|dc.description.abstract||Motivating Operations (MOs) are a fundamental concept in behavioural psychology. Despite this, empirical research into MOs is lacking. The overall aim of this thesis is to contribute to the experimental literature base for MOs.
Experiment 2.1 used an already published video analysis methodology to assess the morphology of food motivated pecks made to a computer screen by hens, after the hens had been trained to emit the peck using either an autoshaping or handshaping procedure. The intention of this was to then be able to use the video analysis to assess the effect of altering two MOs related to two different reinforcers (e.g., food and water) at one time, on morphology. The study showed that both methods produced similarly formed pecks despite the variability inherent in the handshaping procedure. It was then concluded it is the nature of the reinforcer that gives rise to morphology not the autoshaping procedure per se which gives rise to a particular form of elicited responses.
The aim of Experiment 3.1 and 3.2 was to develop a procedure for restricting access to water in laying hens, in order to motivate them sufficiently to respond for water reinforcers. Experiment 3.1 assessed the effect that gradually decreasing time and amount of water access had on food-restricted hens’ water consumption and health. It was found that hens could be restricted to one hr access of water (restricted to the maximum amount that hens would consume when access was ad libitum) without adverse effects to health being apparent. However, when the hens were subsequently exposed to FR schedules with a low response requirement in Experiment 3.2, they did not respond consistently. This indicated that the level of restriction was insufficient to motivate responding and this finding, combined with the difficulty of obtaining ethical approval, meant that the proposed experiments utilising water deprivation as an MO had to be abandoned.
Experiment 4.1 used the autoshaping paradigm to assess the acquisition of food motivated pecks to a stimulus, at two different levels of bodyweight (75% and 95%). An infra-red screen was used to analyse performance separately from learning effects by examining activity levels (location and amount of pecks). It was found that that higher numbers of effective pecks were made by hens maintained at 75% free-feeding bodyweight than hens maintained at 95% (different MO conditions). There were also higher levels of ineffective pecks in the 75% group.
Experiment 5.1 investigated relative preference for stimuli correlated with different MO conditions: high deprivation (no pre-feeding), or low deprivation (pre-feeding), when subjects were maintained at either 75% or 95% of free-feeding bodyweight. The results showed that 6/10 hens demonstrated an increased preference for the stimulus paired with high deprivation conditions (no pre-feeding) when measured by log ratios of responses, and had faster response rates on this stimulus. Overall, the 75% bodyweight hens had faster response rates than the 95% hens (as in Experiment 4.1), and 8/10 hens responded faster on the stimulus that was paired with no pre-feeding. It was also found, as per Experiment 4.1, that higher numbers of effective pecks were made by hens maintained at 75% free-feeding bodyweight than hens maintained at 95% (different MO conditions).
Experiments 6.1 and 6.2 extended the findings of the thesis thus far in that concurrent VI VI schedules were used to assess the effect of bodyweight and pre-feeding as MOs on steady state responding. In total 16 conditions were run exposing hens to three different VI pairs: VI-12, VI-60 (5:1); VI-20, VI-20 (1:1); and VI-60, VI-12 (1:5). Bodyweight values of 85%, 95%, 100%, and 85% with pre-feeding of 40 cc wheat delivered 40 minutes prior to experimental sessions were manipulated between hens finishing a series of the three VI pairs. It was found that 4/6 hens had higher absolute and relative response rates when bodyweight was made lower. For 3/6 of these hens, increasing bodyweight increased sensitivity as measured by the parameter a; this was more distinct when the Generalised Matching Law was applied to response rather than time locations for these hens. Frequency distributions of IRTs showed that for the hens that tended to show increasing sensitivity as bodyweights increased there were more IRTs in bins greater than 0.4 s. This was reflected on the log-survivor plots as the limbs were shallower when bodyweights were higher, indicating that more between-bout responses were occurring. It was also found that pre-feeding increased sensitivity as measured by the parameter a for all hens; this was more noticeable when the GML was applied to response rather than time allocations. Although overall response rates tended to resemble those for the 85% bodyweight condition and remain higher than the 95% and 100% bodyweight conditions, the distribution of left and right response rates showed that hens matched better to the prevailing reinforcer rates when they were pre-fed, than when they were not pre-fed.
Overall, the main findings were: (1) that reducing bodyweights increased amounts of species-specific behaviour; and (2) that reducing bodyweight causes increases in response rate. These findings could explain why changes in preference for stimuli paired with high levels of deprivation are observed during SDVL procedures, and why increased sensitivity to available reinforcement at lower levels of deprivation found in studies utilising the GML have been observed in previous studies. These findings contribute to the empirical data informing the behavioural treatment of motivation and have applied implications. Reinforcement and punishment procedures such as extinction or differential reinforcement of alternative behaviours may no longer be necessary when MOs are manipulated.||