Data from twelve auditory signal detection experiments show that a model general to all signal detection tasks explains results better than specific auditory models. This thesis examines models of the detection of sinusoids in Gaussian white noise, all predicting linear ROC-curves on normal-normal co-ordinates, but differing in their predictions for the ROC-curve slope. Distinction is made between stimulus distributions, transducer distributions and response-inferred distributions. Response-inferred distributions include variance from stimuli, transducers and unstable criteria. The first three experiments (rating) showed that slope increases with increases in p(sn) and in the number of categories available for describing the presence of the signal. An explanation for this data assumes that the 'yes-no' criterion has the least variance and other criteria have variances proportional to their distances from the 'yes-no' criterion. This explanation is developed into a model of selective attention in which the variance of all real criteria is a function of their distances from an optimal criterion. Faulty memory is the assumed cause of criterion variance. Predictions that follow from the model are (i) there is a decline in criterion variance as signal strength increases ; (ii) there is a U- (or inverted U-) shaped function relating slope and signal strength; (iii) criterion variance is less in forced-choice tasks than in 'yes-no' tasks; and (iv) slope is partially determined by task design and any other factor which affects the subject's memory for signal or noise. Experiments 4-6 establish that the concept of criterion variance also applies to yes-no procedures, and Experiments 7-11 substantiate predictions made by the model of selective attention. Data from Experiments 5, 7, 8, 9 and 10 are analysed in terms of models of sinusoidal burst and gap detection. None of the probabilistic models is adequate, but Zwislocki's (1969) deterministic model of temporal summation accounts for the data. A model of response-inferred distributions is presented in which the mathematical relationships of the variances and locations of criteria, of signal strength and of stimulus variances to the slope of the ROC-curve are described. Equations for the optimal criterion, the variance of individual criteria and a measure of sensitivity uncontaminated by criterion variance are derived. Re-analysis of data using the model of response-inferred distributions supports the predictions of the model of selective attention. The model of response-inferred distributions predicts 'peaked' rating ROC-curves which are similar to two-state functions and commonly observed in the literature. However, both two-state and high-threshold theories are unable to explain the data. Experiment 12 compares the model of response-inferred distributions with Pike's (1973) multiple observations model for latency data and finds the latter model inferior. Criterion variance is shown to account fox at least half of the variance of the response-inferred distributions, and consequently, it is argued that TSD results should be viewed more in terms of the general processes which produce criterion variance than modality-specific models of signal processing.

Eight marsupial phalangers, (Trichosurus vulpecula) with mid-saggital section of the optic chiasma were trained monocularly in a visual discrimination task, as were four which had, in addition, mid-saggital sections of the anterior commissure, fasciculus aberrans and hippocampal commissure. All were tested for transfer to the untrained hemisphere. The results showed that those with the commissures intact transferred information to the hemisphere not trained directly, while those with commissures sectioned did not. It appears that transfer of visual information between cerebral hemispheres can take place in an animal which lacks a corpus callosum, and that the forebrain commissures may be the functional equivalent, in this marsupial, of the corpus callosum in eutherians. A further eight animals received mid-saggital section of the optic chiasma. Two had all other interhemispheric pathways intact, two had mid-saggital section of fasciculus aberrans and hippocampal and anterior commissures, two had fasciculus aberrans only sectioned, and two had fasciculus aberrans only intact. All were trained monocularly in a visual discrimination task and tested for transfer to the opposite hemisphere. Those with all commissural pathways intact and those with only fasciculus aberrans intact demonstrated transfer. Those with all commissures sectioned and those with only fasciculus aberrans sectioned did not show transfer. It seems probable that integrity of fasciculus aberrans is a necessary and sufficient commissural condition for interhemispheric transfer of visual information in this marsupial.

Organisms trained to respond at a high rate to one stimulus, S1, and at a low rate to another stimulus, S2, lying in the same physical dimension as S1, may, when presented successively with a number of stimuli lying in the dimension, respond at the highest rate to a stimulus away from S1 in a direction opposite S2. This "peak shift" effect has been found following training in which S1 and S2 were presented successively, but not when they were presented simultaneously. In the present study peak shift was obtained when S1 and S2 were presented simultaneously during training. In order to isolate the conditions necessary to produce peak shift after simultaneous training, different groups of pigeons were given variations in procedure, but peak shift was obtained with all groups. The variations included using stimuli from various physical dimensions, using different training procedures, and using various methods of presenting stimuli during the tests for peak shift. The results were generally more consistent with discrimination theories that postulate the interaction of generalisation gradients around S1 and S2, rather than with theories that postulate control by some relationship between S1 and S2, and in particular, they were most consistent with explanations of peak shift that propose that the responding occurring in the presence of S1 is inhibited in the presence of S2.

