Accurate on average, wrong one person at a time
The most direct evidence comes from a survey of 338 university students whose self-estimated daily screen time was compared with objective screen time data [1]. Contrary to what the researchers expected, the two did not differ significantly at the group level — in other words, if you averaged everyone's guesses, the group landed close to the truth [1].
That group-level match is not the same as individual accuracy. The same study found that individual errors were substantial and followed a pattern rather than scattering randomly [1]. So a student asking "am I good at estimating my own usage?" cannot take comfort from the group average; the useful information is in which direction that particular person tends to err.
Which way the error goes depends on the app
The pattern in the errors was tied to platform. Instagram users tended to underestimate how much time they spent, while TikTok users typically overestimated [1]. The researchers concluded that misperceiving one's screen time is platform-specific rather than a universal blind spot, which is why they argue media literacy efforts should be tailored to how each platform is actually used [1].
This is a correlational, group-level observation from a single survey, so it describes a tendency among users of each app rather than a rule that predicts any one person. It also means the honest answer to "are students accurate?" is: it depends on where they are scrolling.
Heavy use may stretch time beyond the videos themselves
A separate line of work looked at time distortion — over- or underestimating how long something lasted — among college students. In a three-phase survey of 111 students, the frequency of short-form video use significantly predicted overestimating both daily short-form video time and the time spent on an unrelated survey task [2].
A second experiment in that paper assigned 62 students to 15 minutes of TikTok or 15 minutes of reading, then asked them to judge durations. The TikTok group significantly overestimated both the priming activity and a following academic task, and more frequent daily use predicted greater overestimation; the reading group was relatively accurate [2]. This suggests the distortion may not stay confined to the app, though the study reports an association and a short-term experimental effect rather than a proven mechanism.
What a student can actually do with this
The practical takeaway is that self-report is a weak measuring instrument for this behavior — reliable enough to describe a group, unreliable enough to mislead an individual [1]. Because the direction of error appears tied to the platform, a person's own guess is more useful as a signal of habit than as a measurement: an overestimate on TikTok and an underestimate on Instagram may both reflect heavy, automatic scrolling rather than careful counting [1].
The distortion also appears to be short-term and state-dependent rather than a fixed trait: in the experiment, a brief TikTok session shifted duration judgments for the next task, while a reading session did not [2]. That is consistent with the idea that the effect fades with the activity, but it has not been shown to persist. Notably, the control sample in the survey study showed no significant short-term changes in screen time and no platform-based differences over the measurement period [1], so the stability of these patterns over weeks and months remains an open question.
Sources used in this answer
Awareness of Time Spent Watching Short-Form Videos among University Students
Among 338 university students, average self-estimated daily short-form video time did not differ significantly from objectively measured screen time, but individual errors were substantial, with Instagram users tending to underestimate and TikTok users typically overestimating.
Time distortion for short-form video users
In 111 college students, short-form video use frequency predicted overestimating daily video time and an unrelated task's duration, and a 15-minute TikTok session (versus reading) led to overestimating both that activity and a subsequent academic task.
The Doomscrolling Dilemma: How Multilingual Short-Form Videos Reshape the Attention of University Students and the Decline of" Adda" in Bangladesh
Qualitative interviews link multilingual algorithmic short-form video feeds to extended scrolling, distorted time perception, and compulsive use among Bangladeshi university students, but the study does not measure time-estimation accuracy.
Exploring short-form videos binge-scrolling behaviors among college students
The study examines college students' short-form video consumption habits and binge-scrolling, touching on actual time spent versus estimation, but the available abstract does not report a specific accuracy result.
A motivation-based theoretical framework for understanding short-form video use and mental health among university students
The paper proposes a motivation-based theoretical framework in which avoidance motives are expected to produce stronger time distortion in short-form video use, framing relations for future studies to estimate rather than reporting measured accuracy.