John B. Shea

The Influence of Reminder Trials on Contextual Interference Effects

Two experiments investigated the proposition of the elaboration explanation for contextual interference that more than one task is present in working memory when multiple tasks are practiced in a random schedule but that only one task is present in working memory when multiple tasks are practiced in a blocked schedule. Three motor tasks were performed as fast as possible in either a random or blocked practice schedule. At the end of practice, a reminder trial for each task was either given or not given. Acquisition performance was slower for the random practice conditions than for the blocked practice conditions. Retention performance was faster for the random practice conditions than for the blocked practice condition that did not receive a reminder trial for each task. Importantly, performance differences were not found between the random practice conditions and the blocked practice condition that did receive a reminder trial for each task. A blocked practice condition with a beneficial acquisition and reminder task order pairing performed faster during both acquisition and retention than a comparable random practice condition. Reminder trials can facilitate detailing of task characteristics, and their effectiveness is determined by the elapsed time and number of intervening tasks during acquisition and retention.

When Forgetting Benefits Motor Retention

Recent research (Lee & Weeks, 1987; Weeks, Lee, & Elliott, 1987) investigating the processes responsible for the contextual interference phenomenon has used a modified short-term motor retention paradigm to support the reconstruction explanation (Lee & Magill, 1985; Magill, 1989; Magill & Hall, 1990). The present experiment was an extension of these experiments in which forgetting of an acquisition task was induced through performance of either a similar or dissimilar distractor task during the intertrial interval. The effects of an extra practice trial with the acquisition task as well as no activity during the intertrial interval were also investigated. In addition, forgetting of the acquisition task was assessed prior to a reconstruction trial, which immediately preceded a 2-min filled retention interval. Both similar and dissimilar distractor tasks caused equivalent amounts of forgetting of the acquisition task prior to the reconstruction trial. However, retention of the acquisition task was significantly improved if its reconstruction occurred following forgetting due to interference from performance of a similar distractor task. These findings suggest forgetting and subsequent reconstruction alone are not sufficient for improved retention. These processes must occur in the context of a similar task for improved retention.

A Movement Constraint Interpretation of the Response Complexity Effect on Programming Time

This paper reinterprets data from five recently published experiments on response complexity and programming time in which number of movement parts of rapidly executed limb responses was advanced as the principal element influencing length of response programming time. Alternatively it is argued here that programming time, in these experiments, was predominantly a function of the constraints placed upon the output of the motor system by the demand for movement accuracy. The quantification of this accuracy demand is achieved by using the metric of Index of Difficulty (Fitts, 1954). In discrete and straight-line tapping responses to circular targets, response complexity may be conceptualized in terms of the angular constraint imposed on movement initiation at the start key. When responses require changes of direction between movement parts it is proposed that programming time may be a function of the cumulative movement constraints imposed by the task. The discussion focuses on the process by which an increased accuracy demand requires a more constrained motor system output which is brought about by a larger and/or more precise muscle synergy recruitment pattern, resulting in an increase in programming time.

Levels of Processing and the Coding of Position Cues in Motor Short-Term Memory

The present study investigated the appropriateness of the levels-of-processing framework of memory for explaining retention of information in motor short-term memory. Subjects were given labels descriptive of the positions to be remembered by the experimenter (EL), were given no labels (NL), or provided their own labels (SL). A control group (CONT) was required to count backwards during the presentation of the criterion positions. The inclusion of a 30-sec filled retention interval as well as 0-sec and 30-sec unfilled retention intervals tested a prediction by Craik and Lockhart (1972), when attention is diverted from an item, information will be lost at a rate appropriate to its level of processing - that is, slower rates for deeper levels. Groups EL and SL had greater accuracy at recall for all three retention intervals than groups CONT and NL. In addition, there was no significant increase in error between 30-sec unfilled and 30-sec filled intervals for groups EL and SL, while there was a significant increase in error for groups CONT and NL. The data were interpreted in terms of Craik and Lockharts (1972) levels-of-processing approach to memory.

