Ecological Momentary Assessment (EMA) in Psychology

What is the likelihood that someone will remember how they felt the day before? There is research that suggests not much. People tend to remember the worst events and emotions of the last week, forget other minor events completely, and colour their memory of the events based on how they feel at the time of recall (Shiffman et al., 2008). This is a basic issue in psychology, because the most common approach to gathering data has been to ask people to reflect on the past and describe what they experienced, a method called retrospective self-report.

The answer to this problem is a new method called “Ecological Momentary Assessment” (EMA). Ecological Momentary Assessment provides real-time data collection from individuals in their natural setting at multiple times throughout the day, as it occurs rather than as it is recalled. In this article, we will learn what EMA is, why it is important and what it reveals about the human mind in day-to-day life.

Ecological Momentary Assessment (EMA): What is it?

Ecological Momentary Assessment is a research and clinical procedure that entails sampling a person’s experiences, thoughts, emotions, behaviours, and context at multiple points in their daily life in their natural environment. The term is derived from a specific method developed by Paul Stone and Saul Shiffman and their colleagues in 1994, which aimed to reduce the need for retrospection, to understand the influences of the situation on experience, and to analyse the processes of psychological states across time (Stone & Shiffman, 1994). Usually, Ecological Momentary Assessment requires the person to wear a device, such as a dedicated electronic diary or pager (or most often, a smartphone), and to fill out brief questionnaires at various times during the day. These prompts can be given in some ways.

• Signal-contingent EMA prompts the person throughout the day (for example, six to ten times) at random or fixed times, without regard to activity.
• In Event-Contingent EMA, participants answer an assessment at specific events, like when a craving occurs, when they have an argument, or when they feel extremely stressed.
• Interval-contingent EMA elicits responses at fixed intervals over time that are informed by whether a user has performed an interval-contingent response or not, such as every two hours (Csikszentmihalyi & Larson, 1987; Shiffman et al., 2008).

The questions themselves are short, typically taking 2-5 minutes to answer and are designed for the moment, describing how one feels at the moment, where they are, who they are with, what they are doing, and what is the level of pain/craving/anxiety (on a scale of 1-10), etc. As these responses are made at the time of the event, instead of recalling it from memory, they are not distorted by the process of retrospective recall.

They provide a more fine-grained and detailed description of psychological life than other methods commonly used (Bolger et al., 2003). Experience Sampling Method (ESM) (Csikszentmihalyi & Larson, 1987) was a term used in the 1970’s and 1980’s by the psychologist Mihaly Csikszentmihalyi to describe his groundbreaking study of flow and the experience of normal everyday emotions. In current literature, the two terms are commonly mixed, and they refer to the evaluation of a momentary state and the sampling of an experiential episode, respectively. Both have as a common denominator real-time, repeated, naturalistic data collection.

Read More: The Dark Side of Psychological Research: 5 Experiments That Changed Research Ethics

Why are people better off remembering instead of using EMA?

Retrospective questionnaires, cross-sectional surveys and clinical interviews are the three main measures of self-report used in traditional psychological research, describing past experience, general experience and symptoms over a specified time, respectively. Both are valid, but both have one major limitation: they require people to re-create psychological experiences instead of reporting them as they actually are. Memory is reconstructive- when a person remembers something from the past, they reconstruct it in part using present knowledge, mood and expectations (Shiffman et al., 2008). The Ecological Momentary Assessment does not have this issue in a few particular ways:

• It greatly diminishes the possibility of recall bias, which is the distortion in memory that occurs when only certain experiences are remembered while others are forgotten. The Ecological Momentary Assessment data reflects what has actually happened, and not what is later remembered about it, since assessments are made at the time of experience.
• Ecological Momentary Assessment measures the within-person variation over time (i.e., day-to-day, hour-to-hour) instead of reducing that variation into one score. Clinically significant: Two individuals could score the same average anxiety on a weekly questionnaire, but see very different changes in anxiety from day to day, one person having moderate anxiety levels all week, while the other had high levels one day and lower the next (Trull & Ebner-Priemer, 2013).
• Ecological Momentary Assessment records the context in which psychological states are shaped- the physical environment, the social situation, the event that has just happened, and so on- in real time. A traditional questionnaire can provide a researcher with information that a person tends to be anxious. EMA can show that anxiety especially increases in the 30 minutes before work meetings, when the person is alone at home in the evening, or right after using social media. This specificity of context shifts the nature of the interventions that can be designed in response to it (Myin-Germeys et al., 2018).
• An advantage is ecological validity, or the extent to which research results correspond with experience in the real world, as opposed to the manipulated conditions of a laboratory experiment. The results from the Ecological Momentary Assessment are directly applicable to the behaviour that it is intended to describe, since it takes place in a person’s natural setting during their usual activities. In contrast to laboratory experiments, which allow tight control of variables and may be conducted under clean conditions, their results do not necessarily reflect those of experiments conducted under the messier conditions of the real world (Bolger et al., 2003).

