Skip to content

In Irish universities, males get more firsts, females get more seconds and everyone gets more of both

January 21, 2013

New Report on University Grades
A new report by Martin O’Grady of the Network for Irish Educational Standards shows that while grade inflation is showing clear signs of slowing down, it has, nevertheless, continued in the university sector up to 2009, the latest year for which figures are available. In searching for possible explanations for differences in the percentages of firsts and upper seconds across the seven universities, interesting findings have been unearthed in relation to gender differences, the relevance or non-relevance of mature students and differences between universities in size and emphasis on postgraduate study.

Firsts Hit a Plateau but Upper Seconds Continue to Climb

image002

FIGURE 1: The 7 Universities Combined – Rate of Firsts and Upper Seconds 1994-2009

The report updates the previous analysis, by O’Grady and Guilfoyle (2007), of the 1994-2004 figures, adding in those for 2005-2009. A glance at Figure 1  taken from the report shows that something interesting did seem to happen after 2005. Having peaked at 17.5% in that year, following many years of growth, the percentage of firsts in the combined universities showed a dip the following year and stayed level thereafter, at least up to 2009, the latest year for which figures are available.

Interestingly, this pattern was replicated to some extent in each one of the seven universities. In NUIM and DCU there was a marked and continuing decline in firsts after 2005. In UCD there was a sharp decline in 2006 but followed by an equally sharp recovery in 2007 and then a levelling off up to 2009. TCD’s rate levelled off after 2005 but showed signs of a pick up again in 2008 and 2009. At UCC, NUIG and UL, the rates of firsts largely levelled off after 2005 up to 2009.

Are there signs that the universities may finally be getting grade inflation under control? An examination of the upper second figures, however, suggests that this would be a premature conclusion.

As evident in Figure 1, the pattern in upper seconds is different to that of firsts. While there was a clear levelling off in 2006 and 2007, the two highest rates ever for upper seconds occurred in 2008 and 2009. The percentage of upper seconds across the seven universities, which stood at 28.8% in 1994, reached 47.6% in 2009, a figure exceeded only in 2008 at 48.4%. In 1994, 7.2% of university graduates obtained a first class award. The comparable figure in 2009 was 16.4%, down from a peak of 17.5% in 2005.

An analysis of Institute of Technology grades in 2011 compared the average of grades awarded in the final two years of the previous 1994-2004 analysis with the average of the most recent two years for which figures were available, 2008 and 2009 and a similar comparsion using the aggregate university grades. Averaging figures from successive years in this way mitigates the impact of random year to year fluctuations in the figures. Because the HEA had never published the university figures for 2003, the figures for 2002 were used instead, including those for 2005 to balance off the downward pull on the average induced by the 2002 figures. Remember that the percentages of firsts and upper seconds had been increasing year on year.

Across the universities, the average rate of firsts for 2008 and 2009 was 11.4% higher than the average for 2002, 2004 and 2005. A similar comparison for upper seconds found a 13.9% increase. Not quite as much growth as in the Institutes of Technology, where firsts and upper seconds increased in a similar period by 15.7% and 17.5% respectively, but substantial, nonetheless. Definitely too soon to herald the end of grade inflation in the universities.

Better Students?

Of course, better grades might be expected if better students were getting into the universities. The opposite however is the case. On the basis that the most common duration of an undergraduate degree is four years, O’Grady compared CAO points four years in advance for the two graduate groups described above. Comparing all CAO applicants in 1998, 2000 and 2001 with those in 1994 and 1995, 28.8% of the latter had points of 400 or more as compared with 26.4% of the former. The proportion with 500 or more points had also increased from 6.7% to 8%.

This follows a pattern of inflation in Leaving Certificate grades dating back at least to the early nineties. It may be taken therefore that better grades became easier to get over the period under examination. Despite that, there was a general decline over the same period in the minimum points necessary to gain entry to undergraduate courses in the universities.The biggest change between the 1998/2000/2001 and the 2004/2005 figures was the increase from 39% to 49%  of university courses which could be entered with points below 400. An additional 3% of courses requiring over 500 points did not come near to counter-balancing the decline in points’ requirements overall.

On average, university courses became easier to access. Nevertheless, entrants with weaker points’ profiles graduated with superior grades. When the rate of firsts and upper seconds should have been in decline, they were instead in the ascendant. This is the typical grade inflationary pattern.Overall then, it would seem fair to say that while the universities have shown some definite signs of grappling with the grade inflation issue, their report cards should state: ‘Signs of going in the right direction but a great deal of ground to be made up; must work harder to reassert standards.

