Peder J. Johnson

Assessing implicit learning with indirect tests: Determining what is learned about sequence structure.

These studies investigated the unconfounding of learning of simple frequency information from complex structure during sequenced serial reaction time (RT) trials. Experiment 1 demonstrated that disruptions in RTs that occur when Ss are transferred to random trials (e.g., A. Cohen, R. Ivry, & S. W. Keele, 1990) may result from a change in simple frequency information. The implication is that negative transfer effects cannot be attributed to the learning of complex information if simpler information is also allowed to vary between training and transfer trials. Experiment 2 used a procedure in which simple frequency information remained constant between training and transfer trials, which provided evidence of complex sequence structure learning while subjects were engaged in a secondary tone-counting task

Locus of the Predictive Advantage in Pathfinder-Based Representations of Classroom Knowledge.

Earlier T. E. Goldsmith, P. J. Johnson, and W. H. Acton (1991) reported that a Pathfinder-based representation of domain knowledge was more predictive of classroom exam performance than the underlying data from which the Pathfinder representation was derived (relatedness ratings of pairs of concepts from the domain). The question is, what affords Pathfinder this predictive advantage? Analyses reveal that whereas the relatedness ratings are based on a 7-point scale, all of the predictiveness is contained in the most highly related pairs. Moreover, Pathfinders advantage is related to the manner in which it dichotomizes the relatedness data into directly linked and indirectly linked concept pairs

The effects of perceptual pretraining on concept identification and preference.

Abstract The effects of perceptual pretraining on preferences and concept identification performance in five- and eight-year-old S s were investigated. The results indicated that perceptual pretraining facilitated performance on a concept identification task with a non-preferred relevant cue for the five-year-olds, but not for the eight-year-olds. However, perceptual pretraining did not affect preference in either age group and S s were as consistent in their preference responding before and after perceptual pretraining as was a control group that had not received perceptual pretraining. Preferences were highly consistent in both age groups over a one week period. Perceptual pretraining was interpreted to increase the usability of the less preferred dimensions for the younger S s, and to leave the affective preference response unaltered in both groups.

Resource allocation and the attentional demands of letter encoding.

The idea that familiar events can be encoded automatically has gained general acceptance in cognitive psychology since Posner and Boies (1971) first reported that reaction times to a secondary probe were not interfered with by letter encoding. More recently, Ogden, Martin, and Paap (1981) used a more valid control for estimating baseline probe performance and found secondary task interference, suggesting that letter encoding does require attentional resources. The present series of experiments began with the aim of evaluating Ogden et als evidence against automaticity when the first letter was not terminated after a brief exposure, as was done in their study. In the first set of experiments we found evidence of encoding interference when the interval between the two letters was varied (50 to 1,000 msec), but this interference disappeared when there was a constant 1,000-msec interval between the letters. On the basis of these findings, we hypothesized that changes in the primary task (e.g., the exposure duration of the first letter or the interval between the two letters) may influence the momentary allocation of resources between the primary and secondary tasks. More specifically, we hypothesized that any momentary reduction in the resources demanded by the primary tasks results in a reallocation of resources to the secondary task, which in turn reduces the sensitivity of the secondary task to the demands of the primary task, that is, probe performance is moved into the data-limited region of processing (Norman & Bobrow, 1975). This idea was tested by reducing resource allocation to the probe task at the time of encoding by reducing the expectancy (i.e., the probability) of probes in the temporal proximity of the first letter. The results showed that this manipulation produced a large and significant increase in encoding interference. Moreover, when the intensity of the tone (probe) was decreased from 70 to 60 dB, the magnitude of encoding interference was further increased. In regard to the specific issue of automaticity, the findings suggest that encoding familiar events does require resources, which will result in secondary task interference given that the secondary task is in the resource-limited region of processing. More important, the findings suggest that the magnitude of secondary task interference is dependent on within-trial changes in resource allocation between the primary and secondary tasks. This possibility has general implications for dual-task methodology and the measurement of attentional demands.

Application of Psychometrics to the Calibration of Air Carrier Evaluators

The FAAs Advanced Qualification Program (AQP) encourages airlines to implement proficiency-based training programs and requires collection of reliable and valid performance assessment data. We present applications of traditional and innovative psychometric methods to this domain.

Factors Influencing Children's Concept Identification Performance with Nonpreferred Relevant Attributes.

