Navegando por Assunto "Assimetria"

Agora exibindo 1 - 4 de 4

Acesso aberto (Open Access)
Determinação da função de transferência de enlaces metálicos a partir de medições de impedância de entrada
(Universidade Federal do Pará, 2012-04-02) RODRIGUES, Roberto Menezes; COSTA, João Crisóstomo Weyl Albuquerque; http://lattes.cnpq.br/9622051867672434
The digital subscriber line (DSL) technology aims at exploiting the full potential of the telephone metallic lines on providing broadband access. On the other hand, the telephone lines may have distinct transmission capacities due to differences on their topologies. Therefore, it is important to measure the actual transmission capacity of each line before the DSL service deployment. This process is called line qualification. The determination of the transmission capacity of metallic lines requires previous determination of their transfer function. The existent qualification techniques determine the transfer function from the communication between equipments at the central office and the customer’s premises or indirectly, from knowledge about the topology of the line under test. Both processes are not in line with a pre-deployment scenario since they imply additional costs with the dispatching of technicians to the subscriber’s site, dependency of updated records about the telephone network (rarely available) or use of sophisticated line topology techniques. Therefore, the goal of this work is to propose a method for determining the transfer function of metallic lines with the following features: it does not need previous knowledge about the line topology, it uses information collected just at the central office (CO) and it does not require any human intervention at the subscriber’s site. Essentially, the general form of the proposed method analytically describes the transfer function of the line under test in function of its short and open-circuited input impedances, taken from the CO, and its asymmetry. Additionally, an algorithm that derives the short-circuited input impedance from the envelopes of the open-circuited one was developed. By applying this algorithm together with the general form of the proposed method, it is possible to determine the transfer function from just an open-circuited input impedance measurement and without any human intervention at the subscriber’s site. The proposed method was evaluated in two steps. The first step concerns the evaluation of the general form of the method. Specifically, the tests involve baseline comparison using simulated data for threeline topologies, application of the method to a bunch of simulated data generated from an arbitrary lines generator, and evaluation using measurements performed in laboratory for eight real test lines. The second step concerns the joint application of the general form of the method and the algorithm that estimates the short-circuited input impedance to measured data for two of the test lines reproduced in laboratory. The results for simulated data indicate that the general form of the proposed method is efficient, providing estimates well below the defined threshold (< 3 dB per DSL tone). For measurements, the general form of the method has estimated the transfer function of seven from the test lines with deviation per tone below 1.5 dB, but it failed for the line with two bridged-taps, being one very close to the subscriber’s site. Concerning the joint application of the general form of the proposed method and the algorithm that estimates the short-circuited input impedance, the transfer function estimations have been equivalent to those provided by the general form of the method, but with significant deviations in the beginning and in the end of the frequency band. These significant deviations are due to the current version of the envelope detector block of the algorithm that has limited performance, especially for lines with more than two serial sections.
Acesso aberto (Open Access)
Punição: uma replicação sistemática de skinner (1938)
(Universidade Federal do Pará, 2015-03-06) FIGUEIRA, Renata Almeida; MAYER, Paulo César Morales; http://lattes.cnpq.br/5360949596306254; CARVALHO NETO, Marcus Bentes de; http://lattes.cnpq.br/7613198431695463
Skinner (1938) produced a set of data that led him to conclude the effects of punishment on the response probability would be only indirect. In the main experiment conducted at the time, the initial suppressive effects of punishment (“bar slap”) disappeared after the second session of extinction. This data is still used as a proof in favor of asymmetric interpretations of punishment. The present study is a systematic replication of Skinner (1938) with four manipulations: (a) use of electric shock as aversive stimulus; (b) adoption of 10 sessions in the baseline; (c) the adoption of five extinction sessions; (d) introduction of a yoked group with non-contingent shock. Two experiments were conducted, A and B. In Experiment A, 20 rats were divided into two groups, punished group (APUN) and control group (ACON). The rats were submitted to a training session to the feeder and to lever-press shaping session followed by three sessions of FI- 4min., all lasting 60 min. Later, two sessions of extinction of 120 minutes were carried out. To APUN, superimposed to extinction during the first 10 minutes of the first extinction session, each lever-press produced an electric shock. In Experiment B, 30 rats were divided also into three groups, punished (2PUN), control (2CON) and yoked group (2ACO). The experiment started with a magazine and lever-press training session, followed by ten sessions of FI-4min., all lasting 60 min. Later, the subjects were subject to five sessions of extinction (60 minutes each). To 2PUN, superimposed to extinction during the first 10 minutes of the first session of extinction, each leverpress produced an electric shock. For 2ACO shocks were delivered according to the moment the paired subject from the punished group received it. During each session it was recorded the number of lever pressing per minute contrasting the analysis of Skinner (1938) with Boe and Church (1967) intergroup and intragroup. The results showed that, for all the manipulations performed, there was no response recovery after punishment was discontinued. This suggests it is possible to produce lasting punishment effects provided that certain methodological steps are adopted.
