by Mohamed W. Soliman*, Tamer M. Kamel , Nagwa Elshamy Elshamy Mohamed
ABSTRACT
This study examined the effect of early-warning alert timing, represented by proactive versus remedial alerts,
within a learning analytics-based personalized learning environment, on educational technology students’
self-regulated learning and academic decision-making. The study adopted a quasi-experimental design with
two experimental groups and pre- and post-measurements. The first experimental group studied the e-learning
course through a personalized learning environment based on proactive alerts, whereas the second
experimental group studied the same course through the same environment, with the alert pattern modified to
provide remedial alerts. The sample consisted of 100 male and female fourth-year students from the
Department of Educational Technology, Faculty of Specific Education, Alexandria University. The participants
were equally assigned to two experimental groups, with 50 students in each group. Six main measurement
instruments were used: the Self-Regulated Learning Skills Scale, the Self-Regulated Learning Performance
Rubric within the environment, the Learning Analytics and Alert Response Log, the Academic Decision
Making Scale, the Academic Decision-Making Situational Test, and the Academic Decision Quality Rubric. In
addition, four instruments were used for control and interpretive purposes: the Design Standards Checklist for
the Personalized Learning Environment, the Alert Scenario Validation Rubric, the Environment Usability and
Alert Clarity Questionnaire, and the Semi-Structured Interview Guide. The data were analyzed using the
independent-samples t-test to verify pre-measurement equivalence, and analysis of covariance (ANCOVA) to
compare the two groups in the post-measurement after controlling for the effect of the pre-measurement.
Partial eta squared, Cohen’s d, Moodle log analysis, and thematic analysis of the interviews were also used.
The results revealed statistically significant differences in favor of the proactive-alerts group across all main measurement instruments, with effect sizes ranging from moderate to large. These findings indicate that
providing alerts before actual academic difficulty occurs gives students a better opportunity to plan, manage
their time, use support resources, review their performance, and make preventive academic decisions based on
learning data, compared with remedial alerts that are provided after difficulty has already appeared. The study
recommends that personalized learning environments should incorporate interpretable proactive alert
mechanisms linked to clear learning indicators, in ways that support self-regulated learning and academic
decision-making without replacing students’ responsibility for their own learning.
