Smartphone-based study reminders can be a double-edged sword

The study addresses whether smartphone-delivered reminders can effectively promote regular, self-regulated studying—specifically distributed practice—among lower secondary school students. While self-regulated learning is essential for adapting to changing demands, students, especially younger ones, often fail to apply effective techniques such as distributed practice. Mobile technologies can support self-study, and reminders are a common feature in behavior-change interventions. However, reminders may carry drawbacks (e.g., annoyance, guilt, hindered habit formation) and could lead to overreliance, undermining autonomous study routines. The authors investigated if reminders increase day-level studying and improve test performance, preregistering four hypotheses: H1) Within the Reminder Group, students are more likely to study on days with a reminder than on days without; H2) The Reminder Group studies more frequently than the Control Group; H3) The Reminder Group outperforms the Control Group on vocabulary tests; H4) More frequent studying predicts better test performance.

Prior research highlights that distributed practice robustly benefits long-term retention, yet students underuse it, often favoring cramming or evaluating strategies by short-term outcomes. Digital and mobile interventions can scaffold self-regulated learning, but many target behaviors during study sessions rather than prompting the initiation and scheduling of study needed for spacing. Nudges and reminders in mHealth have increased behaviors requiring regular repetition (e.g., medication adherence, water intake, physical activity), with effectiveness varying across individuals and contexts and influenced by timing, frequency, and content. Potential negative effects of reminders include annoyance, guilt, and interference with habit formation; reminders may support repetition but hinder the development of automaticity, risking overreliance whereby behavior occurs mainly when prompted. Educational reminders are less studied; some benefits have been found for students and through stakeholders like parents/teachers. Emerging work suggests interventions that aid habit formation (e.g., implementation intentions that link actions to stable cues) may sustain regular behavior better than simple reminders. This context motivated examining both benefits and hidden costs of study reminders for distributed practice.

Open practices: The study was preregistered (OSF: https://doi.org/10.17605/OSF.IO/KA2NV). Anonymized data and analysis code are available on OSF. Participants: German-speaking fifth graders (Fall 2021). 96 enrolled; after preregistered/explained exclusions, N=85 (Control n=42; Reminder n=43). Mean age 10.67 years (SD=0.36; range 10.08–12.01); 50.59% female. Most attended German schools (one in Austria); 71.76% attended Gymnasium. Most (85.88%) were unfamiliar with the vocabulary app. Incentives: one-year premium account for the vocabulary app (29.99 €) and vouchers tied to study app usage (study app engagement rewarded; vocabulary app use not incentivized). Ethics: DIPF ethics approval (DIPF_EK_2021_33). Design: 55-day study with three phases: 1) Introductory day; 2) 36-day intervention phase; 3) 18-day follow-up. During the intervention, students used both a study app and a vocabulary learning app (cabuu); during follow-up, only cabuu was used. All students watched a video explaining distributed practice on the introductory day. Random assignment to Control (no reminders beyond video) or Reminder Group (in-app reminders on ~16 of the 36 intervention days). Reminders alternated in 2–3 day blocks; order counterbalanced to reduce predictability. Study-task structure: Students prepared six vocabulary lists (~40 word pairs each) drawn from their textbooks; missing items were supplemented from grade-six materials. Each list was to be studied for eight days, followed by a test on day 9, forming a study-test cycle. Over the 36-day intervention, four cycles were intended; the follow-up included two cycles. Adherence varied across students. Interventions and apps: Daily at 6:30 am all students received a notification to open the study app and complete an ambulatory assessment (AA). For the Reminder Group, on half of the study days a reminder screen appeared immediately after the AA (screenshot from the video + brief text: study every day for best outcomes). This was not a push notification; students only saw it if they opened the study app that morning. Thus, both groups had the same notification load; the reminder was an extra in-app screen. The study app also provided a gamified reward system (diamonds). Cabuu supported multiple study methods, adaptive scheduling, and built-in vocabulary tests. Measures and outcomes: Primary daily outcome was whether a student studied in cabuu on a given day (Learning Event: 1 if interacted with ≥1 vocabulary word; 0 otherwise). For H1, within-person predictor: presence of reminder (1/0) for Reminder Group. For H2, between-person predictor: Group (Reminder vs Control). For H3/H4, test performance was percentage correct on vocabulary tests; tests were included if within ±3 days of planned test day, length ≥5 words; multiple same-day/list tests were combined; study events were aggregated between tests; one outlier student with ~3000 studied words was excluded from test analyses. Data analysis: Logistic mixed-effects models for H1 and H2 with random intercepts and slopes, using likelihood-ratio tests for p-values. H1 included only days without scheduled tests and only instances where reminders were actually seen before studying; scheduled-but-unseen reminders were coded as no reminder. Exploratory analyses contrasted Reminder Group day types (with vs without reminders) to matched subsets of Control Group days (mock reminder schedules created via odd-even split mirroring reminder orders). Time trends examined interactions with day or study-test cycle. H3 used a linear mixed-effects model predicting test performance from Group with test-length weighting. H4 modeled test performance from number of learning days and amount of vocabulary studied (person- and grand-mean centered); convergence issues and noisy adherence led authors not to report H4 results (details in supplements). All analyses in R 4.0.3.

