What Is Ecological Momentary Assessment (EMA)? A Researcher's Introduction

Ask someone how anxious they have felt over the past two weeks and you get a single value. But what does that value actually represent? Their average level of anxiety over the past two weeks? How frequently they felt anxious over that time period? Their peak intensity of anxiety experienced over the last two weeks? In reality, how people retrospect - how they think about their past experiences - reflects a multidimensional cognitive process that reflects a combination of how they feel today, which experiences stand out in memory, and how they distill two weeks of experiences into a single value. Ecological momentary assessment (EMA) takes a different approach. Instead of asking people to report on past experiences, EMA asks them, in real time, to report on what is happening right now.

A definition

Shiffman, Stone, and Hufford (2008) define EMA as "repeated sampling of subjects' current behaviors and experiences in real time, in subjects' natural environments." Each part of that definition matters:

• Repeated. Each person completes many brief assessments, often several a day for days or weeks, rather than one questionnaire.
• Current. Questions ask about this moment or the very recent past (e.g., "How stressed do you feel right now?"), not last month.
• Real time. Assessments happen while the experience is happening, triggered by a schedule or by the event itself.
• Natural environments. Participants answer wherever they are, at work, at home, or on the bus, rather than in a lab or clinic.

In the past, participants would be sent home with dozens of pen-and-paper questionnaires and instructed to complete them at pre-defined moments. In a modern EMA study, smartphones enable digital notifications that prompt participants to answer a short questionnaire at fixed or random times throughout the day. Participants can also self-initiate a report themselves when a target event occurs. On Inclivio, participants can receive notifications via the Inclivio mobile application, SMS, WhatsApp, or email.

Where EMA came from

The term EMA comes from behavioral medicine. Arthur Stone and Saul Shiffman introduced it in a 1994 article in Annals of Behavioral Medicine (Stone & Shiffman, 1994). The idea of sampling everyday experience is older. Psychologists had been doing it under the name experience sampling method (ESM), which Csikszentmihalyi and Larson (1987) described as an instrument for capturing "variations in self-reports of mental processes" in everyday life, with evidence for its reliability and validity.

Today the labels overlap heavily. Trull and Ebner-Priemer (2013) note that some researchers use ambulatory assessment, ecological momentary assessment, experience sampling, and e-diary methods "interchangeably." They treat ambulatory assessment as the broadest umbrella, one that also covers physiological and behavioral monitoring.

Why measure in the moment?

Shiffman et al. (2008) give EMA three aims: to "minimize recall bias, maximize ecological validity, and allow study of microprocesses that influence behavior in real-world contexts."

1. Less recall bias

Retrospective reports are reconstructions, not recordings. Stone et al. (1998) demonstrated that people's recollections of how they coped with stress, reported within 48 hours, can meaningfully diverge from what they reported in the moment. Since then, researchers have demonstrated bias in retrospective reports among hospitalized patients with depression (Ben-Zeev & Young, 2010), people with schizophrenia, for emotions more than for psychotic symptoms (Ben-Zeev et al., 2012), psychiatric outpatients with borderline personality disorder or depression (Solhan et al., 2009), people with chronic pain (Stone et al., 2005), and smokers entering a cessation program (Shiffman, 2009), just to name a few.

2. Ecological validity

Lab findings do not always carry over to daily life. EMA collects data "in real-world environments, increasing the ecological validity of findings" (Trull & Ebner-Priemer, 2013). Myin-Germeys et al. (2018) ground the method in ecological psychology, which holds that "behaviour is radically situated." To understand a symptom, you study it in the settings where it happens.

3. Within-person processes over time

Repeated measurement lets you ask how states change and what comes before what within the same person. A classic example is Shiffman et al.'s (1996) study of smokers trying to quit. Participants recorded, in real time, their first lapse, episodes of increased urges to smoke, and assessments completed at randomly scheduled times. Negative affect was highest during lapses, lower in the context of increased urges, and lowest at random assessments. Lapses were more likely when smoking was permitted, when cigarettes were easily available, and when other smokers were present. Only cognitive coping strategies, not behavioral ones, were effective. Comparing a person's lapse with that same person's non-lapse experiences (assessed randomly) is impossible with a single retrospective report. Some conditions also demand repeated measurement, because they are "defined and characterized by instability, a feature that can be measured only by multiple and frequent assessments" (Trull & Ebner-Priemer, 2013).

How EMA studies choose their moments

EMA designs differ mainly in what triggers an assessment:

• Signal-contingent: the device prompts the participant, often at random times within set windows.
• Interval-contingent: assessments happen at fixed times, such as every evening.
• Event-contingent: the participant starts a report when a predefined event occurs. Signal- and interval-contingent EMA schedules can also be triggered on the occurrence of certain events.

Signal-contingent prompting is the most common design. In Wrzus and Neubauer's (2023) meta-analysis of 477 EMA articles, 57.9% of studies used signal-contingent schedules, 18.1% interval-contingent, 5.0% event-contingent, and 15.7% a mix.

