Experience Sampling in Media Research
- Introduced by
- Mihaly Csikszentmihalyi & Reed Larson
- Year
- 1987
- Type
- Momentary self-report of media use and experience in real time
- Family
- Intensive longitudinal self-report methods
What does it do?
The experience-sampling method (ESM), also called ecological momentary assessment, prompts participants at sampled moments throughout daily life to report what they are doing, using, and feeling right now.
Applied to media research, it captures media use and its momentary correlates — mood, context, motivation — in real time and in situ, minimizing recall bias and revealing how media and experience interrelate moment to moment.
The intuition behind the method
Memory smooths and distorts: people forget the dull minutes scrolling a feed and remember the highlights, so retrospective surveys misstate both how much media they use and how it feels. ESM beats this by sampling the present. A signal pings the participant's phone at unpredictable times and asks, in that instant, what they are using and how they feel. Repeated dozens of times across a week, these momentary snapshots build an accurate, fine-grained portrait of media in everyday life — and because the same person is sampled repeatedly, ESM can show, for instance, whether a person's mood dips after each bout of social-media use, a within-person link no one-shot survey can detect.
History
Mihaly Csikszentmihalyi and Reed Larson developed the experience-sampling method in the 1970s and 1980s, validating it in their 1987 paper, originally using pagers and booklets to study the flow of daily experience. Parallel work in health psychology produced ecological momentary assessment. Smartphones transformed the method, enabling precise prompting, context capture, and passive sensing. In communication research ESM has become central to studying mobile and social-media use and its momentary links to mood and well-being, often combined with diary, log, and physiological data.
When to use it
Use experience sampling when your question is about momentary, in-the-moment media experience and its dynamic links to states like mood, attention, or stress — for instance, how social-media use relates to well-being moment to moment, or how media fits into the flow of daily life. It excels at within-person process questions and at reducing recall bias. It assumes participants can and will respond promptly to prompts without those prompts unduly altering behavior, and that momentary self-report is valid. It is less appropriate for rare events that random sampling will miss, when burden threatens compliance, or when the construct is reflective and retrospective rather than momentary — and like all self-report it benefits from pairing with objective logs or passive sensing.
Limitations and considerations
- High burden and frequent interruptions can reduce compliance and bias who completes the study.
- Random prompting can miss rare or brief media events of interest.
- Repeated self-monitoring may reactively change behavior or mood.
- Intensive nested data require multilevel and time-series modeling to analyze properly.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Captures media experience in the moment, minimizing retrospective recall bias.
- High ecological validity: data come from real life as it is lived, not the lab.
- Repeated within-person sampling enables analysis of momentary and lagged dynamics.
- Smartphone delivery allows precise timing, context, and even passive-sensing integration.
Applications
- Studying momentary links between social-media use and mood or well-being.
- Capturing how media use is distributed across daily life and contexts.
- Examining attention, motivation, and gratifications as they occur during media use.
- Investigating second-screen and mobile media behavior in real time.
- Combining momentary self-reports with passive smartphone sensing of actual app use.
How it works
How it works
The researcher specifies the sampling protocol — how many prompts per day, over how many days, and whether timing is random within blocks, fixed, or event-triggered — balancing temporal coverage against participant burden.
Prompts are delivered (historically by pager, now by smartphone app) at the sampled moments, asking a brief set of questions about current media use and state while compliance and response latency are monitored to ensure reports are momentary.
Each prompt captures the in-the-moment situation: what media is being used, where, with whom, and the participant's current mood, motivation, or other states, kept short to protect compliance.
Because each person yields many momentary observations, multilevel models separate within-person processes (how state changes with media use across moments) from between-person differences, supporting analyses of momentary and lagged associations.
Formula
y_{ti} = \beta_{0i} + \beta_{1i} x_{ti} + e_{ti}Common pitfalls
- Over-prompting to the point that fatigue and non-compliance degrade the sample.
- Treating delayed responses as momentary when participants answered long after the signal.
- Using ordinary regression on nested ESM data, ignoring within-person dependence.
- Inferring causation from momentary correlations without lagged designs or careful temporal modeling.
Frequently asked questions
How is experience sampling different from a diary study?
Experience sampling is essentially a signal-contingent diary focused on momentary, in-the-moment states: participants are prompted at sampled times and report what is happening right now. Diary methods are a broader category that also includes end-of-day (interval-contingent) and event-triggered reporting. The defining feature of ESM is the emphasis on the present moment, sampled repeatedly, to capture experience as it unfolds rather than as it is later reconstructed. In smartphone studies the two approaches often merge.
How many prompts per day should an ESM study use?
There is no universal number; it depends on how dynamic the phenomenon is and how much burden participants can sustain. Many media studies use roughly five to ten prompts per day over one to two weeks, balancing temporal coverage against the fatigue and non-compliance that excessive prompting causes. Too few prompts miss within-day dynamics; too many degrade compliance and data quality. Pilot testing and monitoring response rates help calibrate the protocol.
Can experience sampling establish causation?
Not on its own, but it can strengthen causal inference relative to cross-sectional surveys. Momentary associations are correlational, but ESM's repeated within-person measurement supports lagged analyses — for example, testing whether social-media use at one moment predicts mood at the next, within the same person — which helps establish temporal order and rules out stable between-person confounds. True causal claims still require experimental manipulation, but ESM's intensive longitudinal structure is a meaningful step beyond static designs.
Related methods
Sources
Csikszentmihalyi, M., & Larson, R. (1987). Validity and reliability of the experience-sampling method. The Journal of Nervous and Mental Disease, 175(9), 526–536.
Open DOI referenceCopied.Bolger, N., Davis, A., & Rafaeli, E. (2003). Diary methods: Capturing life as it is lived. Annual Review of Psychology, 54, 579–616.
Open DOI referenceCopied.