Foster, Mary T.Temple, WilliamDeMello, L.R.2026-05-042026-05-041989https://hdl.handle.net/10289/18251Changes in discriminative performance as a function of procedural variations were studied, using domestic hens in a visual acuity task. Square-wave gratings and grey stimuli were presented in a successive conditional discrimination. All probe sessions were conducted with increasingly fine gratings. Several methods of analysis were evaluated and various threshold criteria used. For the first set of experiments, Experiments 1.1 to 1.5, a spatial discrimination procedure was used. A comparison of an ascending and descending series of stimulus presentation showed no consistent difference in discriminative performance. An increase in response requirements to the sample key, from FR1 to FR5 and to FR10 in Experiment 1.2, generally resulted in higher accuracy at most grating values. Increasing response requirements further, resulted in inconsistent improvements in performance. Discriminative performance under a limited range of luminances was assessed in Experiment 1.3. Data showed greater accuracy at the highest compared to the lowest luminance, for most hens and most grating values. In Experiment 1.4 the effects of training hens to stability using a finer grating showed that extended practice on a finer grating value did not consistently improve discriminative performance. A one session probe is thus an accurate description of stable performance. A comparison of performance under a controlled and uncontrolled reinforcement procedure in Experiment 1.5 showed greater overall accuracy of performance with the uncontrolled procedure. The second set of experiments were a systematic replication of Experiments 1.1, 1.2, 1.3 and 1.5, using a non-spatial conditional discrimination procedure. Results of the comparison of the ascending and descending series of stimulus presentation were replicated. Increased response requirements from FR1 to FR5 and FR10 generally resulted in increased accuracy at only the coarsest grating. The effects of changing luminance were replicated. The controlled reinforcement procedure gave slightly higher accuracy. Negative exponential functions fitted to group data described the data well and indicated a lower initial accuracy and greater rate of decay of discriminative performance as a function of grating spatial frequency, for the non-spatial discrimination procedure. Threshold estimates were consistently lower (measured in cycles/degree) for performance with the non-spatial discrimination procedure, than with the spatial discrimination task. Several methods of data analysis were used, and thresholds were estimated using various criteria. Threshold estimates changed, generally reflecting changes in discriminative performance, and as such provided a useful measure of the effects of procedural variations on discriminative performance. However, thresholds estimated from least-squares fitted lines may arguably be more useful than those taken from a unitary value, because they include more of the available data. Problems associated with this and with the use of various criteria are discussed. Furthermore, the use of a constant and variable stimulus as used in these studies, is discussed with regard to the role it may play in encouraging response biases at finer gratings. The results of these experiments indicate that the measurement of psychophysical phenomena, as exemplified by visual acuity in the domestic hen, may be consistently changed by variables which directly effect the sensory functioning, such as luminance. Discriminative performance was, however, confounded by the effects of non-sensory procedural variations, and the most important of these appears to be task difficulty.enAll items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.Thesis