to evaluate web information needed knowledge about – web structure & mode of operandi of the internet & domain name system – notion & characteristics of cognitive authority – criteria adapted for the web © Tefko Saracevic, Rutgers University 6 evaluated © Tefko Saracevic, Rutgers University 7 cognitive authority “influence on one’s thoughts that one would consciously recognize as proper” Patrick Wilson • related to assignment of credibility – two components: competence & trustworthiness • ascribed to particular individual, institution, organization, action © Tefko Saracevic, Rutgers University 8 problems • on the web – traditional authority indicators difficult to attribute - often absent • authorship? title? version? place of origin? • author qualification? credentials? – no filtering – vanity publishing © Tefko Saracevic, Rutgers University 9 problems ... – sometimes even attribution difficult • identity? reputation? qualifications? • can be published by anyone • anyone can claim to be somebody else • assigning credibility to Web information a BIG problem © Tefko Saracevic, Rutgers University 10 solutions • rigorous evaluation – more detailed than print sources • depending on known authority sites, authors, organizations • following many sites that did evaluation already © Tefko Saracevic, Rutgers University 11 evaluation criteria • many traditional criteria remain but with new interpretations • a number of new criteria have emerged specific to digital nature of resources & access • many are stated & can be found on the Web e.g. library sites © Tefko Saracevic, Rutgers University 12 web & cognitive authority need to carefully asses (always, web or no web): document, author, institution & affiliation on criteria of: authority accuracy currency objectivity coverage © Tefko Saracevic, Rutgers University 13 specific evaluation criteria • what & why? - documents, objects – content? purpose? scope? viewpoint? • by whom? - creators, authors, institutions – identity? authority? credibility? reputation? qualification? refereeing? • where? - affiliation, connections – identity? overt? covert? authority? credibility? reputation? © Tefko Saracevic, Rutgers University 14 criteria ... • for whom? - orientation – intended audience? needs satisfied? fit with user community? • when? - timeliness – currency? up-to-date? revisions? persistence estimate? • how? - treatment, coverage – accuracy? credibility? objectivity? style? clarity? organization? usability? © Tefko Saracevic, Rutgers University 15 criteria ... • in what way? - presentation – format? layout? interface? search capabilities? access? • how much? - economics – effort? price? cost-benefits? license? • in comparison to? - competition – other similar resources? © Tefko Saracevic, Rutgers University 16 model for evaluating information on the web • INPUT: filter & assessment for – document – author – institution – affiliation • OUTPUT: combined assessment & ascription of cognitive authority © Tefko Saracevic, Rutgers University 17 model ... input from the web assess document assess author filter assess institution assess affiliation ascription of cognitive authority © Tefko Saracevic, Rutgers University 18 examples • newspapers – New York Times; many others • governments – in the US: Census Bureau; State Department; Nat. Inst. of Standards • organizations – www consortium (w3c) © Tefko Saracevic, Rutgers University 19 examples ... • international – UN, European Union agencies • professional – Assoc. for Computing Machinery • health – Mayo Clinic; Rx list for pharmaceuticals © Tefko Saracevic, Rutgers University 20 examples ... • science – national academies of many countries • commercial – encyclopedias, reference sources – Britannica has evaluated web sites • publishers – evaluated sites e.g. Choice © Tefko Saracevic, Rutgers University 21 libraries & web authority • emerged as an important source – many provide evaluated sites & links • judicial & trusted selection: – a key value-added contribution by libraries internationally – trust extends to digital collections – makes all the difference between a library & other collections © Tefko Saracevic, Rutgers University 22 library examples • national libraries: many links – Library of Congress; UKOLN (UK) • academic libraries: great many – U of Michigan: law – U of California Berkeley: many domains – Virtual Library (Switzerland) – etc. etc. etc. GREAT sources! © Tefko Saracevic, Rutgers University 23 reference, journals • elaborate online reference sites – Martindale’s reference desk – some commercial e.g. Ask Jeeves – reference questions answered online • pathways, guides • publications - some free other licensed - licensing now a big deal © Tefko Saracevic, Rutgers University 24 specialized • digital libraries emerged in great many domains, fields – history, national memories – arts, museums, music, poetry ... – science, technology – geography, climate, weather – cooking, stamp collecting, sports ... © Tefko Saracevic, Rutgers University 25 conclusions • web sources HAVE to be evaluated • many evaluations follow traditional criteria e.g. as for news accounts • many new criteria evolved • many tools already there • hard but possible! © Tefko Saracevic, Rutgers University 26 sources • URLs for the mentioned & many other sites can be found on: http://scils.rutgers.edu/~tefko/D-Lib_Edu/ © Tefko Saracevic, Rutgers University 27 © Tefko Saracevic, Rutgers University 28