Submission by AMSI Management Committee 3: Mathematics and statistics research (including interdisciplinary research) in universities and related institutions

Prof Geoff Prince
AMSI Director

The full submission by AMSI management can be viewed here.

Theme 1: Enhancing our PhD programs: A graduate school offering a full year’s graduate coursework delivered across Australia?

There appears to be broad agreement in the AMSI membership that Australian PhD students in the mathematical sciences need more coursework. Indeed the ATN has already moved that way in the creation of the collaborative Industrial Training Centre in the Mathematical Sciences. However, it is clear that the provision of a formal program is beyond the individual resources of most of our member departments and a collaborative approach will be required. Even so there are a number of impediments, such as the funding of an additional year of candidature and stipends and the breadth of offerings required to cover the sub-disciplines.
At the moment AMSI offers the Winter School which is a purely postgraduate program, the Summer School, half of whose attendance is made up of PhD students, and the bioinformatics symposium, BioInfoSummer. These programs are jointly funded by the Commonwealth but at the moment there are some restrictions on multiple attendances which would prevent annual attendance by most students. Then there are the AGR honours courses and short courses which are offered co-operatively by AMSI members. Expanding AGR offerings presents a logistic problem because there is only partial coverage by AGRs and because subjects can’t be run in parallel. At the moment the rooms are used around 16 hours per week including seminars.

Having said this AMSI is the natural facilitator of a future formal program because of its existing network and because its members expect us to deliver programs of this sort. If decadal plan process indicates widespread support and ideas for the program AMSI will undertake to actively pursue the proposal.

Relevant AMSI policies:

Provide a dedicated allocation of Australian Postgraduate Awards (APAs) in mathematics and statistics to the universities to improve retention from honours and masters programs. Incentives for universities which provide such a dedicated allocation from their own award program. Especially important in smaller and regional universities.
Provide a stipend top-up on APAs to improve retention from honours and masters programs. This is particularly important in statistics where employment demand is severely reducing retention.
Re-weight the funding of PhDs in mathematics and statistics to match those in the physical sciences.

Theme 2: Breadth and Depth of mathematical/statistical research

Breadth and depth continue to be problematic in the Australian Mathematical Sciences, exacerbated by poor communication amongst the research sites. The poor staffing position in maths and stats in Australian universities has been a brake on research diversification for nearly 20 years. Breadth is hampered by the ageing cohort of researchers in continuing positions and the existence of strong groups in a small number of areas. Smaller universities do not have the staff cohort or the research funding to develop significant breadth. Depth is significant in several key areas, but poor in many areas of international significance, some emerging and some long established. There has been no national strategic plan produced by the learned societies in mathematics to remedy the weak overall position in certain critical areas, and universities tend to take a conservative approach to new appointments, usually making them in areas of existing strength. Statistics from the AMSI workshop program support this profiling, showing that typically only around 3 Australian sites provide 75% of attendees at any given workshop, with a small number of remaining sites sending only one or two participants.

Australian institutions represented at AMSI workshops

The ERA results are useful in analysing the discipline’s strengths and weaknesses but not definitive.
The 2012 AMSI Member Survey and the 2013 Discipline Profile indicate capacity and resourcing of the discipline but not by sub-discipline.

Poor communication amongst research sites also prevents the achievement of critical mass. While AMSI has been successful in creating a national workshop program supporting 15-20 workshops annually, collaboration between the university sector and the government agencies and between the university and private sectors remains under-developed. AMSI, the learned societies and the parties concerned need to remedy this problem with a targeted program.

Theme 3: Structure and form of Research funding

There are broad issues here of concern to all disciplines around the heavy burden associated with application to the ARC. Our discipline should be engaged in these discussions.

There are a few observations about the current funding situation for the mathematical sciences:

We have had the best success rate in ARC DPs of any discipline for a number of years.
The current ARC schemes re-enforce research concentration at the larger sites.
Our performance in ARC Linkage Grants is woeful and makes us vulnerable.
Absence of a CoE is symptomatic of our inability to engage with the government’s research priorities.
Absence of a national centre is degrading our international standing.

On May 29th 2013 the ARC released a document profiling our discipline’s track record with them, this can be found here.

The discipline must not wait until the decadal plan is released to address these matters. AMSI’s Research and Higher Education Committee has a number of these items on its agenda and the learned societies need to discuss issues around support for research outside the main centres.

Relevant AMSI policies:

The full cost of successful research programs should be funded by the ARC.
The ARC should institute a small grant scheme. This will significantly boost international collaborations.
AMSI must have Commonwealth funding for a rolling 4 year research program in order to sustain Australia’s international collaborations in the mathematical sciences. This should include increased e-research capacity within mathematics departments.

Theme 4: Interdisciplinary Mathematics/Statistics

I do not know of any international comparative study against which we can measure our success, but I suspect that if the universities bench-marked themselves against CSIRO CMIS for example we would find a lot less multi-disciplinary engagement, even amongst the traditional areas of statistics and applied mathematics. But then other disciplines are probably similar because of the silo effect and factors like the ERA.

In our submissions to Committees 4 & 5 we have discussed the issue of improving the external engagement of university departments, and it is critical for those departments to look beyond their faculties and universities for multidisciplinary engagement. AMSI and the learned societies should take the lead in creating opportunity for multidisciplinary engagement by running dedicated events, including workshops and information sessions for prospective Linkage Grant applicants. The discipline should embrace PhD internships. Where possible we should consider joint appointments in areas such as systems biology, engineering and environmental science.

This is an area where initiative is our own responsibility!

Theme 5: Research Centre in Mathematics/Statistics

See our submission to Committee 6.

Further information on relevant AMSI initiatives:
Summer School

Winter School

BioInfoSummer

AGR subjects and courses

AMSI workshop program

Further reading:
Australian Mathematical Sciences Institute, Discipline Profile of the Mathematical Sciences 2013 (2013)

Australian Mathematical Sciences Institute, AMSI Member Survey 2012 (2013)

Policy measures in the national interest (March 2013)

Department of Innovation, Industry, Science and Research, Research Workforce Case Study (2011)