Experimental extensions and theoretical explanations of a psychoacoustic experiment concerning temporal summation of human hearing are investigated. The experiment has the well-established result that, ceteris paribus, a brief tone burst is more readily letected than a brief gap in a tone. Two models, one based on the neuroelectrical activity of the auditory nerve and one on energy-detection theory, are presented and developed in detail: both are shown capable of predicting the result of the bursts and gaps experiment. Seven experiments are reported. Experiment 1 found the reaction time to the start of a faint tone embedded in noise generally shorter than that to its end. Experiment 2 examined the effect of rise/decay time on the detectability of brief bursts of, or gaps in, broadband noise: when the energy change was constant, 1here was no effect except at long (100 msec) rise/decay times. Experiments 3 and 4 assessed the detectability of bursts of, and gaps in, a tone embedded in noise where energy change to noise power per cycle was constant but duration varied. The data demonstrate that the effect of duration on detectability was different for burst than gap events, a result predicted by both models. Experiment 5 showed that, when the energy in and duration of a tone burst in noise were constant but the temporal distribution of the burst energy varied, detectability remained unchanged. The repetition rate of noise bursts and gaps in continuous noise was varied without effect on detectability in Experiment 6. Bursts, however, were consistently mere detectable than gaps. Experiment 7 compared the detectability of increments and decrements of a continuous tone embedded in noise. The quantitative predictions of the two models are compared with the data resulting from Experiments 3, 5, and 7. The qualitative results of these and the other experiments are also discussed in terms of their theoretical implications. Taken all together, the experiments do not definitely favour either model over the other.

Background Depictions of mental illnesses in the mass media have been analysed and criticised for more than forty years with little improvement, and that has serious implications for the ongoing efforts to destigmatise both mental illnesses and those who suffer from them. Aims To examine media depictions of mental illnesses within a social constructionist framework. To identify why media depictions take the form they do and to indicate ways in which such practices may be more effectively addressed. Method Items chosen from factual media genres were subjected to discourse analysis. This form of analysis attends to the preferred meaning of the items and how that meaning is constructed within the item. Exemplars of such analysis are contrasted with the content analyses more commonly performed on media materials before reporting analyses of items from everyday media reports and of materials that were expected to be less stigmatising. Results Irrespective of the form of analysis it is found that media depictions of mental illnesses are dominated by representations of dangerousness, criminal violence, unpredictability, and social incompetence. The same features were found in a destigmatisation documentary and a series of backgrounders on mental health services, in both of which madness was utilised to create interest and drama. It is argued that these characteristics occur because media personnel, like most laypersons, represent mental illnesses as forms of madness. Conclusions That my social constructionist analysis is able to account for the lack of change in media depictions over forty years. That the preference for a public mental health approach to destigmatisation is misplaced because it is unable to address the fear generated by lay understandings of mental illnesses. That the attempt to avoid conflation of the person with the disorder in Diagnostic and Statistical Manuals beginning in 1980 was an inadequate step in an appropriate direction in that it sought to remind clinicians that a mental disorder does not make a person non-human. The thesis findings are interpreted as showing that destigmatisation requires a new way of depicting mental illnesses, one that privileges the individual's experience and their ordinary humanness.

Pigeons were studied in concurrent chain schedules in which the terminal links were variable-interval (aperiodic) and fixed-interval (periodic) schedules. Choice was studied at various maintained percentages of body weight, (Experiment 1) and after various amounts of food had been pre-fed (Experiment 2). No effects on choice of the level of deprivation were found except in those choice conditions in which both terminal-link schedules were variable-interval schedules. Choice behaviour was further studied with equal, but increased magazine durations (Experiment 3) and no reliable effects were found. Explicit discrimination training established control over choice behaviour by the amounts pre-fed (Experiment 4) and generalisation gradiets of choice showed that this control was reliable. The differences found between aperiodic and periodic schedules were probably related to the different patterns of behaviour maintained by these schedules and an attempt was made to relate the results to overall models of behaviour and the 'value' of food.