What a typical study looks like

The same meta-analysis gives a picture of current practice. The median study scheduled six assessments a day for seven days, and average compliance was 79% (Wrzus & Neubauer, 2023). Compliance was higher in studies that paid participants. The total number of assessments did not predict compliance or dropout, and studies that asked for more assessments per day tended to run for fewer days. Protocols in psychiatry are often denser: Myin-Germeys et al. (2018) note that many studies have used ten assessments a day over six consecutive days, and they advise that a questionnaire should take no longer than two minutes to complete.

The trade-offs

EMA asks a lot of participants, but not every design choice adds the same burden. Eisele et al. (2022) gave students a 30- or 60-item questionnaire three, six, or nine times a day for 14 days: longer questionnaires increased burden and reduced data quantity and quality, but more frequent prompts did not. Reactivity is the other open question: does answering questions about your mood change your mood? Summarizing the literature, Barta et al. (2011, as cited in Trull & Ebner-Priemer, 2013) concluded that EMA self-reports are generally "not likely to be affected greatly by reactivity," though "some studies do show such an effect." Myin-Germeys et al. (2018) call measurement reactivity "a key challenge for ESM research, yet [it] remains an under-researched phenomenon." In Wrzus and Neubauer's (2023) meta-analysis, 95.8% of samples were not tested for it. And high compliance is not guaranteed. It depends on burden, study length, and how easy the tools are to use (Trull & Ebner-Priemer, 2013).

What this means for running an EMA study

The method's advantages only hold if the software keeps up with the design. In practice, that means support for random, fixed, and participant-initiated assessments within each person's waking hours. It means time stamps on every prompt and response, response windows that close, and compliance you can watch while the study is still running, so you can follow up before the data are lost. Platforms like Inclivio are built around these requirements.

EMA does not replace the retrospective questionnaire. It answers different questions. When your question is about how experience unfolds in daily life, measuring it as it happens is the most direct way to find out.

References

Ben-Zeev, D., McHugo, G. J., Xie, H., Dobbins, K., & Young, M. A. (2012). Comparing retrospective reports to real-time/real-place mobile assessments in individuals with schizophrenia and a nonclinical comparison group. Schizophrenia Bulletin, 38(3), 396–404. https://doi.org/10.1093/schbul/sbr171

Ben-Zeev, D., & Young, M. A. (2010). Accuracy of hospitalized depressed patients' and healthy controls' retrospective symptom reports: An experience sampling study. The Journal of Nervous and Mental Disease, 198(4), 280–285. https://doi.org/10.1097/NMD.0b013e3181d6141f

Csikszentmihalyi, M., & Larson, R. (1987). Validity and reliability of the experience-sampling method. The Journal of Nervous and Mental Disease, 175(9), 526–536. https://doi.org/10.1097/00005053-198709000-00004

Eisele, G., Vachon, H., Lafit, G., Kuppens, P., Houben, M., Myin-Germeys, I., & Viechtbauer, W. (2022). The effects of sampling frequency and questionnaire length on perceived burden, compliance, and careless responding in experience sampling data in a student population. Assessment, 29(2), 136–151. https://doi.org/10.1177/1073191120957102

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

Shiffman, S. (2009). How many cigarettes did you smoke? Assessing cigarette consumption by global report, Time-Line Follow-Back, and ecological momentary assessment. Health Psychology, 28(5), 519–526. https://doi.org/10.1037/a0015197

Shiffman, S., Paty, J. A., Gnys, M., Kassel, J. A., & Hickcox, M. (1996). First lapses to smoking: Within-subjects analysis of real-time reports. Journal of Consulting and Clinical Psychology, 64(2), 366–379. https://doi.org/10.1037/0022-006X.64.2.366

Shiffman, S., Stone, A. A., & Hufford, M. R. (2008). Ecological momentary assessment. Annual Review of Clinical Psychology, 4, 1–32. https://doi.org/10.1146/annurev.clinpsy.3.022806.091415

Solhan, M. B., Trull, T. J., Jahng, S., & Wood, P. K. (2009). Clinical assessment of affective instability: Comparing EMA indices, questionnaire reports, and retrospective recall. Psychological Assessment, 21(3), 425–436. https://doi.org/10.1037/a0016869

Stone, A. A., Schwartz, J. E., Broderick, J. E., & Shiffman, S. S. (2005). Variability of momentary pain predicts recall of weekly pain: A consequence of the peak (or salience) memory heuristic. Personality and Social Psychology Bulletin, 31(10), 1340–1346. https://doi.org/10.1177/0146167205275615

Stone, A. A., Schwartz, J. E., Neale, J. M., Shiffman, S., Marco, C. A., Hickcox, M., Paty, J., Porter, L. S., & Cruise, L. J. (1998). A comparison of coping assessed by ecological momentary assessment and retrospective recall. Journal of Personality and Social Psychology, 74(6), 1670–1680. https://doi.org/10.1037/0022-3514.74.6.1670

Stone, A. A., & Shiffman, S. (1994). Ecological momentary assessment (EMA) in behavioral medicine. Annals of Behavioral 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, 151–176. https://doi.org/10.1146/annurev-clinpsy-050212-185510

Wrzus, C., & Neubauer, A. B. (2023). Ecological momentary assessment: A meta-analysis on designs, samples, and compliance across research fields. Assessment, 30(3), 825–846. https://doi.org/10.1177/10731911211067538