Abstract It has been demonstrated that young children have difficulty solving simple concept problems when the relevant dimension is the less preferred dimension. Three experiments were conducted to determine the influence of the following factors upon task difficulty: (a) age (4 yr 6 months-5 yr 11 months and 7 yr 0 months-8 yr 9 months); (b) the percentage of redundancy between relevant and irrelevant cues (0, 25, and 50%); (c) saliency of reinforcement; (d) discriminability of relevant nonpreferred dimension; and (e) learning-set pretraining to reject preferred irrelevant dimensions. The results indicated that when compared with the younger children, the older children are far less affected by having the preferred dimension irrelevant. The learning-set pretraining condition resulted in substantial facilitation effects indicating that the inability to inhibit responding to dominant irrelevant cues is an important factor and that an effective rejection strategy can be acquired by young children with brief training. Partial reinforcement of irrelevant dimensions, saliency of reinforcement, and discriminability of the nonpreferred relevant dimension were shown to not contribute significantly to the difficulty experienced by young children on these problems.

Effects of Enforced Attention and Stimulus Phasing Upon Rule Learning in Children.

Abstract Two experiments investigated the effects of enforced attention (EA) and stimulus phasing upon rule learning (RL) in 6-year-old children. In Exp. I. Ss learned the disjunctive or biconditional rule either under standard RL instructions (S told to watch for the presence or absence of relevant values circled on attribute cards which were left in Ss view throughout the experiment), or EA (S required to point to values on attribute card contained in each successive stimulus). Enforced attention resulted in a reliable improvement in performance with virtually all of the improvement occurring on the biconditional rule. In Exp. II Ss learned the biconditional rule with the stimulus population divided into four general classes (based on the presence or absence of two relevant values) and introduced suceessively in one of two orders. Half the Ss received EA in addition to the stimulus phasing. While stimulus phasing alone improved performance independent of order, optimal performance was achieved when EA was included. Performance on a second biconditional problem indicated that EA, in addition to reducing attention to irrelevant attributes, enhanced the acquisition of an efficient coding system.

Concept of dimensionality and optional shift performance in nursery school children

Abstract The ability to order correctly a set of stimuli varying in brightness has been shown to be a good predictor of performance in a mandatory reversal shift problem (Johnson and White, 1967). In the present experiment, nursery school children were administered concept of dimensionality tests (DT) during which they were required to order sets of stimuli varying in size or brightness. DT-training effects were assessed by presenting the DTs immediately before, immediately after, a week before, or a week after an optional shift problem. Results showed that Ss who made few errors on a DT were more likely to make optional reversal shifts. However, performance on one DT dimension was not a reliable predictor of performance for another relevant dimension during initial discrimination learning. DT performance was shown to be a better predictor of performance on the optional shift problem than was age. No significant DT-training effects were found. The results were taken to indicate that the concept of a dimension underlies the ability to make a reversal shift response.

Attribute identification in children as a function of stimulus dimensionality

Abstract Thirty-two four-year-old and 32 six-year-old children were presented a conjunctive concept that involved either two values from within a single dimension (unidimensional condition) or two values from different dimensions (bidimensional condition). As predicted, the results showed that the uni- vs bidimensional factor had no effect upon the performance of the younger children, while the older children solved more rapidly in the unidimensional than in the bidimensional condition. The results were interpreted to suggest that the younger childrens responding is not under dimensional control and that a major source of difficulty for older children solving the conjunctive concept is the necessity of responding to two dimensions, not to two values.

The stability and transferability of conceptual coding in children

Abstract The stability and transferability of conceptual coding systems learned under special rule-learning (RL) procedures was investigated in 6- and 9- year old children. All children were presented three successive biconditional RL problems and were then transferred to a conditional RL problem. The experimental S s received special enforced-attention (EA) instructions and procedures on the first two biconditional problems which were designed to direct their attention to the relevant attributes and encourage a truth-table coding of the stimulus population. On the third biconditional problem and the conditional rule problem they were given standard RL instructions, identical to what one of the control conditions received on all four problems. A second control condition received the EA instructions on all four problems. The results from problems 1 and 2 showed that the experimental procedures facilitated biconditional rule-learning performance in both age groups and tended to reduce the absolute superiority of the older children. Omission of EA on problem 3 had a significant disruptive effect on biconditional RL performance, but on the conditional transfer problem the experimental S s in both age groups solved as rapidly as the control receiving EA and with significantly fewer errors than the control conditions receiving standard RL instructions on all four problems.