Stimulus control and generalization of point-loss punishment with humans

Two experiments demonstrated stimulus control and generalization of conditioned punishment with humans. In both studies, responses first were reinforced with points exchangeable for money on a variable-interval schedule in the presence of one line length (S(D)). Next, a second line length was introduced, and point loss followed every response in the presence of that line (S(D)p). In the final training condition, points were deducted at session end. Response rate was lower in the presence of the S(D)p despite equal rates of points for money in the presence of both stimuli. In generalization testing for Experiment 1, the two lines were included in a 10-line continuum; S(D)p fell in the middle and the trained SD was at one end. Lines were presented randomly, and point delivery and loss contingencies were as in training but with points available in the presence of all lines. For all subjects, response rates were lowest around S(D)p and increased towards the SD end of the continuum. Because testing included only one or two lines beyond S(D), this pattern did not rule out S(D) generalization. Thus, in Experiment 2, stimuli beyond S(D) were added to generalization tests. Response rates did not decrease as a function of distance from S(D), clarifying the demonstration of punishment generalization.     

Peak shift revisited: a test of alternative interpretations

In Experiment 1, 2 groups of human subjects were trained to respond to 1 of 2 light intensity stimuli, S2 or S4, and then were tested for generalization with a randomized series of increasing values from S1 to S11. Both groups, including the group trained to respond to dimmer value, showed peak shifts to a brighter more centrally located test stimulus. In Experiment 2, which used line angle stimuli, both the size of the difference between S+ and S- and the range of test stimuli that extended beyond S+ were varied. The larger the S(+)-S- separation and the larger the range, the greater was the peak shift obtained. In Experiment 3, training involved an S- (line angle) surrounded by 2 S+ values with testing symmetrical about the training values and covering either a narrow or a wide range. The wide range produced greater peak shifts in both directions from S-. All 3 experiments support an adaptation-level interpretation of intradimensional discrimination learning and generalization test performance in human subjects. Related work with animals suggests the presence of similar processes.     

Generalization of a tactile stimulus in horses.

Using horses, we investigated the control of operant behavior by a tactile stimulus (the training stimulus) and the generalization of behavior to six other similar test stimuli. In a stall, the experimenters mounted a response panel in the doorway. Located on this panel were a response lever and a grain dispenser. The experimenters secured a tactile-stimulus belt to the horse's back. The stimulus belt was constructed by mounting seven solenoids along a piece of burlap in a manner that allowed each to provide the delivery of a tactile stimulus, a repetitive light tapping, at different locations (spaced 10.0 cm apart) along the horse's back. Two preliminary steps were necessary before generalization testing: training a measurable response (lip pressing) and training on several reinforcement schedules in the presence of a training stimulus (tapping by one of the solenoids). We then gave each horse two generalization test sessions. Results indicated that the horses' behavior was effectively controlled by the training stimulus. Horses made the greatest number of responses to the training stimulus, and the tendency to respond to the other test stimuli diminished as the stimuli became farther away from the training stimulus. These findings are discussed in the context of behavioral principles and their relevance to the training of horses.     

Peak shift as a function of multiple schedules of reinforcement

Pigeons were trained to respond to two stimuli on the wavelength continuum, 550 nm and 570 nm, each correlated with an independent schedule of reinforcement. The multiple schedule component in effect during 550 nm (S1) was always a variable-interval 1-min. During the 570-nm stimulus (S2) the second component of the schedule was either variable-interval 30-sec, 1-min, 2-min, 5-min, or extinction for different groups of birds. Generalization gradients were obtained after this training, with the following results: (1) response rate to S1 during training was related to the reinforcement frequency associated with S2; the distribution of responding during generalization testing was a function of the schedules of reinforcement used during training and the response rates they produced. Decreases in the relative frequency of reinforcement correlated with S2 resulted in increases in the distribution shift of responses away from S2 during generalization testing.     

