The influence of performance on action-effect integration in sense of agency

Sense of agency refers to the subjective feeling of being able to control an outcome through one’s own actions or will. Prior studies have shown that both sensory processing (e.g., comparisons between sensory feedbacks and predictions basing on one’s motor intentions) and high-level cognitive/constructive processes (e.g., inferences based on one’s performance or the consequences of one’s actions) contribute to judgments of sense of agency. However, it remains unclear how these two types of processes interact, which is important for clarifying the mechanisms underlying sense of agency. Thus, we examined whether performance-based inferences influence action-effect integration in sense of agency using a delay detection paradigm in two experiments. In both experiments, participants pressed left and right arrow keys to control the direction in which a moving dot was travelling. The dot’s response delay was manipulated randomly on 7 levels (0–480 ms) between the trials; for each trial, participants were asked to judge whether the dot response was delayed and to rate their level of agency over the dot. In Experiment 1, participants tried to direct the dot to reach a destination on the screen as quickly as possible. Furthermore, the computer assisted participants by ignoring erroneous commands for half of the trials (assisted condition), while in the other half, all of the participants’ commands were executed (self-control condition). In Experiment 2, participants directed the dot as they pleased (without a specific goal), but, in half of the trials, the computer randomly ignored 32% of their commands (disturbed condition) rather than assisted them. The results from the two experiments showed that performance enhanced action-effect integration. Specifically, when task performance was improved through the computer’s assistance in Experiment 1, delay detection was reduced in the 480-ms delay condition, despite the fact that 32% of participants’ commands were ignored. Conversely, when no feedback on task performance was given (as in Experiment 2), the participants reported greater delay when some of their commands were randomly ignored. Furthermore, the results of a logistic regression analysis showed that the threshold of delay detection was greater in the assisted condition than in the self-control condition in Experiment 1, which suggests a wider time window for action-effect integration. A multivariate analysis also revealed that assistance was related to reduced delay detection via task performance, while reduced delay detection was directly correlated with a better sense of agency. These results indicate an association between the implicit and explicit aspects of sense of agency.     

The CIOMS view on the use of placebo in clinical trials

Based on worldwide consultations with experts in science and ethics the revised CIOMS 2002 International Ethical Guidelines for Biomedical Research Involving Human Subjects provide guidance on when the use of placebo as a comparator in clinical research is ethically acceptable. The article reviews the main points of the CIOMS Guidelines and commentaries including the use of placebo in situations where the best current method is available and the relation of placebo to established effective intervention. It discusses the use of placebo in externally sponsored research in low-resource countries and requirements for informed consent related to placebo studies. An earlier version of this paper was presented at an international conference, “Placebo: Its Action and Place in Health Research Today,” held in Warsaw, Poland on 12–13 April, 2003.     

The influence of goals on sense of control

We examined the influence of goals on sense of control relative to that experienced when taking action randomly. In the experimental task, participants controlled the direction of a moving dot by pressing the left and right keys at will without a specific goal (the control condition), directed the moving dot to a destination as often as possible (the strong goal condition), or kept the moving dot in the central area of the screen (the weak goal condition) for as long as possible. The results showed that the strong goal impaired the sense of control, but the weak goal did not exert an influence. We concluded that the goal-based expectation influenced sense of control, but the goal-directed action selection did not. Furthermore, we proposed a modified comparator model of the sense of control, offering a promising approach to integration of the predictive and postdictive processes involved in the sense of control.