Перегляд вихідного коду сторінки Tricks Of GSK J4 : Ideas On How To Crank ErbB In Half A Second

The subject wore a [http://en.wikipedia.org/wiki/ErbB ErbB] portable eye-tracking system, the EMR-9 (NAC, Tokyo, Japan) while standing on the WBB. The subjects were allowed to move their body freely instead of their feet on the WBB. A flat marker was placed on the shelf for later image processing of the  EMR-9's head-mounted camera. The height of the shelf dh and the distance between the shelf and WBB dw were kept constant throughout the experiments. A webcam was installed to synchronize the data from the WBB and EMR-9. The whole setup was covered by thick cloth to avoid presenting unwanted visual stimuli to the subjects. Two normal lamps above the shelf were installed to control the light level. 2.3. Experimental protocol Each subject had 12 trials each of which had a different pair of items. They were asked to select one of the pairs with no time restriction. The response time was defined as the duration from when the subject started observing the items to when the subject made a decision. To know the decision timing, we asked the subjects to point to the item they preferred with their finger. Unlike previous studies, we did not employ a physical button, in order to eliminate unnatural choice behaviors that would not occur in a real-shop situation. Before each trial began, the whole setup shown in Figure ?Figure11 was kept dark with the lights turned off. The subject sat on the chair in front of the WBB, and was asked to stand up on the WBB's footprint  mark and look straight ahead. The subjects were asked to pay no attention to the shelf top before each trial's start. A trial began when the operator turned the light on and started recording all the signals (WBB, webcam and EMR-9). At the same time, the subject started looking at the pair of items. The operator monitored the subject outside the booth with the webcam until the subject pointed to either of the items. Finally, the operator told the subject to sit back down on the chair, turned the lights off and changed the items for the next trial. Each subject performed 12 trials executed in a pseudo-random order as shown in Table ?Table1:1: three snack trials, three apparel trials, then the same numbers of trials in the same order.  The trial categories were fixed, but the pairs of products were randomized with no duplication. Also, we did not fix the sides of the alternatives to avoid bias stemming from left-right alignment. Table 1 Pseudo-random item categories and WBB angles in the 12 trials. We manipulated the angle of the WBB to study the effect of body-orientation on final choice (Figure ?(Figure2).2). We tested three angles, left, right, and straight, relative to the longer edge of the shelf, and we switched the angle after every trial as shown in Table ?Table11. Figure 2 Body orientation manipulation: the WBB was set at three different angles. Circles and squares in the figure represent the markers placed on the floor.