Visual pattern recognition: Size preprocessing re-examined

Template theories of visual pattern recognition assume the operation of preprocessing routines to deal with irrelevancies such as discrepancies in stimulus size. In three experiments where size was an irrelevant dimension, observers classified pairs of forms as either “same” or “different”. In Experiment I, the classification “different” was required when the stimuli shared the same form but a different orientation, and “same” when the stimuli shared the same form and orientation. Under these conditions RT was an increasing function of the magnitude of the size disparity between stimuli with equal slopes for “same” and “different” judgements. In Experiment II, “different” classifications were made to stimuli that had different forms, and “same” to figures with the same form. This stimulus set produced a size disparity function that interacted with response type; “different” responses had a shallower slope. Experiment III consisted of a mixed stimulus set drawn from both Experiment I and II. Stimuli that produced additive effects of size disparity and response type in Experiment I now produced an interaction between these two factors similar to the one observed in Experiment II. The results of these experiments are interpreted as evidence that previous contradictory results reported in the literature stem from differences in the way the stimulus set is constructed, and that size transformations can not be a necessary operation, at least when “different” judgements are made. The results are problematic for the view that size disparity effects in matching tasks are easily interpretable in terms of a primitive size normalization stage that precedes any comparison operations.     

Discriminability and central intermittency in same-different judgements

Four sets of paired visual stimuli (OO, XX, XO, or OX) were judged by 48 subjects to be either “same” or “different.” Decision latencies of the same and different judgement were studied as a function of the inter-stimulus interval (ISI). In Experiments I and II, in which stimulus durations were 70 millisec., decision latencies showed marked increases when the ISI was reduced to 100 millisec., but in Experiments III and IV, in which the stimulus durations were only 40 millisec., comparable increases did not occur until the ISI was reduced to 50 millisec. These increases were more marked for “same” than for “different” judgements, although overall decision latencies were generally shorter for “same” judgements. The effects of varying ISIs and stimulus durations are interpreted in terms of masking; they fail to support an hypothesis of central intermittency.     

Asymmetry of generalization decrement in causal learning

Two experiments required volunteers to learn which of various “planes” caused high levels of pollution. Novel test items were then rated as causes of pollution. Items created by adding novel features were rated at the same level as that of the original training items but items created by removing features received reduced ratings. This asymmetry of generalization decrement was not predicted by a well-known configural model of stimulus representation (Pearce, 1987, 1994) but was predicted by a recently proposed model of stimulus representation, the replaced-elements model (Brandon, Vogel, & Wagner, 2000).     

The Perceptual Centre of a Stimulus as the Cue for Synchronization to a Metronome: Evidence from Asynchronies

In tasks where subjects are required to tap in synchrony to a sequence of evenly spaced uniform auditory stimuli (a metronome), tap onsets typically tend to anticipate the metronome's stimulus onsets. We investigated this phenomenon, called “negative asynchrony”, as a function of (1) the duration of the stimuli (1 or 2, 50, 100, and 300 msec), (2) the rise time of the stimuli (0%, 40%, and 80% of stimulus duration), and (3) the interstimulus onset interval duration (500, 700, and 900 msec). The results from three experiments with 28 different subjects showed a significant reduction of the negative asynchrony with longer stimulus durations, and the reduction was not significantly affected by the tempo of the stimulus sequence. Also, a prolongation of the rise time of the stimuli caused an analogous reduction of the negative asynchrony. Findings were taken to suggest that subjects use the perceptual centre rather than physical onset of stimulus as the cue with which to synchronize     

An observation on set and visual comparison processes

Tasks involving the comparison of sequentially presented stimulus items to determine identity have generally only required interpretation in terms of simple matching. Some studies have attempted to show effects of selective set in these tasks but with little success except in tasks involving comparison of sentences with pictures. By simplifying the task a reliable set effect is demonstrated, showing that a mismatch of certain stimulus attributes only has an identifiable effect on a subject's reaction-time if he may reasonably be supposed to know that the attribute is critical and is set to examine stimulus items for the attribute. Such conditions result in a relatively fast judgement. The finding suggests that the sequential “same“/“different” judgement is a richer source of data concerning recognition and comparison than has been supposed.     