Descriptives: Control Group studied on 71% of intervention study days (SD=27%; ~26/36 days). Reminder Group studied on 62% (SD=29%; ~22/36). Within Reminder Group: study probability 69% (SD=31%) on days with reminders vs 60% (SD=29%) on days without. Study app usage frequency was similar across groups (Control 77% of days, SD=24%; Reminder 76%, SD=23%; t(82)=0.04, p=0.972, 95% CI [-0.10, 0.10]). Test participation: mean tests per participant M=2.86 (SD=1.21); mean score 76% (SD=17%); Control 76% (SD=18%), Reminder 77% (SD=16%). H1 (within-person effect of reminders): Significant positive effect of receiving a reminder on the likelihood to study that day (b=0.57, 95% CI [0.28, 0.826]; χ²(1)=14.39, p<0.001). Odds ratio OR=1.77 (95% CI [1.32, 2.37]). The reminder effect increased over time (b=0.31, 95% CI [0.01, 0.62]; χ²(1)=3.78, p=0.046). H2 (between-group frequency): No significant difference in day-level study likelihood between Reminder and Control Groups (b=-0.64, 95% CI [-1.41, 0.14]; χ²(1)=2.56, p=0.106). Follow-up decomposition vs Control: On days with reminders, Reminder Group did not differ from Control (b=-0.02, 95% CI [-0.92, 0.89]; χ²(1)=0.00, p=0.960). On days without reminders, Reminder Group was less likely to study than Control (b=-0.79, 95% CI [-1.53, -0.05]; χ²(1)=4.23, p=0.037), OR=0.45 (95% CI [0.22, 0.95]). Over time, a trend suggested the no-reminder disadvantage in the Reminder Group increased (day type × day interaction b=-0.05, 95% CI [-0.09, 0.00]; χ²(2)=6.86, p=0.061); no such trend on reminder days (b=-0.01, 95% CI [-0.06, 0.05]; p=0.815). H3 (test performance): No significant group difference (b=0.05, 95% CI [-0.38, 0.47]; χ²(1)=0.05, p=0.829). H4 (dose–response): Model encountered convergence issues amid noisy adherence; results deemed unreliable (see supplements). Overall, reminders increased same-day studying when present but did not increase overall study frequency versus control; rather, Reminder Group studied less on days without reminders, indicating potential overreliance.

The findings address the core question of whether digital reminders foster regular studying for distributed practice. Reminders produced a clear within-person boost: students were more likely to study on days they actually received a reminder, and this effect grew over time. However, comparing groups revealed no net advantage in overall study frequency or test performance for those receiving reminders versus controls. The group-level pattern arose because Reminder Group students studied less on days without reminders compared to controls, consistent with an overreliance mechanism: studying became contingent on receiving prompts rather than internalized or cue-based routines. This undermines the goal of promoting self-regulated learning and habit formation, where the target behavior should persist without external prompting. The authors argue that reminders alone may not effectively establish stable environmental cues necessary for automaticity; instead, interventions that help learners identify consistent situational cues and create implementation intentions may better support habit formation and sustained distributed practice. The null differences in test performance align with equivalent overall study frequencies and the noisy test-taking behavior, limiting sensitivity to detect learning gains. Overall, reminders can be effective short-term triggers but risk creating dependency, highlighting the importance of designing mobile interventions that fade prompts and scaffold cue-based habits.

This study shows that smartphone reminders increase the likelihood of studying on a given day but can engender overreliance: students receiving reminders were less likely to study on days without them, yielding no overall advantage versus controls and no differences in test performance. The work contributes by combining within- and between-person designs with objective app log data to uncover both benefits and hidden costs of reminders in an educational context. Future research should prioritize habit-formation mechanisms (e.g., implementation intentions, cue selection), explore just-in-time adaptive and micro-randomized trial designs to optimize timing and content, and evaluate long-term maintenance as prompts are tapered. Personalization that accounts for learners’ contexts, motivations, and dynamic needs, along with clear rationales for distributed practice, may enhance durable self-regulated learning behaviors and downstream academic outcomes.

Time-trend analyses were exploratory and require replication. Generalizability is limited to German lower secondary students aged 10–12 and vocabulary learning tasks; effects may differ for older learners, other materials, or strategies. Test data were noisy due to variable adherence (missed/multiple/early/late tests and heterogeneous lengths), leading to limited comparability and convergence issues for dose–response analyses. Although both groups had similar study app usage rates and identical notification schedules, potential confounding from study app interactions cannot be fully excluded. The final sample size (N=85) was slightly below the preregistered target (N=102). Reminders were delivered only when students opened the study app, so scheduled-but-unseen reminders were treated as no reminder, which could attenuate effects. Incentives rewarded study app but not vocabulary app usage, possibly influencing engagement patterns.