In addition, cognitive status was assessed by administration of the Mini Mental State Examination (MMSE; Folstein, Folstein, & McHugh, 1975). Data were analysed using one-way ANOVAs (Table 3). On each trial, two characters, a digit (1–9, except 5 and 0) and a letter from the subset A, E, I, U,
F, C, T, X, were presented. The task-relevant stimuli and task-irrelevant PF-562271 cell line distracters were counterbalanced across each trial type. Distracters were presented either to the left or to the right of the target stimulus, to prevent subjects from adopting a constant search strategy. The Rogers et al. (1998) paradigm contained no stimulus or distracter repetitions, so in the present design the task-relevant stimulus and irrelevant distracter also switched on every trial. In the alternating runs task sequence (AABB), subjects switched task on every second trial. Salient spatial cueing was employed, in the form of stimulus
position in a 2 × 2 grid (Rogers & Monsell, 1995), ensuring a cue switch on each trial and thereby unconfounding cue switches from task switches (Logan & Bundesen, 2003). The task mapping Doramapimod order within the grid was counterbalanced within groups. Since foreperiod preparation has been shown to mask parkinsonian switching deficits (Cools et al., 2003) and reduce sensitivity to frontal activation (Wylie et al., 2004), a short (300 ms) response to stimulus interval duration was utilized to maximize paradigm sensitivity to any such deficits. selleck kinase inhibitor No feedback was given. Subjects switched between categorizing a letter as a vowel or consonant, and categorizing a digit as higher or lower than 5 on every second trial, as fast and as accurately as possible, by emitting vocal responses. Successful performance required selection of the task-relevant stimulus in the face of interference from the irrelevant character in the display, the distracter, and application of the correct response rule. Similar to
our previously published study, these tasks were selected based on the following criteria: (1) the vocal responses mapped directly and naturally onto the judgment outcome (‘high/low’, ‘vowel/consonant’), (2) the vocal responses themselves comprised short vocalizations for ease of triggering the voice key, and (3) the tasks, previously piloted to address task dominance and control for asymmetrical switch costs (Allport, Styles, & Hsieh, 1994; Allport & Wylie, 2000), were relatively easy and based on well-learnt rules. The task sequence followed the alternating runs procedure of AABB, so that subjects switched between two vowel/consonant and two high/low judgments on every second trial. The probability of a response repetition was additionally controlled, since the Rogers et al. (1998) procedure contained by definition no response repetitions because responding to the target comprised vocalization of its identity and there were no stimulus repetitions.