<XML><RECORDS><RECORD><REFERENCE_TYPE>3</REFERENCE_TYPE><REFNUM>7942</REFNUM><AUTHORS><AUTHOR>van Biljon,J.</AUTHOR><AUTHOR>Kotze,P.</AUTHOR><AUTHOR>Renaud,K.V.</AUTHOR><AUTHOR>McGee,M.</AUTHOR></AUTHORS><YEAR>2004</YEAR><TITLE>The Use of Anti-Patterns in Human-Computer Interaction - Wise or Ill-Advised?</TITLE><PLACE_PUBLISHED>SAICSIT - Conference of the South African Institute of Computer Scientists and Information Technologists </PLACE_PUBLISHED><PUBLISHER>N/A</PUBLISHER><LABEL>van Biljon:2004:7942</LABEL><ABSTRACT> In this paper the tenability of anti-patterns in Human Computer Interaction is explored. A pattern is a generalized solution to a common problem within a context. Patterns have been accepted as being very useful in software development and more recently also in Human Computer Interaction. Anti-patterns are solutions that look promising but which contain hidden pitfalls. A concerted effort is being made in Software Engineering (SE) to identify and document anti-patterns. Patterns and anti-patterns are essentially about transferring captured expert knowledge, therefore compatibility between the nature of anti-patterns and the nature of the learner’s internal knowledge representation and processing is crucial. This paper addresses the differences and similarities between patterns and anti-patterns and how this impacts on the mental models and cognitive processing of patterns and anti-patterns. We present evidence from theories of mental modelling and reasoning that might highlight possible significant dangers in the use of anti-patterns to teach novices Human Computer Interaction principles. Given the nature of Human Computer Interaction, it is possible that the misuse of anti-patterns in the Human Computer Interaction context could at worst lead to design patterns being completely misconstrued in the first instance or recalled and used in the wrong manner. A case study on the use of anti-patterns in Human Computer Interaction teaching is presented to support this argument. Both patterns and anti-patterns have been proven as potentially successful constructs in knowledge transfer. The use of the less investigated anti-pattern in the context of Human Computer Interaction needs special consideration if the concept is to be used effectively. If the notion that the current representation of anti-patterns is not supporting cognitive processing, is correct, a new approach to structuring anti-patterns is needed. </ABSTRACT></RECORD></RECORDS></XML>