Read More: Ethics of Psychological Research

Applications of EMA in Psychology

There are many areas of psychological functioning where EMA has been used and, in each area, in some way it has uncovered patterns that were not seen at all or inaccurately described using traditional methods.

1. Mental health and emotion regulation

EMA has revolutionised our understanding of the expression of depression, anxiety, bipolar disorder, and borderline personality disorder in everyday life. In borderline personality disorder (BPD), which involves strong, fleeting emotions and unstable relationships, EMA studies reveal that emotional instability is also present across the day, independent of external events, and that the rate of emotional fluctuations over the course of the day (affect lability) is a better predictor of clinical outcome than the average level of emotions (Trull & Ebner-Priemer, 2013). EMA has also been used in depression research to detect that depressed people report fewer positive emotions in social contexts (when they are with others) than non-depressed people, although their reports of negative emotions are similar.

2. Stress and daily life

EMA studies have cast doubt on which events of life are most likely to cause psychological distress. Daily hassles, which are the small, everyday frustrations and irritations of daily life (traffic, arguments, deadlines, minor failures, etc.), are better assessed using EMA and be better predictors of health and well-being than major stressor measures. It seems that the repeated, small stressors, recorded in real time, have more impact than clinicians were aware of (Almeida, 2005).

3. Addiction and craving

EMA was developed in the context of addiction research, and it is very useful in this area, as Shiffman’s early research on EMA was on smoking cessation. The results of EMA studies indicate that craving varies significantly from moment to moment throughout the day, is strongly influenced by situational factors and social environment, and that the link between craving intensity and actual substance use is quite small compared with what retrospective reports indicate. As described in follow-up interviews, people reported that they used when they experienced strong cravings, whereas EMA data revealed that many times they experienced strong cravings they did not use, and some of their use episodes were preceded by relatively mild craving, highlighting critical implications for addiction treatment design (Shiffman et al., 2008).

4. Research in behavioural and positive psychology

Csikszentmihalyi’s seminal ESM work is an example of such research, which examined the experience of flow, or being in the zone of deep engagement in an activity. EMA studies on well-being have repeatedly shown that individuals overestimate the happiness they experience while passively consuming leisure activities (or watching TV, or scrolling social media) and underestimate the happiness they experience when actively engaging in activities that they find challenging and rewarding (or exercising, or socializing, or working on a meaningful project)- findings relevant to understanding the distinction between hedonic and eudaimonic well-being in positive psychology (Csikszentmihalyi & Larson, 1987; Killingsworth & Gilbert, 2010).

Limitations, Challenges and the Future of EMA

However, there are certain limitations to EMA. The most often quoted limitation is participant burden: the number of people who have to fill out several short questionnaires a day over a long period often tires them out and increases the risk of skewed results. Another issue that is important in EMA research is compliance rates, or the percentage of prompts that are completed when asked: If participants systematically ignore some types of prompts (like those that occur at times when they are feeling poorly or at awkward times), the data may be biased in ways that are hard to identify.

Some studies estimate about an 80 per cent average compliance rate, although there is a great deal of variation across studies dependent on the study design, study population, and duration of the data collection (Palmier-Claus et al., 2011). Another concern is reactivity: the potential for repeatedly measuring the same psychological state to alter the state being assessed. Repeatedly asking a person to rate their anxiety more than once per day can help someone become more aware of their anxiety than they normally would be, or it can be stressful to be asked to rate anxiety over and over again.

Studies of reactivity in EMA have yielded conflicting results: some have shown evidence of short-term reactivity that decreases after the initial few days; others have not been able to demonstrate a consistent reactivity effect (Trull & Ebner-Priemer, 2013). The next one is data complexity. Time series data- repeated measures made over time from the same individual at many time points- are generated in large numbers in EMA studies and require careful and sophisticated statistical analyses to be appropriately analysed.

These typically include multilevel modelling, time-series analysis and network analysis, but require statistical skills beyond the reach of most clinical psychology research contexts (Bolger et al., 2003).