The effects on skill acquisition of an interpolated motor short-term memory task during the kr-delay interval.

The effects of an interpolated motor short-term memory task during the KR-delay interval on skill acquisition was investigated. On each trial subjects estimated two positions on a linear-positioning apparatus. One group of subjects was presented two biasing movements, which they immediately recalled, during the KR-delay interval. The interpolated motor short-term memory task interfered with skill acquisition and retention.

Deficient Processing in Learning and Performance

Two experiments investigated the assumption made by Shea and Zimny (1983) in their recently proposed theoretical framework for contextual interference that recognition and recall performance are at least partially separable and can be independently manipulated. In both experiments, recognition training prior to acquisition resulted in fewer tasks being recalled for the recognition training condition than for the non-task recognition and no-recognition training conditions. We offer an explanation in which recognition training enabled subjects to solely use identification processes to keep the tasks in working memory and allow subsequent motor performance. These identification processes, however, did not provide detailed enough knowledge of the tasks to support recall performance (i.e. deficient processing). These findings are interpreted within a parallel-distributed processing view of learning and performance.

A Model for Contextual Interference Effects in Motor Learning

There exists substantial evidence that practice under conditions of high contextual interference can facilitate retention and transfer performance (Magill & Hall, 1990). Contextual interference refers to the situation in which there is interference among different tasks being learned across practice trials. Practice under a condition of high contextual interference (e.g., when multiple tasks are practiced in a random order) typically results in less proficient performance than practice under a condition of low contextual interference (e.g., when multiple tasks are practiced in a blocked order). These findings are reversed for retention and transfer tests, however, with performance being more proficient for the high contextual interference practice condition than for the low contextual interference practice condition. This phenomenon has attracted wide interest among motor skill researchers because it is counter to the common assumption that practice in situations with little or no interference is most advantageous for learning. We describe a contextual interference experiment and prevailing explanations for its findings. We then describe a hybrid connectionist model for contextual interference that has been successful in predicting empirical findings.

Orienting Task Specificity in Incidental Motor Learning

The manipulation of the retention of a linear movement by means of different orienting tasks in an incidental learning paradigm was investigated. Subjects were presented with a target position followed by a different nontarget position on each of four presentation trials. After the presentation trials subjects were required to recall the target and nontarget positions. One group made verbal estimates of the distance between the target and nontarget positions, another group discriminated between the target and nontarget positions, while no verbal responses were required for a third group. The verbal-response groups, while not differing from each other, had less error at recall than the nonverbalizing group. These findings were discussed in terms of the levels of processing framework for memory research (Craik & Lockhart, 1972).

Representational Structure and Strategic Processes for Movement Production

Three experiments were conducted to investigate the strategic processing responsible for the differential retention interval effect on response accuracy and latency measures reported by Shea (1977). It was found that subjects use previously presented movements in addition to well learned memory structures for movement production. These findings emphasize the constructive nature of movement production. A description of motor skill acquisition incorporating both representational structure and its closely associated strategic processing is offered.

The effects of high and low BACs on the Hoffmann reflex

The purpose of this study was to measure and assess the magnitude and latency of the H-reflex and M-response between pre- and post-alcohol consumption. Also of interest was the comparison of the H-reflex and M-response between the rising and falling curves of BAC. Seven male volunteer subjects participated in this study. Testing started with pre-alcohol BAC and EMG recordings from the tibial nerve following an electrical stimulation. After alcohol consumption, BAC readings were obtained every five minutes, and the EMGs were recorded at the following BAC levels (in mg/dl): 75, 100, and falling 75. H-latency, H-amplitude, and M-response were identified within the EMG signals. The analysis focused on the comparisons of these measures between each of the BAC levels and the baseline, as well as between the rising and falling BACs. Results showed that, overall, the H-reflex and M-response were depressed following alcohol consumption. In addition, the depression of H-reflex appeared to be greater during the rising curve than the falling curve.