Variance in Standards Across the Seven Universities

In addition to analysing the trend in grades across the university sector as a whole, the report explores the differences between the individual universities on CAO points of their entrants and on the grades awarded to their graduates. It also seeks to identify mediators of grade increase such as discipline and gender mix, focus on postgraduate study and university size.

Looking at the average annual rate of firsts and upper seconds awarded over the period 2005-2009, Figure 2 below shows that the differences across the seven universities are far from negligible.

image002

Figure 2: Average annual rate of firsts and upper seconds (2005-2009) with weighted mean of 2002 median CAO points in brackets (TCD figure corrected for absence of two subject moderatorship courses)

Such differences O’Grady hypothesised might result from differences in the academic quality of entrants attracted to the various universities. To test for that possibility he used the entry points’ data available on the CAO website to compute the typical entry points for students in each university. The most accurate estimate could be conducted for 2002 because in that year the CAO, not only published the minimum and median points for entry to each university undergraduate course but, uniquely, also published the number of places on each course.

Using those figures, the average of the median points for all courses at each university weighted by the number of places in each course was computed as an index of the typical student entry CAO points. This weighted figure and the simple mean of median points bore a very close relationship and the comparison between universities in the latter figure does not vary much year to year. It proved possible, therefore, to use the 2002 figures as indicative of the typical points on which students entered the universities between 2005 and 2009.

As can be seen in in Figure 2, the typical points on which students enter the different universities does little or nothing to predict the variance among the colleges in award of firsts and upper seconds at graduation. All other things being equal, this strongly suggests differences in standards. It may simply be easier to get a better grade in some universities, notably DCU and NUIM, than in others.

The Gender Effect

There may, however, be explanations for differences in grade rates other than simply differences in standards across the universities. Females perform better than males in virtually every single Leaving Certificate subject. It might be expected, then, that universities attracting more females, perhaps because of the discipline mix on offer, would, therefore, produce a better grade profile among their graduates. Here a surprise awaits. It is generally not females but males who get the bigger share of firsts, though, interestingly, females do generally get more upper seconds.

boysngirls

Girls more points; boys more firsts; girls more upper seconds

Female graduates from the universities have consistently outnumbered male graduates since 1994 with ratios varying from 1.1: 1 in 1994 up to 1.4:1 in 2004. The imbalances within individual universities have been more extreme with NUIM and UCC recording ratios in favour of females as high as 1.8:1 in some years. There are also imbalances in the opposite direction with UL recording an annual ratio of 0.7:1 and TCD of 0.9:1 in certain years.

NUIM and UCC have tended to be the universities with most female graduates whereas UL, TCD and DCU have tended to be the ones with most males. Comparisons of the gender ratio of graduates in given years with the rate of firsts and upper seconds failed to show any consistent patterns across the universities. Gender imbalances do not seem to help explain variance in grade rates.

The Impact of Discipline Mix

The possibility that variance in the mix of disciplines across the universities might help to explain grading variance was explored on the basis that there might be a general tendency to award higher grades in some disciplines than others. If that were the case, differences in the proportions of graduates in various disciplines across the universities might account for differences in the percentage of firsts and upper seconds across the seven universities.

To test that hypothesis, the discipline mix across three universities, TCD, DCU and NUIM was compared. TCD was chosen for its higher CAO points’ profile not being reflected in its rate of firsts and upper seconds. Both DCU and NUIM were chosen for the contrasting pattern of relatively low profiles of CAO points leading to higher than expected rates of firsts and upper seconds. NUIM was also chosen for being the university with, by far, the greatest level of grade increase since 1994.

As with the gender comparison, no predictive relationship could be found between the proportions of graduates in specific disciplines and a propensity to award more or less high grades at graduation across the three universities. Discipline does not seem to be an important mediating variable for the award of more firsts and upper seconds.

The Growth Imperative

Two further hypotheses were tested in an effort to explain grade rate differences across the universities: that such differences might be a product of differing emphasis on sending graduates through to postgraduate courses and that such differences might be due to differences in university size as evidenced in the number of students admitted each year.

The former hypothesis derives from the possibility that universities, more anxious to populate large postgraduate programmes, might be tempted to lower standards so as to increase the number of graduates with a first or upper second, the usual requirement for access to postgraduate study.