Payoff: A neglected factor in reaction time measurement

The aim of this paper is to point out some implications of implicit or explicit payoffs in reaction time experiments. In any psychophysical task the subject is given a set of instructions which define the desired performance, as it were the rules of the game. Instructions may be supplemented by knowledge of results which further define the required performance. When performance admits of being “good” or “bad,” however wide these limits, the instructions may be said to define, explicitly or implicitly, a system of payoffs. In reaction time experiments the subject is instructed, amongst other things, to respond as soon as possible after the stimulus but never before it. These instructions could equally well be represented by a system of payoffs and Figure 1 illustrates the kind of payoff system normally implied in these instructions. If S represents the time at which the physical stimulus occurs the graph shows a fixed penalty for anticipations (response before S), maximum positive score for responses immediately following S declining at an arbitrary rate as a function of elapsed time between S and the response until an unacceptably long time involves a constant maximum penalty. Some experimenters choose to treat very short RT's as anticipations. This is represented by the dashed line in Figure 1, but it should be clear that payoff values are, to a certain extent, arbitrary.     

Method of serial extrapolation

A psychophysical procedure is proposed for measuring a subject's power of discerning “lawfulness,” “regularity,” “invariance” in a stimulus sequence. The task consists of supplying the next term in a given series. The method is tested by applying it to a visual series consisting of lines subdivided in geometric progression. The effect upon accuracy of judgement is investigated by varying the settings of the stimulus, in particular the rate of increase or gradient of the series and the number of segments given prior to extrapolation, and of presenting the next-but-one position in the series, thereby converting the task into one of interpolation. Where possible, quantitative predictions are made concerning these variables on the basis of certain explicit assumptions.

Results show that accuracy decreases, as predicted, as the gradient of the series increases, but does not seem to be affected by adding to the number of terms presented. Interpolation shows a reduction in variable error compared with extrapolation more or less as required by theory. Finally, discrepancies in the findings are discussed with special reference to their bearing on the validity of the method.     

Refractoriness in same-different judgments

Subjects judged as “same” or “different” pairs of visual stimuli, each either an X or an O, whose onsets were separated by intervals of 50, 100, 150 or 200 ms. When the interstimulus interval (ISI) was held constant within a block of 40 trials (Experiment I), decision time was independent of ISI. But when ISI was varied unpredictably from trial to trial (Experiments II and III), decision times increased the shorter the ISI. This effect was more marked for “same” than for “different” decisions. These results fail to support a single-channel interpretation, but suggest that variations in decision time depend on the subject's expectancy as to when the second stimulus will follow the first.     

An improved technique for monitoring fixation in tachistoscopic tasks

A monocular lens mounted behind a stimulus screen was used to monitor fixation in a tachistoscopic recognition task. This technique was found superior to the often used “sideview technique” and as accurate as the method involving identification of a fixation stimulus. It was found that the “telescope technique” is preferable in those studies where presentation of a fixation stimulus might interfere with recognition or report of the experimental stimulus.     

The effect of an irrelevant dimension on “same-different” judgements of multi-dimensional stimuli

Subjects made “same”-“different” judgements of simultaneously presented pairs of visual stimuli which could vary either in shape and colour independently, or in shape alone. In both conditions only shape was relevant to the “same”-“different” judgement. In the former condition “same” and “different” reaction times (RTs) were shorter when the states (“same” or “different”) of the relevant and the irrelevant dimension, colour, were the same. This result is interpreted as support for either a perceptual or a response interference hypothesis. The presence of an irrelevant dimension did not appear to affect differentially “same” and “different” judgements. The need for a re-evaluation of the results from other studies of multi-dimensional stimulus discrimination is discussed.     