Maintained nodal-distance effects in equivalence classes

Twelve subjects were trained to select one of two stimuli from a pair (the B pair) when presented with one of two stimuli from another pair (the A pair), thus establishing two AB relations, A1-B1 and A2-B2. In a similar fashion, additional stimuli were used to establish BC, CD, and DE relations. Trials used to train all relations occurred in each session. Once performances were established, probe trials were introduced that tested for the emergence of untrained relations (e.g., B1-D1 or A1-E1). These emergent relations were categorized according to nodal distance (i.e., the number of stimuli across which transitivity would have to hold in order for the relation to emerge). For example, a test for A2-C2 crosses one node (B2), whereas a test for A1-E1 crosses three nodes (B1, C1, and D1). Only 2 of the subjects formed equivalence classes. The evocation of class-appropriate responding by each emergent-relation probe was an inverse function of nodal distance for all 12 subjects. In addition, performance on the originally trained relations was disrupted by the introduction of probes. The 2 subjects who exhibited equivalence classes were then trained to make different numbers of key presses in the presence of each of the four A and E stimuli. In a response-transfer test, the B, C, and D stimuli evoked the responses trained to the A and E stimuli in the same equivalence class. Likelihood of class-appropriate responses was an inverse function of nodal distance, and this pattern persisted across testing. Reaction times in the transfer test were an inverted U-shaped function of nodal distance. Because training of the baseline relations occurred concurrently and the B, C, and D stimuli were presented an equal number of times before the transfer test, the test performances illustrate effects of nodal distance that were not confounded by order or amount of experience with the stimuli. The results imply that ordered, sequential exposure to individual stimulus relations may facilitate the development of equivalence classes and that the relatedness of stimuli within an equivalence class is a relatively permanent inverse function of nodal distance.     

Auditory reaction time and the derivation of equal loudness contours for the monkey

Monkeys were trained to release a telegraph key at the onset of a pure tone. Latency of the response was measured over a 70-db range of sound pressure (re 0.0002 dyn/cm(2)) at six frequencies (250 to 15,000 cps). Latency was found to be an inverse exponential function of intensity at all frequencies. Equal loudness was inferred from the equal latency contours which were constructed from the latency-intensity functions at each frequency. These data indicate peak auditory sensitivity for the monkey near 1000 cps. At the frequencies above and below 1000 cps consistently more sound energy was required for equal latency.     

Stimulus generalization of behavioral history: Interspecies generality and persistence of generalization gradients

Four pigeons were exposed to a tandem variable-interval (VI) fixed-ratio (FR) schedule in the presence of a 50-pixel (about 15 mm) square or an 80-pixel (about 24 mm) square and to a tandem VI differential-reinforcement-of-low-rate (DRL) schedule when a second 80-pixel or 50-pixel square was present. The values of the VI and FR schedules were adjusted to equate reinforcement rates in the two tandem schedules. Following this, a square-size continuum generalization test was administered under a fixed-interval (FI) schedule or extinction. In the first testing session, response frequency was a graded function of the similarity of the test stimuli to the training stimuli for all pigeons. These systematic generalization gradients persisted longer under the FI schedule than under extinction.     

An application of item response theory analysis to alcohol, cannabis, and cocaine criteria in DSM-IV

Item response theory (IRT) is supplanting classical test theory as the basis for measures development. This study demonstrated the utility of IRT for evaluating DSM-IV diagnostic criteria. Data on alcohol, cannabis, and cocaine symptoms from 372 adult clinical participants interviewed with the Composite International Diagnostic Interview--Expanded Substance Abuse Module (CIDI-SAM) were analyzed with Mplus (B. Muthen & L. Muthen, 1998) and MULTILOG (D. Thissen, 1991) software. Tolerance and legal problems criteria were dropped because of poor fit with a unidimensional model. Item response curves, test information curves, and testing of variously constrained models suggested that DSM-IV criteria in the CIDI-SAM discriminate between only impaired and less impaired cases and may not be useful to scale case severity. IRT can be used to study the construct validity of DSM-IV diagnoses and to identify diagnostic criteria with poor performance.     

Stimulus generalization following extra-dimensional training in educationally subnormal (severely) children.

