Hypertext offers users a simple, flexible way to navigate through electronic information systems but at the potential risk of becoming lost in the network of interconnected pieces of information. A study was conducted on information retrieval using a commercial hypertext-based help system. It was found that the predominant search strategy was "browsing" (characterized by scanning tables of contents and paging through topics), rather than employing the indexes ("analytical search"). Although subjects did not get lost, individuals with better spatial visualization skills, as measured by a standardized test, were faster at retrieving information and returned to the top of the information hierarchy less often than those with poorer spatial visualization skills. These results support previous studies that have found a strong preference by users for browsing in hypertext systems and extend those findings to a new domain (help), a different type of user interface, and a different information architecture. In addition, the results demonstrate the importance of spatial visualization ability for efficient navigation and information retrieval in a hierarchical hypertext system.

Pigeons were exposed to fixed-time and fixed-interval schedules that ranged from 30 to 960 sec. The probability of a subject's location in the rear of the chamber (away from the reinforcer dispenser) peaked during the postreinforcer period, and was referenced to proportional time between reinforcers. Increasing the interreinforcer interval generally increased time in the rear. In some sessions (Experiment 1), location in the rear produced an explicit stimulus change (altered the color and intensity of lights, i.e., time-out); this change increased time spent in the rear without affecting its temporal locus or its relation to the interreinforcer interval. During Experiment 2, a keypeck (near the reinforcer site) produced the explicit stimulus change used in Experiment 1. The characteristics of keypeck time-out resembled those of movement to the rear of the chamber (with and without an explicit stimulus change), suggesting that movement away from the reinforcer site is functionally homologous to keypeck time-out.

The present experiments evaluated whether transitions in reinforcer probability are necessary to induce attack in pigeons. In Experiment I, three of six pigeons exposed to response-contingent constant-probability food schedules and a photograph of a conspecific as a target exhibited sustained postreinforcement attack on the target. The postreinforcement pattern of attack developed over the course of the experiment and was accompanied by a reduction in the rate of postreinforcement key pecking and an increase in the postreinforcement pause in key pecking. These effects on key pecking resulted in unprogrammed variations in the probability of reinforcement which may have been responsible for the induction of attack. In Experiment II, the attack-inducing properties of a constant-probability response-independent food schedule were compared to a periodic food schedule matched for overall rate of food delivery and to a no-food condition. In addition to attack, the spatial location of the subjects was monitored during each interfood interval. The periodic and aperiodic food schedules generated very different patterns of spatial location. Postfood attack was induced by both food schedules, although the constant-probability schedule induced attack in fewer birds. The no-food condition was not effective in inducing attack in any birds. These experiments indicate that intermittent food schedules without reductions in reinforcer probability are sufficient to induce attack in some pigeons, although not as effective as schedules with transitions in reinforcer probability.

Hypertext offers users a simple, flexible way to navigate through electronic information systems but at the potential risk of becoming lost in the network of interconnected pieces of information. A study was conducted on information retrieval using a commercial hypertextbased help system. It was found that the predominant search strategy was "browsing" (characterized by scanning tables of contents and paging through topics), rather than employing the indexes ("analytical search"). Although subjects did not become lost, individuals with better spatial visualization ability, as measured by a standardized test, were faster at retrieving information and returned to the top of the information hierarchy less often than those with poorer spatial visualization ability. These results support previous studies that have found a strong preference by users to browse in hypertext systems and extend those findings to a new domain (help), a different type of user interface, and a different information architecture. In addition, the results demonstrate the importance of spatial visualization ability for efficient navigation and information retrieval in a hierarchical hypertext system.