Internet Learning Volume 6, Number 1, Spring 2017/Summer 2017 | Page 14

The Value of Common Definitions in Student Success Research: Setting the Stage for Adoption and Scale
framework that supported acting on
analytic evidence from the dataset. The
education theorists played a critical role
of tying the data science innovations to
what was known about student success.
Researchers within the PAR community
began to explore whether the PAR
dataset could be extrapolated to create
an updated model for retention and
progression. In reviewing seminal retention
studies including Tinto (1987),
Bean and Metzner (1985), and Falcone
(2011), the researchers developed an
updated PAR model of retention as
shown in Figure 2 below.
Based on these initial findings
and the research literature, the PAR
model (Daston, James, & Swan, 2015)
shown in Figure 2 begins with learner
characteristics, the relatively consistent
attributes students bring to the learning
experience. It views these characteristics
as being filtered through instructor
behaviors in the courses they take, the
characteristics of those courses themselves
(course characteristics), and other
supports, supports not aimed at specific
parts of the model, most importantly
financial aid. These influence learners’
feelings of FIT or academic and psycho-
Figure 2. PAR Model of Factors Affecting Student Retention and Progression
social integration, which in turn affects
how the learners behave in their courses
and programs (learner behaviors),
including their decisions to continue
their studies (retention/progression).
The model also shows learner behaviors
feedback to impact the factors contributing
to them and suggests where data-driven
interventions might address
these specific categories of what the
model views as predictors of retention
and progression.
In a process similar to that involved
in the creation of the data definitions,
PAR researchers classified interventions
by the predictors identified
in the research as affecting student success.
The review of the literature used
to substantiate the new framework
also revealed that, previously, student
interventions were often classified by
13