|
| International Paper Session - Experimental Analysis of Human Behavior |
| Sunday, May 27, 2007 |
| 4:00 PM–5:20 PM |
| Madeleine AB |
| Area: EAB |
| Chair: Jade Hill (Jacksonville State University) |
| |
| Effects of Instruction Variation and Accuracy upon Behavioral Sensitivity. |
| Domain: Basic Research |
| JOSELE ABREU-RODRIGUES (Universidade de Brasilia), Myriam Christina Alves Rdrigues (Universidade de Brasilia) |
| |
| Abstract: The present study investigated behavioral sensitivity to contingency change after a reinforcement history with varied or specific schedules and instructions that could be either accurate or inaccurate. College students were assigned to two groups that differed in the training phase. Participants in the Varied Group were exposed to four schedules (VI, VR, FT, and FR), and those in the Specific Group to one schedule (FR). Each group was divided in three subgroups. For the Accurate Groups, there was correspondence between instructions and schedules; for the Inaccurate Groups, instructions did not correspond to the schedules; and for the Control Groups, there were no instructions. In the testing phase, all participants were exposed to a FI schedule. The Accurate Groups were similarly insensitive regardless of a history with varied or specific schedules and instructions. Also, the Varied Inaccurate and Varied Control Groups were more sensitive than the Specific Inaccurate and Specific Control Groups. These results suggest that a history with varied schedules promote greater behavioral sensitivity than with only one schedule. When instructions are provided, however, this difference in sensitivity is observed with inaccurate instructions (more precisely, when there is contact with the discrepancy between instructions and schedules), but not with accurate instructions. |
| |
| Research Participation Under Informed Consent: Factors Influencing the Choice to Withdraw. |
| Domain: Basic Research |
| DOUGLAS NAVARICK (California State University, Fullerton) |
| |
| Abstract: After receiving a standard informed consent statement allowing withdrawal without penalties, college students repeatedly chose between a 25-second cartoon video followed by a 5-second timeout and a 5-second cartoon video followed by a 25-second timeout in two, 15-minute segments. After showing a preference for the former schedule, they were instructed to choose the latter. Of 15 participants, 14 complied on almost all trials and none withdrew. In an Escape group, the withdrawal process was simplified by placing a desk bell on the console and instructing participants that they could end the session by tapping it. In Escape/Approval, participants were additionally instructed that sufficient data were already collected and leaving would not interfere with the research. In Escape/Approval/Norm, participants were also instructed that most participants did leave before the experiment was over. Withdrawal rates rose from 0, to 13, to 40 percent, respectively. Almost all withdrawals occurred on the early trials, a result predicted by Delay Reduction Theory due to the increasing advantage of staying over quitting as trials increased, signifying progressively shorter delays to conventional escape. Possibly the most powerful variables in psychological experiments are also the least investigated: the contextual factors that keep participants in their seats. |
| |
| Number of Repetitions: A Learning History Variable. |
| Domain: Basic Research |
| LAURILYN DIANNE JONES (The Mechner Foundation), Francis Mechner (The Mechner Foundation) |
| |
| Abstract: An issue in experimental behavior analysis is how learning history variables interact with prevailing contingencies in determining the properties of the learned operants. Single-subject studies of the effects of learning history variables require the comparison of matched sets of equivalent operant classes that have received different learning histories and were learned by the same individual.
One important history variable is the number of repetitions of an operant class during learning. The operant classes used were magnified operant units consisting of short keystroke sequences executed on a computer keyboard. Subjects repeatedly learned groups of six different but equivalent operants, half of them practiced a smaller number of times and half practiced a larger number of times. When preferences between these were examined under subsequent test conditions using a forced-choice format, it was found that the effects depended on both the ratio of repetitions (high vs. low) and the absolute values. When the lower number of repetitions was high enough, the high-repetition alternative lost its advantage, even if the ratio remained the same. These results may be due to repetition-induced automatization. In addition, cumulative learning effects were observed.
In previous single-subject history experiments, learning sequence was shown to produce both recency and primacy effects. |
| |
|
| |