Physiological measures were collected during Experiment 2 only. To test if autonomic physiology variables added additional predictive power compared to self-report, we included 5 PCA components extracted from the autonomic variables to the PCA components from self-report, and then predicted intervention condition.
Physiology was extracted from three tasks questionnaire, emotional memory, situation construal , and each was added to the self-report variables in separate models see Table Further, models trained only on physiology resulted in chance test performance. Table Train and test accuracy for machine learning models of autonomic variables during the questionnaire task. Prediction of the intervention condition from EEG components yielded chance performance for all models. This is not surprising given the low sample size. To investigate if cognitive and physiological measures are correlated with self-report variables in the context of our intervention, we ran series of models predicting each self-report measure from cognitive or physiological measures.
We found one significant model, where Positive Affect was predicted by physiological variables during the emotional memory task Table Regression coefficients for predicting positive affect from physiological variables during emotional memory task. Well-being science, and the field of social psychology in general, have traditionally focused on self-report to measure target variables Baumeister et al. However, the potential bias apparent in these measures, even when rigorously validated, has propelled interest in alternative, more objective approaches. Here we investigated the efficacy of cognitive and physiological measures in determining the effect of positive versus negative activity interventions on well-being.
While self-report variables produced significant results, we were unable to detect robust effects of well-being change in any cognitive or physiological measures. Our null results were possibly the results of chance, low power, or methodological limitations see below , and we cannot conclusively establish whether our intervention had no cognitive or physiological effect.
Additionally, we found limited evidence for a relation between any individual cognitive or physiological measures and any specific self-report variable. Although we did not detect significant effects of positive versus negative activity manipulations using cognitive and psychological measures, it is important to note some of the limitations of our approach and outline recommendations for future work.
While considered more objective, biological indicators are not themselves unambiguous measures of human happiness or unhappiness Oswald and Wu, Previous theoretical work has argued that only arousal levels, and not the valence of emotion, are detectable via physiology Schachter and Singer, In our design, we pitted induced negative affect versus positive affect, and, while the valences were clearly different, the arousal levels may have been similar, resulting in a null effect.
We quantified physiological measures while subjects performed cognitive tasks, thereby allowing for event driven analysis. We therefore recommend future studies also record physiological measures during a prolonged period of low arousal, where no stimulus is present. Further, physiological signals are inherently noisy, and prone to interference from movement, sweat, or electromagnetic sources. Future investigators should be cognizant of their low signal to noise ratio, and the need for substantial preprocessing i.
Furthermore, reducing noise sources e. The devices used in our research Feel wristband, Modified Emotive Epoch , while low-cost and wireless, were found to be limiting for rigorous research methodology. Neither device had the ability to embed stimulus triggers. The Feel device was synced with the stimulus computer, and it was assumed that its internal clock was robust and remained synchronized throughout the experimentation procedure.
Furthermore, piloting revealed that the Emotiv device was becoming unsynchronized during the experimentation procedure. This was overcome by co-opting two EEG channels as a bipolar stimulus channel, driven by a photodiode connected to the stimulus presentation computer. It is possible that the addition of the synchronizing photodiode added artifacts that were not removed by our artifact rejection procedure, and this may have influenced our results.
Further, the low sample rate Hz of the device may have hidden effects. Hence, the lack of significant EEG results may be due to low sample size, poor device quality, or artifacts. While considerable care was taken to ensure all signals were synchronized and denoised, the process was time consuming, and much could have been mitigated with the use of dedicated research grade devices.
Although we selected self-report, cognitive and physiological measures based on their prior utility and theory, our goal was not to develop a model that unifies self-report and physiological measures. However, we note the lack of any published theoretical model linking these measures.
A useful addition to the field would be a set of repeated measures studies to determine the relationship between these measures in a variety of contexts, focusing on measure reliability, as well as their convergent and discriminate validity.
Conscious Architecture: Spatial Sensation and the Mindful Body by Benjamin Daniel Fisher - Issuu
Due to the exploratory nature of this study, a model-free analysis approach was taken, where variables were combined using a number of machine learning techniques to predict the intervention positive vs. We made heavy use of cross validation techniques to alleviate the possibility of overfitting or false positives. We highlight this as especially important, because due to the sheer number of variables produced by each device, and the number of statistical comparisons that could be made, false positives are highly probable.