The use of discrimination learning paradigms in the study of attentional transfer is discussed. The technique of go/no-go discrimination learning followed by stimulus generalization testing is contrasted with the more familiar simultaneous learning paradigm followed by a shift in the relevant cues. In the former paradigm the effect of training a discrimination on one dimension on the slope of the stimulus generalization gradient on an independent gradient dimension (extra-dimensional training) is assessed. A steepening of the gradient relative to appropriate control procedures is taken as evidence of positive attentional transfer. The relevance of the technique to the detailed study of attentional transfer in educationally subnormal (severely) (ESN(S)) children is considered. In Expt. I nine ESN(S) children were trained in a go/no-go discrimination involving stimuli differing in orientation, and were generalization tested on a dimension that was orthogonal, namely hue. Of the six subjects who learnt the discrimination five showed clear decremental gradients on the hue dimension. In contrast a Pseudo-Discrimination group (PD) of eight subjects matched to those in the TD group showed no gradients. These subjects were not trained in the orientation discrimination, but were reinforced for responding on 50 per cent of each of the S+ and S- stimulus presentations. They thus received equal exposure to, but no differential training on, the orientation dimension. An S+ only group of four subjects who received no exposure to the orientation stimuli showed no gradients when stimulus generalization testing on the hue continuum was carried out. The result is discussed in terms of transfer deriving from stimulus control by relational aspects of the stimuli; in terms of control by constant irrelevant stimuli; and in terms of the study of stimulus control in ESN(S) children. In Expt. II the influence of the codability of the colours on the location of the peak of the stimulus generalization gradients in the TD group is investigated.     

Peak shift but not range effects in recognition of faces

University students were trained to discriminate between two gray-scale images of faces that varied along a continuum from a unique face to an average face created by morphing. Following training, participants were tested without feedback for their ability to recognize the positive face (S+) within a range of faces along the continuum. In Experiments 1 and 4, the range of stimuli presented during testing was manipulated. In Experiment 2, participants viewed different ranges of faces during an adaptation period that followed training and preceded testing. In all experiments, generalization functions revealed peak shifts or area shifts (fewer “yes” responses to novel faces on the negative side of the S+), but no systematic effects of the test or adaptation range. Peak shift was found both for upright and inverted faces and occurred even if the orientation of the face was reversed between training and test. Using similar methods, either an area shift or range effect (but not both together) was demonstrated for line tilt stimuli (Experiment 3), and the appearance of these effects depended on instructions. It appears that peak shift and area shift are robust across many different kinds of stimuli, but range effects may not readily occur with complex multidimensional stimuli.     

Stimulus factors in aversive controls: conditioned suppression after equal training to two stimuli

Pigeons trained to peck a key for food were periodically presented with tones ending with electrical shock until tone presentation consistently suppressed ongoing pecks. Shock was then discontinued and gradients of stimulus generalization were assessed by presenting tones with frequencies above, below, and at the frequencies of those used to develop conditioned suppression. When the training tones had frequencies at 670 and 1500 cps, resulting gradients were bi-modal with peak suppression at 670 and 1500 cps. Of the other test tones, 1000 cps produced the most suppression. When the training tones had frequencies at 450 and 2250 cps, bi-modal gradients were again obtained with peak suppression to the 450 and 2250 cps tones. Of the other test tones, 1000 cps produced the least suppression. These results support the hypothesis that generalized response tendencies summate.     

The interaction of temporal generalization gradients predicts the context effect

In a temporal double bisection task, animals learn two discriminations. In the presence of Red and Green keys, responses to Red are reinforced after 1-s samples and responses to Green are reinforced after 4-s samples; in the presence of Blue and Yellow keys, responses to Blue are reinforced after 4-s samples and responses to Yellow are reinforced after 16-s samples. Subsequently, given a choice between Green and Blue, the probability of choosing Green increases with the sample duration-the context effect. In the present study we asked whether this effect could be predicted from the stimulus generalization gradients induced by the two basic discriminations. Six pigeons learned to peck Green following 4-s samples (S(+)) but not following 1-s samples (S(-)) and to peck Red following 4-s samples (S(+)) but not following 16-s samples (S(-)). Temporal generalization gradients for Green and Red were then obtained. Finally, the pigeons were given a choice between Green and Red following sample durations ranging from 1 to 16 s. Results showed that a) the two generalization gradients had the minimum at the S(-) duration, an intermediate value between the S(-) and the S(+) durations, and the maximum at the S(+) as well as more extreme durations; b) on choice trials, preference for Green over Red increased with sample duration, the context effect; and c) the two generalization gradients predicted the average context effect well. The Learning-to-Time model accounts for the major trends in the data.     

