An investigation of the mechanism of information reduction
Contemporary theories of skill acquisition emphasise qualitative changes in processing as
expertise is acquired (Anderson 1981; Cheng 1985; Newell 1990; 1998; Lewis 2001). In
particular, describing this qualitative change as a switch from calculating the answer to the
retrieval of the solutions from memory, is popular (Campitelli and Gobet, 2005; Logan
1988,2002; Nosofsky & Palmeri, 1997; Palmeri, 1997, 1999)
Against this background, Haider and Frensch (1996, 1999a, 1999b, 2002) have recently
identified some of the quantitative changes that may also occur with practice, changes
which they term Information Reduction (IR). They demonstrated that people could 'reduce'
to processing task-relevant segments of a stimulus, without instruction to do so. Further
they found this effect was not stimulus-specific, transferring to novel item sets. This latter
point was particularly troublesome for any theory reliant on the retrieval of exemplars from
memory, since such a strategy will become unsuccessful for novel items.
The work presented in this thesis further explored the factors that may playa role in IR.
Study 1 both replicated the basic effect and also found reduction when the visual regularity
of the stimulus was varied. Study 2 realised a new target search task (TST) in which IR also
developed, generalising the strategy beyond the original alphabet arithmetic task (AAT).
The third study of the thesis investigated further attributes that could inhibit or facilitate
reduction. The final study determined the regularity of task-redundancy necessary for IR to
take place. The results are discussed in terms of the residual processing of task irrelevant
items and the overall part IR must play in skill acquisition.