Specifying levels of processing

This study examines the nature of stimulus processing under semantic and nonsemantic orienting instructions. Two experiments are reported in which subjects were presented with a series of trials each beginning with the presentation of a “decision word” about which they made either a semantic or non-semantic orienting decision. This decision was followed by a word in coloured ink whose colour subjects were required to name as quickly as possible. On half the trails the coloured word was the primary associate of the decision word whilst on the other half the two words were normatively unrelated. On completion of the experiments the subjects were given an unexpected free recall test. The semantic orienting condition led to longer colour naming latencies on associate trials whilst no such difference was found in the non-semantic condition. The semantic condition also produced higher levels of incidental recall although paradoxically an analysis of associative clustering in recall failed to show any difference between the two orienting conditions. The results are interpreted as support for the “Levels of Processing” approach to memory since they provide an index of processing depth which is independent of retention performance.     

Tactile short-term memory

Subjects tried to recall the location of a tactile stimulus on the underside of the forearm after delays of 0, 3, 5, 10, 15, 30, 45 and 60 sec. When “rehearsal” was prevented by requiring subjects to count backwards during the delay, accuracy of recall decreased systematically reaching an asymptote after 45 sec. When subjects were left free to “rehearse,” this did not affect the decline in accuracy over the first 10 sec. Between 10 and 15 sec. there was a significant increase in accuracy followed by a slow decline which had not reached asymptote by 60 sec. It is suggested that tactile STM (short-term memory) depends on two processes, a fading sensory trace which is unaffected by distraction and a less labile system which does not appear to be verbal but which depends on “rehearsal.”     

Selective interference during the retrieval of visual images

Twenty-six subjects memorized lists of (low I and high I) noun pairs under imagery or verbal mediation instructions. At recall the subjects were presented a digit (“1” or “2”) either auditorily or visually as an interfering stimulus.

Visual interference was found to selectively affect the retrieval of high I response terms. Also, the retrieval of nouns studied by imagery mediation was found to be selectively disrupted by visual interference.

These results suggest that the qualities of a visual image are retained all the way from image acquisition to retrieval, and that the visual components of images generated at the acquisition stage are probably not lost by subsequent coding processes.     

Information processing by performers and observers in a paired associate learning task

The effects of two training procedures on learning and performance are compared. Performers select a response alternative for each stimulus on Trial 1 and receive feedback in terms of “Right” or “Wrong”. Observers receive the same information by listening to the experimenter. Experiment I tests the hypothesis that performers and observers are using a different learning strategy when there are only two response alternatives available for each stimulus on Trial 1. A recognition procedure was used on Trial 2; each stimulus was followed by four alternatives, two of them being the same as presented on Trial 1. Subjects have to recognize the two “old” alternatives. Performers are always better at recognizing the chosen alternative, whereas observers are better at recognizing the correct alternative. Experiment II extends the comparison between performers and observers to a task with four response alternatives on Trial 1. There are no longer differences in performance between the two training procedures.     

The psychological refractory period as a function of performance of a first response

Proponents of a “single channel” theory of the psychological refractory period have not specified whether the single channel occupies only the decision component of the response selection, only the motor or response component, or both. In this experiment, the delay in the RT to the second of two successively presented stimuli was examined as a function of whether or not an overt motor response was made to the first stimulus, keeping the decision component constant. It was found that in both conditions RT₂ was delayed, suggesting that the decision component was a part of the single channel. However, RT₂ was delayed by a significantly greater amount if a motor response was required, indicating that the motor component is part of the single channel as well. Implications of the results for an expectancy theory of the psychological refractory period are discussed.     

