Article Preview
TopIntroduction
User interfaces for software download on the Internet are the front end of software publishers for remote users. The design of these interfaces is critical for users who may make unconscious or misguided decisions if interfaces are hard to understand and use (Brustoloni & Brustoloni, 2005; Kormann & Rubin, 2000). Problems may also arise for online publishers if hard tasks and complex interfaces reduce user trust, satisfaction and system usability at large. Software downloading also has security implications, and Brustoloni and Villamarin (2007) show that bad interface design can cause unjustified risks (breaches in the policy adopted to classify risks).
Addressing similar issues, our aim is to investigate the users’ behavior while dealing with software download interfaces. In line with Brustoloni and Villamarin (2007), we observed incoherent behaviors in users interacting with the software download interface (Dini, Foglia, Prete, & Zanda, 2006, 2007). We define an incoherent behavior as a download decision which is not coherent with the motivation given a-posteriori. In Dini et al. (2006) we observed that names of well-known (WK) publishers have a very strong brand effect that increases users’ downloads, even if users claimed they wished to download free software only. From this, we inferred that users could have problems in understanding the information in the interface: they accepted software downloads according to the publisher’s name, not the item cost. The interface was also problematic for publishers with no brand effect who had low acceptance rates. Hence, we proposed in Dini et al. (2007) a 3-step wizard interface asking for an explicit input on software cost. We observed improvements in user behaviors, with significant reductions of incoherent behaviors. However, the wizard forced users to be aware of cost and as a consequence users mainly accepted free software and refused software which entailed a charge.
This paper proposes and assesses the inclusion of a Reputation System (RS) in the common browser download interface. The RS was adopted to investigate whether it can mitigate the incoherent behaviors and increase trust in publishers with little or no brand effect.
In online service or product provisioning the client party often has little information on the service or product provider. Internet RSs collect, distribute, and aggregate feedback about past performances of a service or product provider. These systems have enjoyed widespread diffusion proving valuable in providing users with relevant information from previous users: RSs support the user in deciding who to trust, foster trustworthy behaviors, and deter malicious parties. “Reputation systems seek to establish the shadow of the future to each transaction by creating an expectation that other people will look back on it” (Resnick, Kuwabara, Zeckhauser, & Friedman, 2000). Because of such positive effects they have been adopted with success in many e-commerce systems (Chia, Heiner, & Asokan, 2010; Dellarocas & Resnick, 2003; Chen & Liu, 2011).