Media-Use Diary Method
- Introduced by
- Diary-methods tradition (Bolger, Davis & Rafaeli)
- Year
- 2003
- Type
- Self-report data collection of media use in natural settings over time
- Family
- Intensive longitudinal self-report methods
What does it do?
The media-use diary method has participants record their media use repeatedly over days or weeks, close to when it happens, capturing everyday media behavior in its natural context with minimal retrospective bias.
It yields intensive longitudinal data that reveal how media use varies within individuals across time and situations, not just averaged across people.
The intuition behind the method
If you ask people at the end of a month how much news they watched, you get a rough, often flattering guess. Catch them in the moment — 'what are you using right now, and why?' — repeated day after day, and you get a far truer picture of media life as it is actually lived. Diaries trade the single, distant snapshot of a survey for a stream of close-to-real-time entries. Because each person contributes many entries, the data show not only who uses more media but how a given person's use ebbs and flows with mood, time of day, and circumstance — within-person dynamics invisible to one-shot surveys.
History
Diary methods have a long history in psychology and sociology for studying daily experience, and they were systematized for the social sciences in Bolger, Davis, and Rafaeli's 2003 Annual Review treatment distinguishing interval-, signal-, and event-contingent designs. They are closely allied with the experience-sampling method validated by Csikszentmihalyi and Larson. In communication research, diaries have long been used to measure media exposure (including in audience-ratings traditions) and, with smartphones, have become a flexible tool for capturing everyday and mobile media use in context.
When to use it
Use the media-use diary method when you need accurate, contextualized accounts of everyday media behavior and how it varies within people over time — for example, linking daily media use to mood, studying routines, or capturing fleeting use that surveys miss. It suits questions about dynamics, context, and process rather than one-time levels. It assumes participants will comply and report honestly and that frequent reporting does not unduly change the behavior. It is less appropriate when the behavior is rare or hard to notice, when participant burden would harm data quality, or when objective logs (server-side or device tracking) are feasible and preferable — diaries can be combined with such traces to offset self-report limits.
Limitations and considerations
- High participant burden can erode compliance and data quality over time.
- Self-reports remain vulnerable to honesty, awareness, and timing problems (backfilling entries).
- Repeated self-monitoring can reactively change the very behavior being studied.
- Intensive longitudinal data require multilevel modeling expertise to analyze correctly.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Captures media use close to when it occurs, sharply reducing retrospective recall bias.
- Preserves context, letting use be linked to mood, situation, and motives as they actually co-occur.
- Separates within-person dynamics from between-person differences through repeated measures.
- Studies media life in natural settings, improving ecological validity over lab or one-shot designs.
Applications
- Measuring everyday exposure to news, television, and social media in natural settings.
- Linking daily media use to mood, stress, well-being, or social interaction.
- Studying media routines and how use varies by time of day and context.
- Capturing mobile and second-screen media behaviors as they occur.
- Complementing or validating server-side and device-tracking logs of media use.
How it works
How it works
The researcher selects a diary design: interval-contingent (entries at fixed times, e.g., each evening), signal-contingent (prompted at random or scheduled moments), or event-contingent (an entry whenever a defined media event occurs), matching the design to the phenomenon's timescale.
A short, low-burden instrument is built — paper, web, or mobile app — asking about media use and relevant covariates (mood, context, motives), kept brief to sustain compliance over the study period.
Participants record entries over the study window, with attention to compliance and timing (mobile timestamps and reminders help verify entries are made near the event rather than backfilled), echoing experience-sampling validity concerns.
Because entries are nested within persons, data are analyzed with multilevel (mixed-effects) models that separate within-person variation (how use changes across occasions) from between-person differences, with random intercepts and slopes capturing individual variability.
Formula
y_{ti} = \beta_{0i} + \beta_{1i} x_{ti} + e_{ti},\quad \beta_{0i} = \gamma_{00} + u_{0i}Common pitfalls
- Designing an instrument so long that fatigue degrades compliance and responses.
- Allowing retrospective backfilling, undermining the method's core advantage of immediacy.
- Analyzing nested diary data with ordinary regression, ignoring within-person dependence.
- Generalizing from highly compliant volunteers whose media habits may be atypical.
Frequently asked questions
What are the main types of diary designs?
Interval-contingent diaries ask for entries at fixed intervals, such as each evening, suiting reflective summaries. Signal-contingent diaries prompt entries at random or scheduled signals, capturing momentary states (this overlaps with experience sampling). Event-contingent diaries ask for an entry whenever a defined event occurs, such as each time the participant uses social media, capturing rare or specific behaviors. The choice depends on whether the phenomenon is continuous, momentary, or event-based, and on how it is distributed in time.
Why are diary data analyzed with multilevel models?
Diary studies produce many observations per person, so entries from the same individual are correlated rather than independent. Ordinary regression that ignores this clustering underestimates standard errors and conflates within-person and between-person effects. Multilevel (mixed-effects) models nest occasions within persons, with random intercepts and slopes, allowing researchers to separate how a person's media use varies across occasions from how people differ from one another — the central questions diaries are designed to answer.
How is the diary method different from experience sampling?
They overlap heavily and are sometimes used interchangeably. Experience sampling is essentially a signal-contingent diary that prompts participants at random or scheduled moments to report their current state, emphasizing momentary, in-the-moment experience. The diary method is a broader umbrella that also includes interval-contingent (e.g., end-of-day) and event-contingent reporting. In practice, smartphone-based studies blur the line, combining scheduled prompts with event-triggered and end-of-day entries.
Related methods
Sources
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.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.