A search for perceptual differentiation produced by nondifferential reinforcement

In each of four experiments (three using pigeons as the subjects and one using rats) animals in an experimental condition were given exposure to a pair of stimuli that subsequently they were required to discriminate between. During the exposure phase response to either stimulus was consistently followed by reward. These subjects were uniformly found to be retarded in learning the discrimination when compared with control subjects that received equivalent pre-training in the absence of the critical stimuli. These results provide no support for the suggestion that stimuli may become “perceptually differentiated” during the pre-training phase but are consistent with the proposal that novel stimuli possess a high level of associability that declines as a result of experience.     

Redundancy effects in short-term memory

Five experiments are reported whose purpose was to demonstrate that short-term memory is improved by redundancy within the material. In Experiment I “tune” containing two, three, four and five tones of differing frequencies had to be coded into digits 1-5, to indicate the order of the pitches in a tune. Performance on stimuli containing correlated amplitude and duration were compared with the uni-dimensional condition. Experiment II repeated I, but required intensity to be coded. Experiment III required pitch coding under three conditions including that when amplitude and frequency were uncorrelated, and compared the performance of musically trained sunjects with nonmusicians. Experiment IV repeated III, but subjects were informed of the relation between dimensions. Experiment V involved “shadowing” the tunes by whistling simultaneously with the stimulus.

It was concluded (a) that intercorrelation improves, but zero correlation impairs short-term memory; (b) that knowledge of the relation between dimensions improves performance in the correlated condition, but does not prevent impairment under zero correlation; and (c) the performance of musically trained subjects exceeds that of controls and is unaffected by the presence of a correlated or uncorrelated dimension.     

Win-stay behaviour in the rat

The Law of Effect dictates that animals will repeat the just-reinforced response “win-stay”, and yet there have been apparent violations of this by rats running for food in mazes, in the form of “win-shift” behaviour. Four experiments analysed the conditions determining win-stay and win-shift behaviour. All the experiments employed a schedule in which reinforcement was distributed across two choices, and in which the probability of reinforcement for the first response after previous reinforcement was equivalent for the two options. Despite this lack of programmed differential reinforcement, rats showed a significant win-stay tendency in Experiment 1, with no spatial bias. Experiments 2 and 3 used respectively a Y-maze and an operant chamber in which responding required a return to a central choice point. In both situations, significant win-shift behaviour resulted initially, at relatively high frequencies of reinforcement, but this win-shift changed to significant win-stay over many sessions, and with intermittent reinforcement. Experiment 4 introduced explicitly non-reinforced trials and demonstrated a “lose-shift” tendency, parallel to win-stay. Possible artefacts underlying win-stay and win-shift behaviour are rejected, and various mechanisms for these effects, including theories of optimal choice and associative learning, are discussed.     

Acoustic factors in visual search

Previous work has shown that in searching for existing or absent “e.s” in printed prose, the presence or absence of silent “e.s” was less likely to be detected than that of pronounced “e.s.” It was suggested that the acoustic or kinaesthetic “image” was searched for evidence of an “e” in addition to the visual stimulus and that evidence from both sources was considered in making the appropriate response.

The present experiment employs mainly substitutive errors within words, which may or may not change their pronunciation. The results suggest that the form of the acoustic correlates has no bearing upon whether the words are detected as wrongly spelt, but that the presence or absence of an acoustic event corresponding in time to the spatial location of the error is important.     

Conditional probability and components of rt in the variable foreperiod experiment

If the passage of time during the foreperiod in a variable foreperiod experiment is marked by a series of tones, RT decreases with the conditional probability of stimulus occurrence. RTs at short foreperiods, however, are rather slower than would be expected on the basis of a simple conditional probability effect. It is suggested that this is attributable to an independent “initial slow reaction” effect, and it is shown that the degree of this effect is influenced by the duration of the prior foreperiod. The results are related to those of variable foreperiod a-reactions in which no conditional probability effect has been found, and it is argued that in a marked reaction of the kind described above, the initial slow reaction effect behaves like an a-reaction component of RT. It is suggested that the absence of a conditional probability effect in the a-reaction and its presence in the marked reaction are related to the fact that a different type of sensory process is used to identify the signal in each case.