S-06-31

S

SIMON FRASER UNIVERSITY

Senate Committee on University Priorities

Memorandum

TO:

Senate

FROM:

John Wate

Chair, SCL

Vice Presic

RE: Department of Statistics and

DATE: February 15, 2006

Actuarial Sciences External Review (06-06a)

At its February 15, 2006 meeting, SCUP reviewed the External Review Report on

the Department of Statistics and Actuarial Sciences, together with a response

from the Department, from the Dean of Science, and input from the Associate

Vice-President, Academic.

Mot ion

S

That Senate concurs with the recommendations from the Senate

Committee on University Priorities concerning advice to the Department of

Statistics and Actuarial Sciences on priority items resulting from the

external review as outlined in SCUP 06-06a.

Background

The report of the External Review Committee for the Department of Statistics and

Actuarial Sciences was submitted on April 20, 2005 following the review team's

site visit, which took place March 2-4, 2005. The response of the Department

Chair, Dr. R. Routledge, was received on August 23, 2005, followed by that of

the Dean of Science, Dr. M. Plischke on August 26, 2005. There was general

agreement on the recommendations in the External Review from the Department

and the Dean.

SCUP recommends to Senate that the Department of Statistics and Actuarial

Sciences and the Dean of Science be advised to pursue the following as priority

items:

-

1. Undergraduate Program

1.1

Develop and implement an undergraduate recruitment plan.

16

---

1.2

The Departments of Statistics and Actuarial Sciences, Mathematics, and

the Faculty of Business should consider appointing a Director of the MSSC

program for a four or five year term.

.

1.3 The Department should consider what offerings are best placed at Surrey,

for example data mining, operations research, and the MSSC program.

1.4 The Department should continue to develop plans to contribute to the new

curriculum requirements. Similarly, the Department and the Faculty of

Science should consider how best to highlight the value of a general

education in statistical literacy for all undergraduates.

2.

Graduate Program

2.1

The Department should review its graduate offerings in the theory of

statistics with a view to adding at least one additional course at the Ph.D.

level in this area. The Department should also discuss with Mathematics

how best to provide some course availability in advanced probability

theory.

2.2 The Department should review its requirements for the Masters degree

and in particular the expectations for the Masters thesis to be sure that

they are-realistic and facilitate timely completion of degrees. The

Department may also wish to consider an option for a course based

Masters degree.

2.3 The Department should proceed with plans to develop a Masters option in

Biostatistics and consider extending this option to the Ph.D. level.

2.4 The Department and the Actuarial Science group should consider

mounting a Masters Program in Actuarial Sciences in three or four years

time, with more immediate consideration of an additional appointment,

course offerings, and a post degree diploma in this area.

3.

Planning and Departmental Resources

3.1

There should be discussion of the allocation of additional resources to the

Consulting Service to increase faculty involvement in the area, and to

make the service more accessible to researchers in all areas.

3.2 The Department should review and renovate student space to make it

more functional and hospitable.

3.3 The Department, in consultation with Mathematics, should develop a plan

for long term computing support and consider appropriate funding models

to support it.

0

2.

---

RECEIVED

SIMON FRASER UNIVERSITY

-AUG-

25 25

Office of the Dean of Science

.

MEMORANDUM

Vice Presid € nt

ACAD EM C

TO: W.R. Krane

FROM: Dr. Michael Plischke,

Associate Vice-President, Academic

Faculty of Science

RE:

External Review of Statistics

&

DATE: August 26, 2005

Actuarial Sciences

I

am

writing with regard to the report of the external review committee (ERC) for

the Department of Statistics and Actuarial Sciences and the Department's response to that

report. My comments will be brief since the Department seems to agree almost completely

with the recommendations of the committee.

The report is highly complimentary of the Department's research and teaching

programs and of the quality of recent faculty hires. This is extremely gratifying since the

Faculty and University have invested significant resources in this unit. Turnin

g

now to

specific recommendations:

Recommendation

2:

It is recommended that the Departments of Statistics and A ctuarial Science

and Mathematics and the Business School consider appointing a Director of the

IvISSC

program

for afour or five year term.

I agree with the ERC and the Department that this quite successful program needs

.

some attention. Rather than simply appoint a Director, I would prefer to move this

program to SFU Surrey where it would become one of the first degree-completion

programs in Science at that campus. All the units involved in the MSSC program will

have a significant presence at Surrey and this program fits naturally with the operations

research program being developed there.

Recommendation

9:

It is recommended that the Department and A ctuarial Science Group

consider the possibility of offering a post degree diploma in A ctuarial Science with appropriate

sharing of proceeds between the University and the Department.

I fully support this recommendation and the Department's response is also

positive. The im

p lication in the recommendation that this could be a program with

differential fees is most interesting from another perspective. The proposed graduate

program in Actuarial Science, to be developed in a few years, would require a fourth

faculty member in that area. If there is sufficient income from differential fees to offset a

significant part of the cost of such an appointment, I would be inclined to strongly

support it.

