This study aimed to investigate the validity of psychological experimental methods within human-computer interaction research (Carroll, 1989) and to examine design guidelines pertaining to hypertext link colour and positioning of navigation menu frames as part of web documents. The results of past research on both link colour and positioning of menus are mixed and guidelines are usually not based on empirical evidence (Tullis, 1997; Shneiderman, 1997). The study used a repeated measures experimental design. Participants carried out both a visual search task and an interactive search task. Task performance on the two tasks did not to correlate (p > 0.05), indicating that the visual search task may lack external validity. Results of the interactive search task suggest that the design convention of blue links (Nielsen, 1999a) should be retained as responses for blue were found to be significantly quicker than red, F(1, 117) = 14.526, p <0.001, MScolour = 89.866. Furthermore, an effect of presentation position, F(3,117) = 8.410, p <0.001, MS position = 61.015, was found, with support for menus on the left (Nielsen, 1999a; Campbell & Maglio, 1999) or right (Nielsen, 1999a). Evidence was also found to support the conjecture that experienced Internet users might have formed automatic attention responses to specific web page designs. The need for validation of behavioural and psychometric methods with task performance and the use of cognitive-perceptual-motor modelling are discussed.