It is often claimed that applied behaviour analysis is founded on basic behavioural concepts involving the direct extrapolation of data from the experimental laboratory to the analysis of human behaviour. It is claimed that such an approach gives theoretical coherence to a set of procedures, and avoids the pitfalls associated with a collection of unrelated techniques. This presumed experimental base is said to make applied behaviour analysis more rigorous, more effective, more systematic, and easier to teach and learn than other approaches to the modification of human behaviour. There is reason to believe, however, that the link between experimental and applied behaviour analysis is more one of commitment than reality. Previous authors have shown that research findings from the experimental analysis of behaviour are becoming increasingly isolated from other areas of psychology in general. The present study provides evidence that such findings are also isolated from applied behaviour analysis in particular. Experimental data are seldom cited in applied work and, when they are, it is seldom specific or current data to which references are made. It is argued that this state of affairs is attributable to the acceptance, by applied behaviour analysis, of a model of behaviour whose complexity is insufficient to cope with the complex interactive nature of human behaviour. The present work provides an outline of the current applied behaviour analytic model, and then proceeds to discuss some of the major types of data of which the model takes no significant account. An extensive examination of data from the experimental analysis of choice behaviour is given, which reveals the level of complexity to which application may be made. Suggestions as to the practical utilization of these data are made, with particular reference to the areas of self-control and commitment. A case study involving the use of a commitment procedure in the treatment of an exhibitionist and another concerning the application of experimentally-derived procedures to a self-control problem, provide clinical evidence of the usefulness of the direct incorporation of experimental data. Following these applied case studies, a number of other areas of experimental research are examined with respect to their relevance to applied behaviour analysis. Significant data from the study of multiple-schedule interactions, stimulus control, and two specific types of analyses of reinforcement phenomena are outlined and suggestions made concerning their applied potential. Further, data which could determine the limits of the applicability of experimental data, in the context of biological constraints on behaviour, are shown to be capable of further increasing the utility of applied behaviour analytic techniques. Finally, the implications of the data analyzed are discussed in terms of the future training of applied behaviour analysts.

Lesions were placed in several areas of the telencephalon and diencephalon of the brain of the pigeon and the effects on the acquisition of inhibitory dimensional and inhibitory stimulus control were observed. The experimental tasks consisted of both visual and auditory interdimensional discriminations each of which had two components: In the first the stimuli were presented successively on the one response key (the main key) according to a multiple variable-interval extinction schedule. In the second the multiple schedule still obtained but a changeover key was added which, when pecked, changed the main-key stimulus, together with its associated schedule of reinforcement, to the next in a randomly ordered series. The use of these two components allowed the separation of two aspects of inhibitory control in learning: response reduction and stimulus reduction thereby permitting the assessment of discriminative ability in the absence of the confounding factor of response reduction. Inhibitory dimensional and inhibitory stimulus control were assessed by post-discrimination generalization tests and combined-cue tests, respectively. It was found that lesions to areas considered limbic - hippocampus, septum, anterior dorsomedial thalamus - had no effect on the learning of a visual discrimination. Lesions to the dorsolateral thalamus produced a complete inability to learn this discrimination, presumably due to disruption of visual fibres en route to the telecephalon. Lesions to the Wulst produced a visual discrimination learning deficit in some birds but not in others, an inconsistency not accounted for by differences in lesion size. Wulst lesions also produced an auditory discrimination learning deficit and in this case the larger the lesion, the larger the deficit. Lesions to ectostriatum produced a deficit in the visual task and lesions to Field L, an auditory projection area, produced a deficit in the auditory task, but not in the visual task. However, in most cases the discrimination learning deficit which was produced was confined to the multiple schedule where the animal had no control over the presentation or duration of the stimuli. Once the changeover key was introduced most birds obtained the learning criterion very quickly by "switching out" of the negative stimulus. This effective changeover responding, together with unimpaired inhibitory dimensional or inhibitory stimulus control, suggested that although the initial learning deficit might be described in terms of an inability to withhold responding in the presence of stimuli previously correlated with reinforcement, this inability could not readily be explained in terms of a lesion-induced impairment in an inhibitory process.

