The experience sampling method, also referred to as a daily diary method, or ecological momentary assessment, is an intensive longitudinal research methodology that involves asking participants to report on their thoughts, feelings, behaviors, and/or environment on multiple occasions over time. Participants report on their thoughts, feelings, behaviors, and/or environment in the moment or shortly thereafter. Participants can be given a journal with many identical pages. Each page can have a psychometric scale, open-ended questions, or anything else used to assess their condition in that place and time. ESM studies can also operate fully automatized on portable electronic devices or via the internet. The experience sampling method was developed by Larson and Csikszentmihalyi.
Overview
There are different ways to signal participants when to take notes in their journal or complete a questionnaire, like using preprogrammed stopwatches. An observer can have an identically programmed stopwatch, so the observer can record specific events as the participants are recording their feelings or other behaviors. It is best to avoid letting subjects know in advance when they will record their feelings, so they can't anticipate the event, and will just be "acting naturally" when they stop and take notes on their current condition. Conversely, some statistical techniques require roughly equidistant time intervals, which has the limitation that assessments can be anticipated. Validity in these studies comes from repetition, so you can look for patterns, like participants reporting greater happiness right after meals. These correlations can then be tested by other means for cause and effect, such as vector autoregression, since ESM just shows correlation. Some authors also use the term experience sampling to encompass passive data derived from sources such as smartphones, wearable sensors, the Internet of Things, email and social media that do not require explicit input from participants. These methods can be advantageous as they impose less demand on participants improving compliance and allowing data to be collected for much longer periods, are less likely to change the behaviour being studied and allow data to be sampled at much higher rates and with greater precision. Many research questions can benefit from both active and passive forms of experience sampling.
Software and related tools
The first mobile device application that could be used as a tool for Experience Sampling Method was the ESP Package. This had limited functionality in that it is designed for older iOS Palm devices and had limited scheduling capabilities. It no longer works on modern mobile devices. iHabit was the first smartphone mobile application designed for Experience Sampling. It was developed in 2011 and used in a study published by PLOS One in 2013. In 2015, it was superseded by the system, which was used in a study published by JAMA Pediatrics in 2016. This system has subsequently been used in numerous studies. The app is a free app available in iOS and Android versions and has since been used in more than 12 academic publications. It can be used for scheduled, random and on-demand surveys. Unlike many platforms, no server is required as data is saved on the device and emailed to the researcher or else retrieved by file sharing. Other early smartphone platforms for ESM include and , , , , and . The largest ESM study was achieved through 's , PSYT’s apps collect data through ESM as well as reporting the data back to users to enable real-time visualisation and tracking of variables. Several other commercial and open source systems are currently available to help researchers run ESM studies, including BeepMe, and Expimetrics. enables researchers to gather and integrate data from commercially available sensors and service providers to use them in ESM, including and . As of 2014, have developed the ability to trigger sampling forms from physiological data such as actigraphy and ECG. provide a platform for both active and passive experience sampling that allows the integration of some 400 data sources.
