Reliable internet access remains a significant barrier to digital learning in many rural and disadvantaged regions. This study examined the effectiveness of offline digital mathematics resources in supporting learning outcomes in a connectivity-constrained madrasah setting in Indonesia. Using a quantitative quasi-experimental pretest-posttest control group design, the study involved 62 Grade 4 and Grade 5 students from Madrasah Ibtidaiyah 01 Sinunukan, a school located in a 3T (frontier, outermost, and disadvantaged) region of North Sumatra. Thirty-one students participated in an intervention using offline-enabled tablets preloaded with GeoGebra, while 31 students received conventional instruction. Data were collected through a validated mathematics achievement test, a Technology Acceptance Model survey, and usage logs. Statistical analyses included descriptive statistics, paired and independent t-tests, effect size calculations, and Pearson correlation analysis. Results showed that students exposed to offline digital learning achieved substantially greater learning gains than those in the control group. GeoGebra usage frequency was positively associated with learning improvement, while students reported favorable perceptions of usefulness and ease of use. Because neither condition used internet-connected tools, the design evaluates the offline GeoGebra-supported collaborative tablet package against conventional individual seatwork rather than isolating offline from online delivery. Within this single-madrasah sample, the findings demonstrate that offline digital resources can operate reliably without internet connectivity and support mathematics learning gains in connectivity-constrained classrooms, informing infrastructure-responsive educational technology planning for rural schools and open and distance learning settings.
Torang Siregar (Researcher)