How ICT Use and Time Fragmentation Shape Well-Being

Abstract

Digital technology is embedded in almost every part of daily life. Most adults carry a connected device from morning to night, and activities that used to be tied to a fixed time and place can now happen anywhere and at any moment. Research links heavy ICT use to lower well-being, but the measured effects are small and depend on context. One possible pathway has received little direct attention. ICT changes the structure of the day itself. Connectivity allows activities to be performed in shorter pieces and switched at any moment, so days become fragmented into more and shorter episodes. Three links matter here: ICT use and well-being, fragmentation and well-being, and ICT use and fragmentation. They form a triangle, and each side has been studied on its own. The understudied link is fragmentation as the channel that ties ICT use to well-being.

The study uses episode level diary data from the 2021 American Time Use Survey and its Well-Being Module (6,902 respondents). Two ICT predictors are constructed from the diary. ICT group breadth counts how many domains of digital life a respondent touched on the diary day, and ICT minutes captures total daily ICT time as an ordinal scale. Three fragmentation indicators are computed from each respondent's full activity sequence. These are the number of activity episodes, turbulence, and entropy. The outcome is life evaluation, measured with the Cantril ladder. The relationships are estimated jointly in a path model in IBM AMOS, which separates the direct association between ICT use and life evaluation from the indirect association that runs through fragmentation. The same model is estimated in six subgroups defined by gender and three age bands, and in a combined model, followed by three sensitivity tests.

The results show that ICT exposure is not a single quantity. The two ICT predictors pull in opposite directions. Group breadth predicts higher fragmentation on all three indicators in every subgroup. ICT minutes predicts lower fragmentation in every subgroup, which is best read as a measurement property of the diary. Long blocks of passive media use collapse into single episodes and make the day look more consolidated than it was. Fragmentation also splits internally. More activity episodes predict lower life evaluation, while higher turbulence and entropy predict higher life evaluation, suggesting that variety in the day can support well-being while frequent switching costs it. The direct paths from ICT use to life evaluation are mostly null, with small negative effects only for ICT minutes among young women and middle aged men. The indirect channels through fragmentation carry opposite signs for the two predictors, and total effects sit close to zero in every subgroup. The fragmentation channel is concentrated in the middle and older age groups, while the direct channel appears only in the younger half of the sample.

These findings challenge the assumption behind current screen time policy that ICT exposure is one quantity where more is worse. Reporting screen time as a single number of minutes is like a nutrition label that only reports calories. The findings suggest that what matters is which face of ICT use and which face of fragmentation dominate, and for whom.

The study has limitations. The data cover one diary day per respondent in a single country during a pandemic year. Respondents round durations to five minute blocks, and short ICT episodes inside long passive activities disappear from the diary. Follow-up research should combine time use diaries with device level usage logs, so the device supplies the measurement and the respondent supplies the meaning.