Perceptual grouping by similarity of surface roughness in haptics: the influence of task difficulty.

We investigated grouping by similarity of surface roughness in the context of task difficulty. We hypothesized that grouping yields a larger benefit at higher levels of task complexity, because efficient processing is more helpful when more cognitive resources are needed to execute a task. Participants searched for a patch of a different roughness as compared to the distractors in two strips of similar or dissimilar roughness values. We reasoned that if the distractors could be grouped based on similar roughness values, exploration time would be shorter and fewer errors would occur. To manipulate task complexity, we varied task difficulty (high target saliency equalling low task difficulty), and we varied the fingers used to explore the display (two fingers of one hand being more cognitive demanding than two fingers of opposite hands). We found much better performance in the easy condition as compared to the difficult condition (in both error rates and mean search slopes). Moreover, we found a larger effect for the similarity manipulation in the difficult condition as compared to the easy condition. Within the difficult condition, we found a larger effect for the one-hand condition as compared to the two-hand condition. These results show that haptic search is accelerated by the use of grouping by similarity of surface roughness, especially when the task is relatively complex. We conclude that the effect of perceptual grouping is more prominent when more cognitive resources are needed to perform a task.

Rewarding imperfect motor performance reduces adaptive changes.

Could a pat on the back affect motor adaptation? Recent studies indeed suggest that rewards can boost motor adaptation. However, the rewards used were typically reward gradients that carried quite detailed information about performance. We investigated whether simple binary rewards affected how participants learned to correct for a visual rotation of performance feedback in a 3D pointing task. To do so, we asked participants to align their unseen hand with virtual target cubes in alternating blocks with and without spatial performance feedback. Forty participants were assigned to one of two groups: a 'spatial only' group, in which the feedback consisted of showing the (perturbed) endpoint of the hand, or to a 'spatial & reward' group, in which a reward could be received in addition to the spatial feedback. In addition, six participants were tested in a 'reward only' group. Binary reward was given when the participants' hand landed in a virtual 'hit area' that was adapted to individual performance to reward about half the trials. The results show a typical pattern of adaptation in both the 'spatial only' and the 'spatial & reward' groups, whereas the 'reward only' group was unable to adapt. The rewards did not affect the overall pattern of adaptation in the 'spatial & reward' group. However, on a trial-by-trial basis, the rewards reduced adaptive changes to spatial errors.

Perceptual grouping determines haptic contextual modulation.

Since the early phenomenological demonstrations of Gestalt principles, one of the major challenges of Gestalt psychology has been to quantify these principles. Here, we show that contextual modulation, i.e. the influence of context on target perception, can be used as a tool to quantify perceptual grouping in the haptic domain, similar to the visual domain. We investigated the influence of target-flanker grouping on performance in haptic vernier offset discrimination. We hypothesized that when, despite the apparent differences between vision and haptics, similar grouping principles are operational, a similar pattern of flanker interference would be observed in the haptic as in the visual domain. Participants discriminated the offset of a haptic vernier. The vernier was flanked by different flanker configurations: no flankers, single flanking lines, 10 flanking lines, rectangles and single perpendicular lines, varying the degree to which the vernier grouped with the flankers. Additionally, we used two different flanker widths (same width as and narrower than the target), again to vary target-flanker grouping. Our results show a clear effect of flankers: performance was much better when the vernier was presented alone compared to when it was presented with flankers. In the majority of flanker configurations, grouping between the target and the flankers determined the strength of interference, similar to the visual domain. However, in the same width rectangular flanker condition we found aberrant results. We discuss the results of our study in light of similarities and differences between vision and haptics and the interaction between different grouping principles. We conclude that in haptics, similar organization principles apply as in visual perception and argue that grouping and Gestalt are key organization principles not only of vision, but of the perceptual system in general.

The effect of perceptual grouping on haptic numerosity perception.

We used a haptic enumeration task to investigate whether enumeration can be facilitated by perceptual grouping in the haptic modality. Eight participants were asked to count tangible dots as quickly and accurately as possible, while moving their finger pad over a tactile display. In Experiment 1, we manipulated the number and organization of the dots, while keeping the total exploration area constant. The dots were either evenly distributed on a horizontal line (baseline condition) or organized into groups based on either proximity (dots placed in closer proximity to each other) or configural cues (dots placed in a geometric configuration). In Experiment 2, we varied the distance between the subsets of dots. We hypothesized that when subsets of dots can be grouped together, the enumeration time will be shorter and accuracy will be higher than in the baseline condition. The results of both experiments showed faster enumeration for the configural condition than for the baseline condition, indicating that configural grouping also facilitates haptic enumeration. In Experiment 2, faster enumeration was also observed for the proximity condition than for the baseline condition. Thus, perceptual grouping speeds up haptic enumeration by both configural and proximity cues, suggesting that similar mechanisms underlie perceptual grouping in both visual and haptic enumeration.

Relative finger position influences whether you can localize tactile stimuli.