Shepard's universal law supported by honeybees in spatial generalization

An animal that is rewarded for a response in one situation (the S+) is likely to respond to similar but recognizably different stimuli, the ubiquitous phenomenon of stimulus generalization. On the basis of functional analyses of the probabilistic structure of the world, Shepard formulated a universal law of generalization, claiming that generalization gradients, as a function of the appropriately scaled distance of a stimulus from S+, should be exponential in shape. This law was tested in spatial generalization in honeybees. Based on theoretically derived scales, generalization along both the dimensions of the distance from a landmark and the direction to a landmark followed Shepard's law. Support in an invertebrate animal increases the scope of the law, and suggests that the ecological structure of the world may have driven the evolution of cognitive structures in diverse animals.     

Examining effects of training duration on humans' resurgence and variability using a novel touchscreen procedure

Resurgence occurs when a previously reinforced and then extinguished target response increases due to reducing/eliminating an alternative source of reinforcement or punishing an alternative response. We evaluated whether duration of reinforcement history for a target response (1) affects the degree to which resurgence is observed in humans and (2) produces different gradients of response generalization around target responding during extinction testing. We arranged a novel touchscreen interface in which university students could swipe a 3D soccer ball to spin any direction. In Phase 1, the first direction swiped became the target and produced points exchangeable for money for 3 or 1 min across 2 groups. The first swipe was recorded but had no programmed consequence in a third group. In Phase 2, swipes 180-degrees from the target resulted in points for 3 min in all groups. Point deliveries ceased for 2 min to test for resurgence in Phase 3. Target responses resurged during testing to a relatively greater extent with longer Phase-1 training but gradients of response generalization did not differ among groups. These findings extend prior research on the role of training duration on resurgence. We discuss methodological and conceptual issues surrounding the assessment of response generalization in resurgence.     

Discriminated timeout avoidance in pigeons: the roles of added stimuli

Two experiments examined pigeons' postponement of a signaled extinction period, or timeout (TO), from an ongoing schedule of response-dependent food delivery. A concurrent-operant procedure was used in which responses on one (food) key produced food according to a variable-interval schedule and responses on a second (postponement) key delayed the next scheduled TO according to a response-TO (R-TO) interval. A series of response-independent stimulus changes on the food key temporally partitioned the R-TO into three equal segments (S1, S2, and S3). Postponement responses, in addition to postponing TO, also reinstated S1, the stimulus correlated with the greatest temporal distance from TO. In Experiment 1, the R-TO interval was manipulated systematically across blocks of sessions (conditions) at a given ratio of R-TO:TO duration. This R-TO:TO ratio was manipulated across blocks of conditions (phases). Postponement response rates varied inversely with R-TO interval in each phase. Changes in the R-TO:TO ratio did not produce consistent differences except at the 1:10 ratio for some pigeons, where it disrupted postponement responding in some conditions. Most of the postponement responses occurred in the presence of S2 and S3, the stimuli most proximal to TO, whereas most of the food-key responses occurred in S1. In Experiment 2, the R-TO contingencies were systematically manipulated in the presence of the time-correlated stimuli. In one set of conditions, the R-TO contingencies were made either ineffective or less effective in the presence of one or more stimuli. Postponement responses typically shifted to stimuli in the presence of which responses were relatively more effective. Postponement responses decreased markedly when the added stimuli were removed, and then recovered when the stimuli were reinstated. Results from both experiments indicate that the added stimuli in a discriminated TO-avoidance procedure serve predominately discriminative functions, delineating periods during which behavior is maximally effective. The results parallel those obtained in shock-avoidance procedures, providing further evidence that TO functions as an aversive stimulus.     

