Int. J. Human-Computer Studies 68 (2010) 22–40 Can inspection methods generate valid new knowledge in HCI? The case of semiotic inspection Clarisse Sieckenius de Souza a , Carla Faria Leit ˜ ao a,Ã , Raquel Oliveira Prates b , S´ılvia Am elia Bim a,c , Elton Jos  e da Silva d a Departamento de Inform  atica, PUC Rio, R. Marquˆ es de S. Vicente 225, 22451-900 Rio de Janeiro, RJ, Brazil b Departamento de Ciˆ encia da Computac - ˜ ao, UFMG, Av. Antonio Carlos 6627, 31270-010 Belo Horizonte, MG, Brazil c Departamento de Ciˆ encia da Computac - ˜ ao, UNICENTRO, R. Presidente Zacarias 875, 85015-430 Guarapuava, PR, Brazil d Departamento de Computac - ˜ ao, UFOP, Campus Morro do Cruzeiro, 35400-000 Ouro Preto, MG, Brazil Received 29 September 2008; received in revised form 8 May 2009; accepted 21 August 2009 Communicated by Prof. E. Motta Available online 28 August 2009 Abstract HCI evaluation methods tend to be proposed and used to verify the interactive qualities of specific systems and design strategies. A discussion about the scientific merits of such methods to advance knowledge in HCI as a field is very rare, although much needed. This paper shows that, under certain conditions, inspection methods can be safely used in scientific research in HCI and extend their advantages beyond the territory of professional practice. Taking the Semiotic Inspection Method (SIM) as an example, we argue that its interpretive results are objective, can be validated, and produce scientific knowledge comparable to that generated by more widely accepted methods. & 2009 Elsevier Ltd. All rights reserved. Keywords: Semiotic engineering; Evaluation method; Semiotic Inspection Method; Scientific application; Technical application; Communicability 1. Introduction Interface inspections are widely used in HCI professional practice. When it is difficult to recruit users or deadlines are approaching rapidly, experts are often called in to evaluate the quality of interaction supported by a system’s interface. Heuristic evaluation (Nielsen and Molich, 1990; Nielsen, 1994) and cognitive walkthroughs (Lewis et al., 1990; Spencer, 2000; Blackmon et al., 2002) are well known inspection methods. Although valuable in the industry, the results of inspection methods are generally considered and explicitly referred to as expert opinions. This has probably contributed to the perception that inspection methods are questionable when used to produce and validate scientific knowledge. Empirical experiments and analytical evalua- tions based on predictive models have been much more widely accepted in this context. Even when backed by theory, most inspection methods are frequently disregarded because they can produce interpretations of reality, knowl- edge that cannot be generalized to predict how users interact with computer-based systems. This paper shows that, under certain conditions, inspec- tion methods can be safely used in scientific research in HCI and extend their advantages beyond the territory of professional practice. Taking the Semiotic Inspection Method (SIM) as a case, we argue that its interpretive results are objective, can be validated, and produce scientific knowledge comparable to that generated by more widely accepted methods. SIM is a semiotic engineering method (de Souza et al., 2006; de Souza and Leit ˜ ao, 2009) with which inspectors can analyze the communicability of interactive computer-based artifacts. It can be used technically, with the purpose of ARTICLE IN PRESS www.elsevier.com/locate/ijhcs 1071-5819/$ - see front matter & 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhcs.2009.08.006 Ã Corresponding author. Tel.: +55 21 3527 1618; fax: +55 21 3527 1132. E-mail addresses: clarisse@inf.puc-rio.br (C.S. de Souza), cfaria@inf.puc-rio.br (C.F. Leit ˜ ao), rprates@dcc.ufmg.br (R.O. Prates), sabim@inf.puc-rio.br (S. Am elia Bim), elton@iceb.ufop.br (E.J. da Silva).