The Web Science Trust

Behavioral research on the WWW

Boesser, Tom (2009) Behavioral research on the WWW. In: Proceedings of the WebSci'09: Society On-Line, 18-20 March 2009, Athens, Greece. (In Press)

[img] Plain Text (preprint) - Repository staff only
[img] Plain Text (Poster Description)
PDF (Poster Artwork) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader


Research on human behavior has around 125 years background of methods development in psychology, medicine, biology, political science, economy, sociology and related fields behind it. The objective has been to establish experimental and observational scientific approaches which can produce valid and repeatable results both to provide a foundation for decisions on practical issues (engineering solutions or societal decision making), and to test hypotheses and theories in empirical sciences. All relevant scientific fields have developed their body of methods and principles (often against scientific positions which questioned the need for a rigid methodology), but it still often requires hard efforts to convince all actors of the benefit which a rigidly applied logic and methodology for empirical research provides. The question is whether we can transfer the traditional and proven empirical methodology from laboratory and field studies to the WWW, and even exploit new possibilities. This is a highly relevant question, because social and economic behavior makes growing use of the WWW, which, in addition to being a new communication medium, lets new behavior patterns emerge. There are two aspects here: Researching the behavior of WWW users, and using the WWW to research human behavior for more generic objectives, or to provide fundamental scientific understanding. Users of new technology and new applications adapt their behavior according to their abilities, goals and preferences. Understanding this process helps to shape the technology into desired directions. This is the objective of applied research, such as human computer interaction. While HCI is an established part of software development approaches, new issues have emerged as the WWW hosts increasingly innovative functionality and enables new means for cooperation: We have studied user needs and preferences, and have measured the changes in user behavior in a number of large-scale applications for collaborative knowledge management in company-internal applications using semantic technology, and in applications for media production in open collaborative virtual organizations. Results show that the workflow can be affected fundamentally by these new facilities, but we also see that the emerging behavior varies widely, and depends on specifics of the organizational context: Cooperative parameters and specific properties of the organizational culture, trust, anonymity, and the expected long-term risks and benefits determine how the individuals in an organization use new functionality, which collaborative behavior they develop, and to which incentives they respond. We pursue this line of investigation further by researching which factors foster cooperation, building on extensive results from experimental economics and human decision making. We compare the effects of trust and anonymity on cooperative behavior on the WWW and in organizations relying primarily on direct communication between individuals. Using the WWW as a research tool still raises questions regarding the validity of the results obtained, due to the limited history of behavioral research on the WWW using a rigid methodology. There are two main issues which will need to be answered: Firstly sampling methods (which are a cornerstone of valid empirical research) taking the specific problems of the selection and identification of individuals, and also reservations about privacy and use of information on the WWW into account, are hard to conceive, and we are not aware of truly satisfactory approaches. This is a problem which has existed in other research domains (such as polls or market research), and where satisfactory solutions were found, but did require extensive research and long experience. A second major issue is that research on decision making and economic theory tells us that people are fairly good at making decisions and at assessing values when these can be based on physical objects and physical variables. Where abstract objects (such as information objects) and inferred variables are concerned (such as probabilities and long-term developments), large errors of judgement and highly variable behavior are observed. We look at the differences between the behavior in well researched lab and field situations, and the decision making within WWW applications and communities, and compare this to models of rational decision making.

Item Type:Conference or Workshop Item (Poster)
Uncontrolled Keywords:human behaviour, behavioral research
Subjects:Web Science Events > Web Science 2009
ID Code:228
Deposited By: W S T Administrator
Deposited On:24 Jan 2009 08:45
Last Modified:25 Oct 2011 16:50

Repository Staff Only: item control page

EPrints Logo
Web Science Repository is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.