The purpose of this study was to investigate the immediate consequences of concussion. Previous work had either concentrated on memory defects only, included cases of cerebral damage as well as concussion, or examined residual defects persisting after the period of impaired consciousness had ended. It was hoped to determine which function or functions might be affected in concussion by testing only young adult cases of 'pure' concussion in the immediate post-traumatic period. In the main experiment a paced serial addition task was administered to a mildly concussed (MC) and seriously concussed (SC) group at twenty-four-hourly intervals during the hospital period, and thirty to forty days after discharge. A normal control and hospital control group were given the task at similar time intervals. Recent concussion impaired performance and SC Ss needed almost five times as long as controls to process each correct response while MC Ss were about twenty per cent slower, However, responses were qualitatively similar in all groups, and concussed Ss differed only in producing many more late response errors. At retest the MC group had regained control level, but although SC Ss had significantly improved PASAT scores they were still poorer than the other three groups. Experiments II and III investigated stages in the information transmission process which may have produced restricted information processing shown by PASAT results. Reaction and movement times were not longer than normal, although central processing time was. MC Ss did not have significantly slower non-symbolic RTs, or symbolic two- and four-choice RTs, but they differed significantly from controls on eight- and ten-choice symbolic trials. Again this difference between groups was not found at retest four weeks later. Results suggested inefficient 'pigeon-holing' (Broadbent, 1971) as a possible factor in reduced information transmission following concussion. However, a message repetition task indicated no difference between concussed and non-concussed Ss in use of this mechanism. Message repetition also demonstrated that patients had no difficulty in auditory perception of verbal material. In the third stage, normal Ss were given message repetition with and without a concurrent distracting task. Performance during distraction was sufficiently similar to that of recently concussed Ss to suggest that the patient group may have been poorer than controls on the same task only because they also had reduced processing capacity. When given PASAT with a secondary paced task, performance of another group of normal Ss was almost identical to concussed Ss in the first experiment. Finally, to test the possibility that reduced processing capacity was the result of increased distractability, that is, processing task-irrelevant stimuli, an attention task (speech shadowing) was given to a small group of MC Ss and normal controls. There were no instances of intrusions or interference from an irrelevant message to indicate a defect in selective attention, although MC Ss had significantly lower shadowing scores. It was concluded that reduction in processing capacity produced by concussion was a function of patients' lower level of arousal, and that performance resembled that reported in the literature from sleep-deprived Ss and cases of brain-stem damage or dysfunction. A feature of results was the extremely competent performance exhibited by concussion patients, given a reduction in total capacity. Similar optimizing behaviour was evident during distraction tasks, and it was suggested that ability to monitor and control distribution of available processing space by instructions ('set') is a general characteristic of ordered behaviour. This ability was not disturbed by degrees of concussion sustained by experimental Ss in this study. Amnesic symptoms of the immediate post-traumatic period were considered in terms of lowered arousal level. It was proposed that these might be explained by a combination of two factors: (i) processing space insufficient to store items while responding to them, and (ii) tagging of items in storage with levels of concurrent background activity, and thus 'state-dependent' memory. Evidence for a processing deficit existing after termination of the period of PTA was reviewed, and the possibility of permanent effect in terms of increased disability following repeated concussions. Some aspects of results important in considering rehabilitation programmes for head-injured patients were noted.

A series of experiments was undertaken to explore the effect of interaural differences in the intensity of masking noise upon the detection of a signal. The signal was a 2-kHz sinusoid, and the masker was composed of either one or two 800-Hz wide bands of noise. The centre frequencies of the two bands of noise, one above and one below the signal frequency, were varied. On most occasions both noise bands were used to create a spectral notch surrounding the signal. The following factors were manipulated: (1) The width of the notch: from 0 to 1900 Hz. (2) The location of the lower and upper frequency edges of the notch relative to the signal: either equidistant from the signal, or with one edge 150 Hz nearer to the signal than the other edge. (3) The spectrum level of the noise: either 20, 35 or 50 dB SPL. (4) The degree of interaural disparity in the intensity of some components of the noise: either 0 dB, 10 dB or infinite. (5) The ear to which a particular masking noise was sent. (6) The manner in which the noise was presented: either diotically or dichotically. (7) The manner in which the signal was presented: either diotically or monotically. The results of the first two experiments were interpreted as showing that threshold signal levels in the presence of interaural differences in the intensity of masking noise depended principally on the ear in which the signal-to-masker ratio at the output of the auditory filter was larger. To test this possibility, auditory filter shapes were derived from two listeners. These were then used to predict thresholds when there were interaural differences in the spectral envelope of a masking noise. The results of a comparison between the predicted and obtained thresholds were consistent with the previous interpretation. Thus it appears that the detector following the filter can discriminate the output of the two ears and base detection on the better output. This ability may be useful in normal listening, where both wanted and extraneous sounds are subject to change from moment to moment.