UC Berkeley

This research investigates the finger-movement strategies used by blindfolded sighted participants when completing a search task on a tactile map. These participants can be considered similar to individuals with recent vision loss, for example due to age-related causes that are responsible for the majority visual impairments in the developed world. The results could also be extended to children with acquired visual impairments, who are likely to encounter a large number of tactile maps and graphics in their continuing education. Tactile maps, and tactile graphics in general, provide opportunities to individuals with low or no vision to access educational, professional, and other graphical materials. However, performance with tactile graphics is highly dependent on the finger-movement strategies used. Results of this dissertation can provide insight into the type of movement strategies that are likely to be used by individuals with a recently acquired visual impairment, and can inform tactile training and the design of tactile displays for these individuals.

Finger-position data were collected from nine right-handed blindfolded sighted participants as they searched for a landmark on a tactile map using either one or five fingers. Three separate sets of analyses were conducted. In the first, the spatial and temporal distributions of finger locations during the haptic search task were analyzed to investigate why five-finger search is significantly faster than one-finger search. Results supported the possibility that five-finger search was benefited by greater throughput of information, with information flowing through multiple fingers, and the larger detection radius of the hand during five-finger search. The results did not support faster index-finger movement in five-finger search, nor the use of fingers as non-moving anchors to promote better allocentric spatial representations.

The second set of analyses investigated the use of systematic strategies: spirals, zigzags, and parallel sweeps, during haptic search. Monte Carlo simulations were used to demonstrate that a systematic strategy is more efficient than a random walk (Brownian, L'evy, or ballistic) When searching for a non-revisitable target in an enclosed space. However, this efficiency benefit decreases with increased detection radius. Systematic strategies were observed and coded in the finger-movement trajectories, and were found to be more prevalent in one-finger search than five-finger search, agreeing with predictions based on the Monte Carlo simulations. This effect was not explained by longer one-finger search times.

Finally, an initial model of finger-movement trajectories was constructed based on the Langevin equation, a stochastic differential equation describing correlated random motion. This model allowed for stimulus features to attract or repel finger movements through instantiation of a potential function. Ornstein-Uhlenbeck schemes of attraction and repulsion were activated through spatial thresholds, determined using Maximum Likelihood Estimation. These models revealed benefits of five-finger search over one-finger search that were not apparent in previous analyses. Namely, attraction to distractors was reduced in five-finger compared to one-finger search.