University completion rates won’t be improved by looking at isolated causes
Type of Publication: Professional commentary
Lead Organisation: Curtin University
Year Published: 2017
Lead Researcher: Tim Pitman
Written by Dr Tim Pitman for The Conversation
Originally published on 18 January 2017
News media are widely reporting on new data released by the government showing that one-third of students starting university in 2009 had not finished their studies within six years.
This stat makes a good headline, but oversimplifies the reality, which is detrimental to improving higher education standards.
When you drill down into this data, the picture is very different for a number of universities. Completion rates range from 36.9% to 88%.
Those in rural and regional areas – in Queensland in particular – struggle the most to retain students, and accounted for seven of the ten lowest-completing institutions. Those based in the city have the highest rates of completion.
This is not necessarily a reflection on the quality of the educational experience that rural and regional universities provide, but reflects the demographics of the students they support.
So why is it that some universities – mainly those in rural and regional areas – are still struggling to reduce drop-out rates?
More studying part-time?
One key message of the report is that type of attendance, among measured variables, appears to have most influence on student completion. That is, part-time students are less likely to complete their studies than full-time students.
A cursory read of the part-time enrolments in individual universities might seem to support this argument. Generally, universities with below-average completion rates have above-average part-time enrolments, and vice versa.
However, the reality is more complex. For example, the Australian National University has one of the best completion rates but above-average part-time enrolments. Conversely, Federation University has one of the worst completion rates, but lower-than-average part-time enrolments.
The fact is, as the report itself acknowledges, despite part-time enrolment having the largest influence of all variables measured, it still explains only 6.31% of the variation in completion rates.
The report acknowledges that the method of analysis may overstate the strength of the relationship between particular factors and completion, and that a range of factors are “less amenable to measurement or unmeasurable”. However, these important caveats, by and large, do not make it into the media releases.
Reporting that part-time students struggle more than full-time students to complete studies, or that one in three students don’t complete, makes for good headlines but does not accurately reflect the real issues nor help experts advocate for meaningful change.
For example, many mature-age students study part-time because of work and family commitments. Encouraging them to move to full-time studies would exacerbate the problem in most cases, not solve it.
Other research has found that factors such as socio-economic status, indigeneity and geographical location are significant predictors of lower completion rates.
These important factors were downplayed in this latest report.
Even more important is the need to consider the effect of multiple factors. In other words, consider a student’s part-time status as sometimes being a cause of disadvantage and at other times a consequence.
One way of thinking about it is to consider a student failing to complete as a boat sinking because of multiple holes. Pointing out one leak and fixing it won’t stop the boat sinking.
Similarly, related analyses of graduate outcomes show that it is not necessarily about being Indigenous, or living in a regional part of Australia, but about how the two – and other factors – come together.
Result of demand-driven system?
Concerns have regularly been raised that increasing access to higher education will lead to universities admitting lower-quality students. And that this may in turn cause a spike in non-completions.
However, there is no evidence yet that the demand-driven system – introduced in 2012 to allow universities to recruit as many students as they wish – has led to an overall decline in academic standards.
Students’ ATAR scores
The report confirms that the higher a student’s ATAR score – an overall score that is given to students when they leave school – the more likely it is that the student will complete. This finding is in line with previous research which found that prior academic performance is the main predictor of subsequent university performance.
But when considered as the only variable, an ATAR explains only 3.86% of variance in completion rates. So having a lower ATAR score in no way means the student is destined to fail. However, there is a danger that this will be a take-home message for many readers.
Students are taking longer to complete
The report looked at how many students had completed within four, six and nine years after starting university. It found that while the four-year data is showing a small but steady decline in completion rates, the longer-series ones indicate that students are still completing, just taking longer to do so than before.
Four-year completion rates have dropped from 47.4% to 45.2% (2005-08 compared to 2011-14). Six-year completion rates have barely moved; being 67.0% in 2005-10 and 66.8% in 2009-14. Nine-year completion rates have also changed little: from 73.6 in 2005-13 to 73.5% in 2006-14.
This is line with previous research showing that increasing access to higher education may cause rates of quicker completion (eg within four years) to fall.
Research in the US suggests that taking longer to complete studies is due to four reasons: bad planning or advice, switching courses, switching universities, and taking unnecessary units. This may prove a useful guide to direct research in Australia to look at why the same is occurring here.
Governments need to stop considering individual factors such as race, age or wealth in isolation when measuring higher education disadvantage. This requires researchers and policymakers to do the same. Ultimately, they are all proxies for real disadvantage, something that is distinct to most individuals.
This article is republished from The Conversation under a Creative Commons license. Read the original article.