Loading...
Predicting student attrition at an urban college
Staman, E. Michael.
Staman, E. Michael.
Abstract
Abstract
The purpose of the study was to determine whether the variables related to students' goals, reasons for attending college, academic background, socioeconomic background, basic demographic characteristic,
expectations about college, source of financial support, participation in
college activities, use of college facilities, and college choice criteria could be successfully incorporated into a multi-variate analysis
capable of predicting attrition at an urban institution. Central to the
study was a test of Tinto's theoretical model of attrition, a model that
suggests in part that while background characteristics are important in
predicting attrition, student expectational and motivational attributes
are at least equally important. Equally important was a test of the suggestion, which appears in virtually every synthesis of the literature,
that the attrition phenomenon is multi-faceted.
The college used for the study is an urban, commuting institution
with a large percentage of non-traditional students. Two groups of students were identified— students who attended the orientation program and
continuing students. Parallel questionnaires were administered to each
group. A total of 1314 students responded. Factor analytic techniques
were used to reduce the items in the questionnaires to a set of predictor variables.
A model was developed to predict which students would or would not
persist in the college studies. Discriminant analyses were used to
determine which variables from the questionnaire to use in the model.
The model demonstrated that for the college being studied, attrition
is indeed a multi-faceted phenomenon. Twenty different variables entered
into the prediction formulas for orientation students, ages 17-21, 22 for
continuing students ages 17-21, and 18 for continuing students, ages 22-
45. The model demonstrated that for the college being studied, the congruency model proposed by Tinto can be accepted. Some background characteristics entered the prediction formulas for non-traditional and
traditional students, and expected involvement factors played an important role for orientation students, satisfaction factors for continuing
traditionally aged students, and job related factors for continuing nontraditional students. The model also demonstrated that the variables
that predict attrition for the non-traditional students are not the same
as the variables which predict attrition for traditional students. In a
purely quantitative sense, five variables that entered the prediction
formulas for non-traditional students did not enter the formulas for
either of the traditionally aged students, and 23 variables which entered
the formulas for traditionally aged students did not enter the formula for
non-traditional students.
Finally, Tinto's model generally agrees with the basic congruency
argument proposed by (among others) Feldman and Newcomb, with the exception that Tinto emphasizes background characteristics to a greater extent
than do earlier versions of the concept. The model developed in this
study incorporated this increased emphasis and expanded Tinto's model to
include current environmental factors which are completely separate from
the college environment. Acceptance of the model may suggest that a
congruency model for non-traditional institutions must incorporate
factors related to expectations and motiviation, variables describingbackground, and factors or variables which relate to a student's environment external to the college. In one sense, as these external factors
change, the commitment to complete college changes— thus a model where
motivational factors dominate may be appropriate. In a non-traditional
environment, however, external factors may play a more dominant role
than in a traditional environment, thereby requiring a more direct incorporation of these factors into the model.
At the college used in this study the congruency model seems appropriate for orientation students and for non-traditional students. For
the traditional, continuing student, however, the model proposed by Astin
seemed more appropriate, suggesting that satisfaction and involvement in
college may be more important than motivation and expectations about
college for the traditional student
Description
Date
1979
Journal Title
Journal ISSN
Volume Title
Publisher
Collections
Download Dataset
Files
Loading...
8002565.pdf
Adobe PDF, 5.89 MB
Rights Holder
Usage License
Embargo
Research Projects
Organizational Units
Journal Issue
Keywords
Citation
Advisor
Department
Education
DOI
https://dx.doi.org/doi:10.25774/w4-qj6v-fh94