To investigate whether the relative positions of the fingers influence tactile localization, participants were asked to localize tactile stimuli applied to their fingertips. We measured the location and rate of errors for three finger configurations: fingers stretched out and together so that they are touching each other, fingers stretched out and spread apart maximally and fingers stretched out with the two hands on top of each other so that the fingers are interwoven. When the fingers contact each other, it is likely that the error rate to the adjacent fingers will be higher than when the fingers are spread apart. In particular, we reasoned that localization would probably improve when the fingers are spread. We aimed at assessing whether such adjacency was measured in external coordinates (taking proprioception into account) or on the body (in skin coordinates). The results confirmed that the error rate was lower when the fingers were spread. However, there was no decrease in error rate to neighbouring fingertips in the fingers spread condition in comparison with the fingers together condition. In an additional experiment, we showed that the lower error rate when the fingers were spread was not related to the continuous tactile input from the neighbouring fingers when the fingers were together. The current results suggest that information from proprioception is taken into account in perceiving the location of a stimulus on one of the fingertips.

The use of proprioception and tactile information in haptic search.

To investigate how tactile and proprioceptive information are used in haptic object discrimination we conducted a haptic search task in which participants had to search for either a cylinder, a bar or a rotated cube within a grid of aligned cubes. Tactile information from one finger is enough to detect a cylinder amongst the cubes. For detecting a bar or a rotated cube amongst cubes touch alone is not enough. For the rotated cube this is evident because its shape is identical to that of the non-targets, so proprioception must provide information about the orientation of the fingers and hand when touching it. For the bar one either needs proprioceptive information about the distance and direction of a single finger's movements along the surfaces, or proprioceptive information from several fingers when they touch it simultaneously. When using only one finger, search times for the bar were much longer than those for the other two targets. When the whole hand or both hands were used the search times were similar for all shapes. Most errors were made when searching for the rotated cube, probably due to systematic posture-related biases in judging orientation on the basis of proprioception. The results suggest that tactile and proprioceptive information are readily combined for shape discrimination.

Haptic search is more efficient when the stimulus can be interpreted as consisting of fewer items.

In a typical haptic search task, separate items are presented to individual fingertips. The time to find a specific item generally increases with the number of items, but is it the number of items or the number of fingers that determines search time? To find out, we conducted haptic search experiments in which horizontal lines made of swell paper were presented to either two, four or six of the participants' fingertips. The task for the participant was to lift the finger under which they did not feel (part of) a line. In one of the conditions separate non-aligned lines were presented to the fingertips so that the number of items increased with the number of fingers used. In two other conditions the participants had to find an interruption in a single straight line under one of the fingertips. These conditions differed in the size of the gap. If only the number of items in the tactile display were important, search times would increase with the number of fingers in the first condition, but not depend on the number of fingers used in the other two conditions. In all conditions we found that the search time increased with the number of fingers used. However, this increase was smaller in the single line condition in which the gap was large enough for one finger to not make any contact with the line. Thus, the number of fingers involved determines the haptic search time, but search is more efficient when the stimulus can be interpreted as consisting of fewer items.

Parallel and serial search in haptics.

We propose a model that distinguishes between parallel and serial search in haptics. To test this model, participants performed three haptic search experiments in which a target and distractors were presented to their fingertips. The participants indicated a target's presence by lifting the corresponding finger, or its absence by lifting all fingers. In one experiment, the target was a cross and the distractors were circles. In another, the target was a vertical line and the distractors were horizontal lines. In both cases, we found a serial search pattern. In a final experiment, the target was a horizontal line and the distractors were surfaces without any contours. In this case, we found a parallel search pattern. We conclude that the model can describe our data very well.

Haptic search with finger movements: using more fingers does not necessarily reduce search times.

Two haptic serial search tasks were used to investigate how the separations between items, and the number of fingers used to scan them, influence the search time and search strategy. In both tasks participants had to search for a target (cross) between a fixed number of non-targets (circles). The items were placed in a straight line. The target's position was varied within blocks, and inter-item separation was varied between blocks. In the first experiment participants used their index finger to scan the display. As expected, search time depended on target position as well as on item separation. For larger separations participants' movements were jerky, resembling 'saccades' and 'fixations', while for the shortest separation the movements were smooth. When only considering time in contact with an item, search times were the same for all separation conditions. Furthermore, participants never continued their movement after they encountered the target. These results suggest that participants did not use the time during which they were moving between the items to process information about the items. The search times were a little shorter than those in a static search experiment (Overvliet et al. in Percept Psychophys, 2007a), where multiple items were presented to the fingertips simultaneously. To investigate whether this is because the finger was moving or because only one finger was stimulated, we conducted a second experiment in which we asked participants to put three fingers in line and use them together to scan the items. Doing so increased the time in contact with the items for all separations, so search times were presumably longer in the static search experiment because multiple fingers were involved. This may be caused by the time that it takes to switch from one finger to the other.