Stimulus generalization following extradimensional and intradimensional training in educable retarded children

Five related experiments investigating stimulus generalization following go/no-go discrimination training of educable retarded children are reported. Experiment 1 employed an Extradimensional paradigm in which generalization testing was on the hue dimension following training on an independent (orientation) dimension. Following True discrimination training only 25% of children showed a decremental stimulus generalization gradient on the hue dimension, though all children exhibited flat gradients in Pseudodiscrimination and S+ only control groups. An increase in difficulty of the orientation discrimination in Experiment 2 did not increase the number of decremental gradients. In Experiment 3, children who exhibited decremental gradients in Experiments 1 and 2 underwent further generalization testing with modified stimuli to establish a symmetrical gradient peaked at a hue S+ to be employed in Experiments 4 and 5. In these experiments an Intradimensional paradigm was employed with S+ and S? stimuli drawn from the hue dimension. Excitatory control by S+ and inhibitory control by S? were demonstrated, as were inhibitory consequences of S? such as peak and area shift.     

Contour deletion as a method for identifying the weights of features underlying object recognition

A prerequisite for comparative work on object recognition is a method for identifying the features actually extracted from the form. The method introduced here with pigeons is discrimination training between two simple line drawings, followed by a generalization test in which contour is deleted from the reinforced drawing. In Condition 1, the line drawings were a square (S+) versus a triangle (S-); for Condition 2, the line drawings were planar projections of a cube (S+) versus a truncated pyramid (S-). The generalization decrement between responses to S+ and responses to test stimuli provides a quantitative index of the weight assigned to each feature. Contour deletion at either vertices or midsegments produced a decrement in the rate of responding, showing that each contour was represented as a feature. The generalization decrement to forms containing vertices with midsegments deleted was larger than the generalization decrement to forms containing midsegments with vertices deleted. Therefore, it appears that midsegments are weighted more strongly as features than vertices. Contour deletion provides a direct method for identifying the visual features underlying object recognition and lays a foundation for the development of comparative theories of object recognition.     

Stimulus generalization of gravity

In two experiments, squirrel monkeys were exposed to centrifugally generated, artificial gravity and trained to respond for food reinforcement at selected gravity (g) levels. Experiment I involved a single g value; in Exp. II, subjects were trained to discriminate among two or three g values. After training, generalization tests were administered over a 1.1-g to 2.1-g range. The major findings were as follows: (a) single-stimulus training yielded a linear relationship between percentage of responding and magnitude of gravity. (b) Two-valued discrimination training produced gradient peaks which were shifted from S(D) in a direction away from S(Delta). This effect was cancelled when S(D) was located equidistant between two S(Delta) stimuli. (c) Gradient form was independent of the S(D)-S(Delta) difference, but related to continuum location and/or intensity of discriminative stimuli.     

Cardiovascular generalization and differentiation: The relationship between heart rate and blood pressure

The present study was designed to study generalization and differentiation and to check the relation between heart rate and blood pressure in cardiovascular conditioning. Five naive mongrel dogs were presented with 10 sec 1000 cps tone regularly reinforced with 0.5-sec shock to a left foreleg. Four other 10-sec tones (600, 800, 1200, 1400 cps), never reinforced, were administered in a randomized and counterbalanced fashion using a differential reinforcing technique. The response measures were heart rate and three indices of blood pressure,i.e., systolic, diastolic, and pulse pressures.“Initial response” of heart rate showed rapid tachycardia to all tones, with no generalization and differentiation among five tones.“Terminal response” clearly indicated generalization gradients that tachycardia to 800 and 1200 cps tones were higher than those to 600 and 1400 cps tones, respectively. Some differences were noted among systolic, diastolic, and pulse pressures, with pulse pressure showing a trend of generalization and differentiation in the“initial response.” No statistical difference, however, was obtained in blood pressure scores. Beat-by-beat analyses of correlations between heart rate and diastolic blood pressure showed that the correlations successively decreased during the“terminal response.” Overall correlations between heart rate and three indices of blood pressure were obtained, with the highest correlation between heart rate and pulse pressure. Considering the above results in relation to the classical conditioning theory, it was tentatively concluded that“initial response” includes non-conditional factors, while“terminal response” reflects mainly a true conditional response, particularly, at a long CS-US interval. This interpretation leads to the hypothesis that conditional response with a short CS-US interval may attain better conditioning scores in terms of the summation effect between a non-conditional factor and a true conditional factor.