Previous literature highlights the inconsistency of physiology in the detection of affect, as well as the possibility of Type 1 errors given that, in each Experiment, a different set of variables was found to be predictive; Calvo and Mello, Therefore, we recommend future studies use appropriate methods to protect against spurious results coupled with appropriate power analysis.
Another promising avenue includes model-based analyses, such as those employed in Bayesian cognitive modeling M. Lee and Wagenmakers, Here, a number of specific theories that build in biologically plausible relations as priors as parameterized by models can be compared, and conclusions drawn are graded, where one theory is favored X time more than the other, as opposed to the all or nothing null hypothesis testing approach.
If we take the self-report variables as unbiased, and a small effect as given, then this small effect may simply not have been sufficient to produce a detectable cognitive and physiological response. Future studies should increase power via a larger sample size, or attempt to increase intervention effect either by including additional well-being activities [e. The low effect size also suggests high variance in our measures.
We recommend the addition of a baseline task in future studies, where the physiological response to a large number of highly emotionally arousing stimuli is recorded in the same participants and used to verify a detectable physiological change. This approach would also allow for the exclusion of subjects with low physiological responses, as well as for individual tuning of methods with the caveat that the results would not generalize to the full population.
We include a table of means and standard deviations of all measures in Supplementary Material Supplementary Tables S6—S11 , as a means for future investigators to estimate the expected effect size and hence required sample size of future studies. Further, no neutral condition was included, in which, for example, participants might be asked to recall memories without any strong emotional component. As a result, we could only compare positive and negative conditions. If some measures respond only to the arousal component of emotion as argued previously , then we would be unable to detect a difference.
In conclusion, our study replicated previous research on the beneficial effects of writing about prosocial events as quantified by self-report.
- MegaZEN Well-Being Journal - Volume 6 by MegaZEN - Issuu.
- Consciousness of Health: Tuning in to Absolute Well-Being?
- Be the Change You Want In Your Life;
- Kickstart Your Life.
Furthermore, although we did not clearly demonstrate the cognitive effects of a PAI, neither did we obtain conclusive evidence for a lack of cognitive of physiological effects. We believe this paper represents a first step in introducing more cognitive paradigms into the positive activity literature, and set a precedent for the use of more objective measures in such research. However, in light of limitations of the methods and study design, we recommend that the field considers these measures, while bearing in mind our recommendations for future work. More research is needed to investigate the conditions under which these measures may be feasible and useful, but we stress that they should not be unilaterally favored over the traditional self-report approach.
The datasets generated for this study are available on request to the corresponding author. Students reviewed and signed a consent form which detailed the study procedures. BY contributed to conceptualization, methodology, data curation, formal analysis, investigation, literature review, visualization, and manuscript preparation. JR contributed to conceptualization, methods, methodology, investigation, and literature review. SM contributed to conceptualization, methodology, and investigation. SL and AS initially conceived the study, administrated the study, and oversaw methods and investigation.
All authors edited and approved the final manuscript. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Consciousness of Health - Tuning in to Absolute Well-Being (Paperback)
The authors would like to thank Panagiotis Fatouros, Haris Tsirbas, and George Eleftheriou at Sentio, for their donation of the Feel device and continual support. The authors would also like to thank the hard work of the research assistants at University of California, Riverside who helped collect the data. Algoe, S. Find, remind, and bind: the functions of gratitude in everyday relationships.
Compass 6, — Bao, F. Baumeister, R. Psychology as the science of self-reports and finger movements: whatever happened to actual behavior? Beaten, D. Guidelines for the process of cross cultural adaptation of self report mesures. Spine 25, — Bolier, L. Positive psychology interventions: a meta-analysis of randomized controlled studies. BMC Public Health Bradley, M. Gainesville: University of Florida. Google Scholar. Calvo, R.
Affect detection: an interdisciplinary review of models, methods, and their applications. IEEE Trans. Well Being Chancellor, J. Everyday prosociality in the workplace: the reinforcing benefits of giving, getting, and glimpsing. Emotion 18, — Chanel, G.
A Language to Map Consciousness
Short-term emotion assessment in a recall paradigm. Chen, L.
Building a profile of subjective well-being for social media users. PLoS One e Coan, J. Frontal EEG asymmetry as a moderator and mediator of emotion.