The latter hypothesis follows from the O’Grady (2011) finding that in the Institute of Technology sector smaller providers tend to award a disproportionate rate of better grades relative to the CAO points’ profile of their entrants. He suggested that this finding was due to greater pressure at smaller, less popular Institutes to retain weaker students in the growth focused culture, which has pervaded higher education in Ireland for many years. The same might, therefore, be true of the smaller universities.

Emphasis on postgraduate study was measured as the ratio of registered postgraduate students in a given year to the number of undergraduate graduates the previous year and by the level of growth in that ratio over time. Very substantial differences were found in emphasis on postgraduate study. For example, in 2009-2010, DCU showed the highest ratio of registered postgraduate students to undergraduate graduates the previous year with a ratio of 2.1:1 as compared with a ratio of 1:1 at both NUIM and NUIG, the lowest ranking universities in this respect. Apart from DCU’s elevated ranking in both measures, there was no pattern of high PG:UG ratios going together with high awards of firsts and upper seconds. Emphasis on postgraduate study was not found to be a predictive factor for higher grades.

size

Does Size Matter? It does in the IOT Sector.

The size of universities did emerge as a potential predictor of high grade awards. Of the 15,703 students who emerged with undergraduate degrees across the seven universities in 2009, the following are the proportions accounted for by each of the universities: UCD (20%), TCD (19%), UCC (17%), NUIG (17%), UL (11%), NUIM (8%) and DCU (7%). The top four are all double the size of the bottom two with the largest having nearly three times as many graduates as the smallest.

It is striking that it is the two smaller universities that stand out in terms of their award of firsts and upper seconds by comparison with the CAO points’ profiles of their students. In Figure 2 above, DCU is second only to UCC in the award of firsts and upper seconds despite having a points’ profile ranked in joint second last place.

While NUIM ranks second last in the award of firsts and upper seconds combined, it has almost identical figures to UCD and its rate approaches closely to NUIG and TCD, despite its points’ profile being 77 points short of TCD. It is perhaps not an accident that since 1994 both DCU and NUIM have increased their share of the overall undergraduate market, having both commanded a 6% share in 1994, while UCD and TCD have experienced significant declines, down from 27% and 25% respectively in 1994.

It is tempting to suggest that the two smaller universities have grown at the expense of larger ones by taking in weaker students and dropping their academic standards. In such a scenario the high rate of top grades would have been a consequence of the overall decline in academic expectations required to enable the weaker students to progress in their courses and ultimately graduate.

A Problem Common to All

Overall, however, the latest report suggests more than anything that factors common to all seven universities account for the bulk of the grade increase over the years with factors by which they can be differentiated, with the possible exception of university size, showing little or no correlation with the rate of firsts and upper seconds. There is clear evidence that despite quite large variance in the academic ability of the entrants they attract, the seven universities have grown much more homogenous in their award of the better grades. It is as if they are benchmarking against each other in a crude quantitative way rather than maintaining similar qualitative standards.

Serious questions need to be asked about the efficacy of the external examination system which is the main quality benchmark for examination standards. It has singularly failed in the past and, on the evidence reviewed above, is still failing to maintain parity of degree standards over time. It is also failing, it would seem, to maintain parity of standards across the various universities. None of this is good for higher education. If, as a nation, we are to produce degrees we can trust, a large measure of objectivity will need to be brought to bear on standards. Grades and qualifications must be known to mean what they should mean, independent of time and location.

Advertisements
2 Comments leave one →
  1. Joe MacDonagh permalink
    January 23, 2013 2:13 pm

    As ever from Martin O’Grady an excellently researched and closely argued piece. As he says in the last sentence:
    “Grades and qualifications must be known to mean what they should mean, independent of time and location.”
    Research such as this, plus his previous work, is essential to a properly functioning educational system; such a system should not be characterised by rote learning and the expectation of high grades but should be one where students engage with their subjects and are willing to read widely for their degrees.
    This is superb research which I would like to be applied to the other end of the distribution; those who are often passed, as he suggests, in order to keep numbers up, and so state funding in place, for smaller colleges. Perhaps this means a lot more students passing than should do so. I’m sure O’Grady will cover this in his next piece of research and I look forward to seeing that when it is finished. In the meantime well done!

  2. May 8, 2014 6:25 pm

    The American grading system is better than this ridiculous 1st, 2nd, 3rd system. In the 50’s and 60’s the % of 1st class degrees in Irish universities was about 2(%). Now it is 15%. What gives?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: