Analysis of Retention Rate and Graduation Rate in Online Universities

This analysis examines the retention rate and graduation rate in online universities in the US, highlighting the issue of dropouts and providing recommendations for improvement.

ECONOMICS AND QUANTITATIVE ANALYSIS
Student ID:
[Pick the date]

ECONOMICS AND QUANTITATIVE ANALYSIS
Purpose
The online education has gain popularity during the last decade or so and led to significant
number of students availing these courses. As a result, it makes sense to analyse if the
performance of the online universities for which the given task focuses on two measures in
the form of RR% (Retention Rate) and GR% (Graduation Rate). Collectively these two are
pivotal factors that highlight the extent of drop-outs highlighted in the online universities in
US. This is pivotal as higher drop outs reflect potential issues that would need to be sorted in
the long run through pragmatic recommendations.
Background
The large scale internet penetration has altered the ecosystem related to product and service
delivery. Using internet as the medium, it became possible to transmit the various services to
far flung corners and avail these in the comfort of chosen location. Education also witnessed
significant changes owing to the growth in online courses which over the time have become
popular. This does not come as a surprise considering the benefits in the form of lower costs,
increased convenience, better time management and flexibility in time and geography for
availing various courses (Lederman, 2018). One key downside of these courses that has been
indicated through various empirical evidence is the high rate of dropout that are experienced
(Wellman, 2018). In order to ensure that these courses make a significant positive
contribution to education and economic development, it is pivotal that the issue of dropouts
must be kept within check.
Method
The sample data provided which contains information about RR% and GR% for a set of 29
online US universities acts as the pivotal first step. Using measures of central tendency along
with dispersion measures, descriptive statistics are represented that indicate the summary of
the sample data. Further, a scatter plot has been drawn to explore the level of association
between RR% and GR%. In order to carry out a detailed analysis on the same, regression
analysis is carried out based on Excel. Besides, inferential statistical techniques such as
hypothesis testing have been used to ascertain the statistical significance of slope and linear
relationship.
1

ECONOMICS AND QUANTITATIVE ANALYSIS
Results
a) The descriptive statistics for the same data are presented below.
b) The scatter plot has been drawn using Excel and illustrated below.
The first noteworthy aspect is the arrangement of scatter points which exhibit an upward
trend which hints at the positive relationship between the two variables under consideration.
Further, the position of the scatter points is that the deviation of these points from the
regression line is not high which implies that association is strong in strength. This is
confirmed from the computation of correlation coefficient which comes out as 0.67 (Hillier,
2016).
2

ECONOMICS AND QUANTITATIVE ANALYSIS
c) The regression model output as derived from Excel is pasted as follows. The independent
variable is RR% with GR% as the dependent variable for this model.
d) The following equation captures the regression model which is pasted above.
In the above equation, 0.285 is the slope coefficient. This may be interpreted as the change
which is likely to be witnessed in GR% when there is a one percent change in RR%. Also,
this change in the dependent variable (i.e. GR%) would be the same as independent variable
(i.e. RR%) as the slope coefficient is positive (Lieberman et. al., 2013).
e) The significance of the slope can be ascertained with the aid of following hypotheses.
For the given hypothesis test, assumed level of significance (α) is 5%.
3

ECONOMICS AND QUANTITATIVE ANALYSIS
It is apparent from above that t stat for the slope coefficient is 4.693 with the p value as
0.000.
The decision rule is to reject the H0 when p value fails to exceeds the significance level. This
has been achieved in the given case as p value (0.00) < significance level and therefore
rejection of H0 is confirmed. This leads to acceptance of H1 (Flick, 2015). The conclusion is
that the slope for the regression line is statistically significant thereby reflecting that
association between RR% and GR% is also significant and cannot be ignored.
f) Two key inputs which ought to be considered while deciding the fit are coefficient of
determination and also if the slope is significant or not. The coefficient of determination is
0.4492 and thereby hints that the independent variable (RR%) is capable of providing
explanation to 44.92% of the changes observed in the dependent variable (GR%). Also,
taking into account the significant slope, it would be fair to interpret that the model despite
moderate prediction ability boasts a good fit in terms of the regression model (Hair et. al.,
2015).
g) In the context of South University, if the GR% and RR% are considered, it is apparent that
even though RR% is quite robust, this is not reflected in GR%. This conclusion is reached
by considering the regression model equation where for an RR% value of 51%, the
expected GR% is sizably greater than the observed GR% value of 25%. It is imperative
that the factors leading to this need to be understood and rectified.
h) In the context of University of Phoenix, a glimpse at the RR% clearly reflects the problem
since it is the lowest for all the sample universities. Also, a 4% rate would indicate that
96% of the students who enrol tend to leave the course before completion which is a very
grave issue which can potentially have grave implications in the future. As a result, drastic
measures need to be introduced by the university for addressing this problem.
Discussion
On an average, the sample universities performance is not too bad as the RR% and GR%
seem to be satisfactory. However, the concerning aspect is that there is high variation in these
performance parameters across the sample universities. Also, a majority tend to default in one
of the two parameters which essentially leads to poor performance. Using the scatter-plot, the
association between performance parameters (RR% and GR%) is not only positive but
4

ECONOMICS AND QUANTITATIVE ANALYSIS
strong. The slope coefficient and the statistical significance associated with the same also
tend to provide support in this regards.
A particular strength of this exercise is that the conclusions have been based on results
produced through statistical techniques based quantitative analysis. However, this objectivity
and reliability may have been compromised if the sample used for the exercise was biased.
This is possible as the underlying sampling technique used for selection of the included
universities is not disclosed (Medhi, 2016). Some concern in relation to reliability of the
sample data is resolved by the fact that the results produced through analysis of sample data
is similar to recent studies on the topic conducted in a similar context. Considering the
potential issues for online universities, policymakers need to address the same through
constructive regulation.
Recommendations
The key recommendations are enumerated as follows.
1) In order to ensure that the quality of teaching, & relevance of course along with
curriculum is maintained, regulation is required or else the intended objectives related to
learning may not be met leading to drop outs.
2) These universities that offer online courses need to conduct review of the services offered
based on student feedback and make necessary amendments so as to make their course
offerings more relevant and convenient for students.
3) With regards to seriousness of students, some entry mechanism may be designed such as
an entry level test so that enrolling becomes little difficult. Besides, a large amount of fees
should be loaded at the beginning of the course so as to result in effective deterrent against
the issue of student drop-out.
5

ECONOMICS AND QUANTITATIVE ANALYSIS
References
Flick, U. (2015) Introducing research methodology: A beginner's guide to doing a research
project. 4th ed. New York: Sage Publications, pp. 56-57
Hair, J. F., Wolfinbarger, M., Money, A. H., Samouel, P., and Page, M. J. (2015) Essentials
of business research methods. 2nd ed. New York: Routledge, pp. 105-107
Hillier, F. (2016) Introduction to Operations Research.6th ed.New York: McGraw Hill
Publications, pp. 145-146
Lederman, D. (2018) Who Is Studying Online (and Where), Available [online] at
https://www.insidehighered.com/digital-learning/article/2018/01/05/new-us-data-show-
continued-growth-college-students-studying [Available February 13, 2019]
Lieberman, F. J., Nag, B., Hiller, F.S. and Basu, P. (2013) Introduction To Operations
Research. 5th ed. New Delhi: Tata McGraw Hill Publishers, pp. 134-135
Medhi, J. (2016) Statistical Methods: An Introductory Text. 4th ed. Sydney: New Age
International, pp. 67-69
Wellman, R. (2018) How to Avoid Dropping Out of an Online College Available [online] at
https://www.usnews.com/education/online-learning-lessons/articles/2018-06-01/how-to-
avoid-dropping-out-of-an-online-college [Available February 13, 2019]
6