It revises the bottleneck theory and indicates that instead of being eliminated, the unattended stimuli are tuned. The dichotic shadowing task is again referenced supporting this. Cherry (1953) in her variation of this experiment, found that semantic information could still be processed in the seemingly unattended ear. Similarly, Treisman, (1964) observed that unattended stimuli could be processed even after delay. She proposes serial processing of information, beginning with basic perceptual processing, in turn trigger further semantic processing. However, this raises the problem of insufficient processing capacity for unattended stimuli as well as implicating late-selection. The feature-integration task is therefore introduced to counter this. In this task, participants were asked to identify a target among a set of distractors. Treisman and Gelade, 1980 proposed that multiple features are processed in tandem by multiple feature maps in the perceptual process. These are then combined to form a whole picture of the target. One problem however, is that only six subjects were recruited for the studies. In addition, some of the participants had been involved with previously similar experiments. Also, the experiment was spatially concentrated, which could be argued does not account for whether these processes could be replicated for with auditory …show more content…
Treisman’s (1964) attenuation model conversely suggests that unattended information is instead moderated although both models concede early selective attention. However, Treisman et al., 1980 like other literature presented in this review, employed a low sample of participants, which could affect results analysis. In addition, some participants were already familiar with the research subject, possibly contributing to the outcome of the results, considering research that suggests practice contributes to performance (Stroop, 1935). In contrast, the Deutsch and Deutsch model adopts the late-selection theory, although this is countered by the studies that prove early neural activation to stimuli. Another problem, especially regarding O’Craven et al., (1999) is the seemingly lack of screening for participants’ sex, IQ or mental health which may account for performance in the experimental tasks. This is proposed for future consideration. Also, spatially-oriented experiments used to support semantic processing could be problematic due to the ambiguity in interpreting these spatially-oriented task results. Regrettably, considering imposed word limits, this review does not provide a more critical evaluation of the models. However, review of the literature presented provide evidence that although imperfect,