Acesso aberto (Open Access)
Simetria e assimetria entre reforçamento e punição
(Universidade Federal do Pará, 2014-08-22) MAYER, Paulo César Morales; CARVALHO NETO, Marcus Bentes de; http://lattes.cnpq.br/7613198431695463
In Behavior Analysis, consequences play a central role in the explanation of behavior. Reinforcement and punishment are the conceptual tools for the basic behavioral relations between the organism and its environment and to the possible resulting effects as well. Despite consensual that certain consequences may increase behavior probabilities (reinforcement) there is an extensive debate on the role of consequences in suppressing responding and the process involved in it (punishment). Two perspectives prevail on this debate, one termed symmetric, which considers reinforcement and punishment as essentially differing only on the direction of the effect of the probability of responding; and the other, named asymmetric, which, under many aspects considers punishment and reinforcement as intrinsically distinguished phenomena and establishes specific bases to consider each. The present dissertation is an investigation of the experimental bases of the asymmetric position and a proposition of re-dimensioning the debate on the issue of symmetry between reinforcement and punishment. For it, three studies investigated some asymmetric assumptions. Study 1 was a replication of Thorndike (1932, Experiment 71), investigating if the magnitude of the effects of punishment over the responding could be proportional to the ones of reinforcement. Ten college students participated a multiple choice task with the verbal stimuli “Right” and “Wrong” as consequences for the choices. As in the original study, in spite of the choices followed by “Right” having their probability increased, choices followed by “Wrong” continued to occur with probabilities close to the chance level. This data imply unequal effects for each consequence. Study 2 was a replication of Skinner (1938) assessing the extension of the suppressive effects of punishment in comparison to extinction. Six rats were trained for lever pressing on a Fixed Interval (FI) schedule and them submitted to two extinction sessions. Half of the subjects received Hot Air Blasts as consequence for lever pressing (FR1 punishment) during the initial 10 minutes of the first extinction session. Once again, the data reproduced the observed in the original study: although lever press was almost completely suppressed during punishment, responding recovered once the punishment schedule was over and by the end of the second extinction session the total number of responses was the same for both groups of subjects. Study 3 was a replication of Arbuckle and Lattal (1987) evaluating the relation between the behavior suppression produced by punishment and negative reinforcement. Seven white rats served as subjects in an operant chamber equipped with two levers and a self administration drug apparatus. After a training of lever pressing on a Variable Interval (VI) schedule, one of the levers produced pellet as consequence (VI-120s) and intravenous infusions of Histamine (behavioral suppressor) on different VI schedules (maximum value of 15s). Responses on the other lever prevented the histamine infusions (avoidance responding). Histamine effectively suppressed responding however, it was not possible to establish and sustain reliable avoidance responding, even when the contingencies were changed to promote specificvii training of unsignaled (Sidman) avoidance. By discussing the three studies altogether the difficulty of discussing the symmetry between reinforcement and punishment through a single approach arouse. Each study, despite related to the general issue discussed, was designed to answer to different dimensions of the question and the validity of such discussion would be at question. On this scenario an additional essay was written aiming at a conceptual evaluation of the symmetry issue. It is proposed a splitting of the theme under suggested categories, which would help finding new and allegedly more fruitful directions to the debate.
Acesso aberto (Open Access)
Thorndike (1932) e a assimetria entre reforçamento e punição: uma replicação
(Universidade Federal do Pará, 2015-03-06) WANZILER, Jesiane Silva; CARVALHO NETO, Marcus Bentes de; http://lattes.cnpq.br/7613198431695463
In a series of researches, Thorndike concluded that punishment does not produce direct effects as reward on behavior because some responses of participants continued to be repeated when followed by announcement of “Wrong” (punishment) and the ones followed by “Right” (reward) were all repeated. His results are widely cited to justify the asymmetric interpretation between the effects of the behavioral processes. The present work systematically replicated a study of Thorndike, through three interlinked experiments (20 university in each), in order to discuss the magnitude of the effects of punishment and to contribute to the debate on symmetry and asymmetry between this operation and the reinforcement. A vocabulary list (from a factorial arrangement of 200 or 50 questions and of five or three response alternatives) was presented six times to the participant and the proportion of responses repetition was analyzed. In Experiment 1, with printed lists, “Right” and “Wrong” were the consequences and this last was more effective (suppression of responding of all participants) on the list with 50 questions and three alternatives, confirming previous studies observing greater effectiveness of punishment as the number of items was reduced. In Experiment 2, with the automated task, there were not consistent results of suppression by “Wrong” and, therefore, revealing the manual task with the greater influence on punishment effectiveness. There were also no systematic results in Experiment 3 and the university repeated more errors even when they generated monetary loss. All experiments replicated results as the original work: repetition of some incorrect responses was not suppressed for the potentially punishing consequences. However, repetition of some correct responses was apparently weakened by the potentially reinforcing consequences. Thus, the effectiveness of the behavioral processes may be more linked to the conditions of test and the symmetry or asymmetry between them to the optical of interpretation.