Read More: Research Aptitude: Understanding Psychological Data and Designing Experiments

Emerging Technologies and Future Directions

The potential for EMA, however, is very promising. As modern smartphones and other wearable devices and passive sensing technologies become more common, EMA is becoming more common. Ambulatory Assessment: multi-modal continuous sensing of psychological and physiological states in natural environments without requiring active participation from the person being assessed.

Passive sensing data can be used to augment the active self-report data, which decreases the burden on participants and increases the amount of contextual information that can be provided (Myin-Germeys et al., 2018). One of the most promising future directions of Ecological Momentary Assessment is Just-in-Time Adaptive Interventions (JITAIs), which are among the most interesting clinical applications of EMA. JITAIs are therapeutic interventions that are delivered via smartphones, at just the right time, as determined in real time from EMA data or passive sensing.

For instance, an app could find that a user is at heightened risk of a craving episode, based on self-reported anxiety level plus passively measured elevation of heart rate, and provide a personalised mindfulness exercise at that moment. Psychology’s future may be defined as effective studies of the mind in everyday life, but also as interventions in the mind; adaptive, tailored and real-time interventions (Nahum-Shani et al., 2018).

Read More: Ecopsychological Interventions in Mental Health Care: How Nature Heals the Mind

Conclusion

The use of ecological moments to assess what is happening to a person is one of the greatest methodological advances in psychology over the last 30 years. It has shifted the focus of data collection from the clinic to the lived moments of people’s daily lives and has remedied some basic misunderstandings about psychology’s knowledge of phenomena such as craving, stress, mood, well-being, and the texture of psychological experience throughout the day.

It has proven that the mind in the lab and the mind on a Tuesday afternoon can be different. The challenges are very real. Participant burden, reactivity, data complexity, and privacy issues with passive sensing all need continued tackling. But the trend is apparent. The distance between psychological research and psychological reality is shrinking. Digital technologies are becoming more complex and more integrated into everyday life. EMA is the key piece in that convergence.

References +

• Almeida, D. M. (2005). Resilience and vulnerability to daily stressors assessed via diary methods. Current Directions in Psychological Science, 14(2), 64–68. https://doi.org/10.1111/j.0963-7214.2005.00336.x 
• Bolger, N., Davis, A., & Rafaeli, E. (2003). Diary methods: Capturing life as it is lived. Annual Review of Psychology, 54(1), 579–616. https://doi.org/10.1146/annurev.psych.54.101601.145030 
• Csikszentmihalyi, M., & Larson, R. (1987). Validity and reliability of the Experience Sampling Method. Journal of Nervous and Mental Disease, 175(9), 526–536. https://doi.org/10.1097/00005053-198709000-00004
• Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is an unhappy mind. Science, 330(6006), 932. https://doi.org/10.1126/science.1192439 
• Myin-Germeys, I., Kasanova, Z., Vaessen, T., Vachon, H., Kirtley, O., Viechtbauer, W., & Reininghaus, U. (2018). Experience sampling methodology in mental health research: New insights and technical developments. World Psychiatry, 17(2), 123–132. https://doi.org/10.1002/wps.20513 
• Nahum-Shani, I., Smith, S. N., Spring, B. J., Collins, L. M., Witkiewitz, K., Tewari, A., & Murphy, S. A. (2018). Just-in-Time Adaptive Interventions (JITAIs) in mobile health: Key components and design principles for ongoing health behaviour support. Annals of Behavioural Medicine, 52(6), 446–462. https://doi.org/10.1007/s12160-016-9830-8 
• Palmier-Claus, J. E., Myin-Germeys, I., Barkus, E., Bentley, L., Udachina, A., Delespaul, P. A., Lewis, S. W., & Dunn, G. (2011). Experience sampling research in individuals with mental illness: Reflections and guidance. Acta Psychiatrica Scandinavica, 123(1), 12–20. https://doi.org/10.1111/j.1600-0447.2010.01596.x 
• Shiffman, S., Stone, A. A., & Hufford, M. R. (2008). Ecological momentary assessment. Annual Review of Clinical Psychology, 4(1), 1–32. https://doi.org/10.1146/annurev.clinpsy.3.022806.091415 
• Stone, A. A., & Shiffman, S. (1994). Ecological momentary assessment (EMA) in behavioural medicine. Annals of Behavioural Medicine, 16(3), 199–202. https://doi.org/10.1093/abm/16.3.199 
• Trull, T. J., & Ebner-Priemer, U. (2013). Ambulatory assessment. Annual Review of Clinical Psychology, 9(1), 151–176. https://doi.org/10.1146/annurev-clinpsy-050212-185510