Recommendation

1:

It is recoin mended that time central administration allocate some base

budget to time Statistical Consulting Service (SCS) to increase fnculh,' involvement in the area and

to

make the service more accessible to researchers in all areas.

I agree with the spirit of this recommendation and would be prepared to allocate a

relatively modest amount of funding, perhaps equivalent to a course buyout, to the SCS.

A more si

g

nificant allocation is not possible at this time.

.

3.

...2

---

Recommendation 14:

It is recommended that the Departine;it change its tea thing load fro

m

four

to three courses per year for research active faculty. This change would also entail a review of

current policies for granting teaching credits in recognition

of

other duties.

I am pleased that the Department has accepted this recommendation. This will

bring their policy on teaching loads and credit for administrative duties in line with

normal practice across the Faculty.

Recommendation

15:

The Department should consider how it can best and most effi'ctively be

involved in the Surrey initiative.

If

it is to be involved in such initiatives in the program in

operations research or data mining, then it needs to plan appropriate appointments and ways to

I

ntegrate such appointments with the Department on the Burnaby campus.

In my view there are significant opportunities for the Department in Surrey and I

am keen to have them pursued. The required faculty positions will be made available

when programs for SFU Surrey are determined.

Michael Plischke

C:

Rick Routledge

.

.

4k

---

C

•

SIMoN

FRASER UNIVERSITY

DEPARTMENT OF STATISTICS

i

BURNABY. BRITISH COLUMBIA

& ACTUARIAL SCIENCE

CANADA V5A 1S6

Telephone:

(604)

291-3803

Fax: (604) 291-43,68

To:

W. Krane.

From:

R. Routlede,

Chair

Associate Vice-President, Academic

Re:

Response to External Review Report

Date:

August

23,

2005

Attached is the Departments formal response to the 2005 External Review Report,

as adopted in principle by strong consensus at a department meeting on May 10. 2005,

with details developed subsequently through working groups and individual

consultations.

• Sincerely,

Rick Routledge

•

---

Response to the 2005 External Review of the

Department of Statistics and Actuarial Science

June 15, 2005

We were ver

y

gratified by the genera

y

ll

favourable

assessment of the Department, and intend to pursue

all the substantive recommendations, subject to

Universit y

support where needed. Following is our

response to individual recommendations and other

pertinent items.

RL'conunendation

1: Jr is recommended that the

Department develop a plan for recruitment of strong

quantitative undergraduates into the statistics major

program .

Response

We recognize the value of encouraging

more students with the appropriate mix of talents to

further their training in statistics. To this end, we

propose to develop the following recruiting strategy.

+ We would like to work with John Simms, high-

school liaison officer in the Faculty of Science, to

enhance the statistics component to his information

packages and presentations. We shall also offer to

participate in some of his high-school visits where

appropriate, and shall continue to participate in his

Scientific l' and other similar programs.

+ We also agree with the informal

recommendation in the text of the report that we

should take advantage of the professional

certification program recently implemented by the

Statisticai Society of Canada. Indeed, three of us (C.

Dean, R. Routlede, and C. Schwarz) are now

formall y

certified. In addition. C. Schwarz played a

key role in developing and implementing the

program and has recently accepted a nomination to

serve as the incoming Chair of the Accreditation

Committee. R. Routledge also served on the Initial

Accreditation Committee. The Initial Accreditation

Committee has developed a proposal for accrediting

courses for the educational component of the

certification. The Societ

y

's Board of Governors has

recommended minor revisions to the proposal. and is

scheduled to vote on a revised proposal at its

October meeting. If the proposal is approved, then

ve anticipate that courses in our majors and honors

programs will be accredited, and that our graduates

would qualif

y

as Associate Statisticians. We

anticipate that this will be a valuable asset for our

students, and shall apply

for

accreditation as soon as

the opportunity arises.

+ We further concur -with the Review Committee's

recommendation that we focus on introductor,'

courses, and especially on STAT 270, that contains

the largest number of students with the appropriate

talents. Our two major service courses. STAT 201

(for life science students) and STAT

203

(for social

science students) have also attracted students with

strong quantitative skills to the discipline. We shall

increase our emphasis on assigning the best possible

instructors to these courses, shall provide further,

more regularized, opportunities for students in these

courses to learn about advantages of further

education in the discipline, and shall invite the

strongest students in these classes to attend a

presentation on career options in statistics with a

reception to follow.

• We have also been discussing with Mathematics

opportunities to participate in high-school events

such as their Math Camps, and to collaborate with

them in developing on-campus events for current

undergraduates such as the one described above.

• We do not intend to focus solely on recruiting

students into our majors and honours programs. We

intend also to encourage students who are already

committed to programs in related disciplines to

consider supplementing their education with

enhanced statistical components beyond the

common one or two semesters of coursework in

applied techniques. We have recruited occasional

students at such a level in the past, with remarkably

positive benefits for both these students and their

classmates. We shall also continue to promote such

Opportunities in our above-described recruiting

efforts.

6 We also recognize the value of interdisciplinary

programs in providing attractive opportunities for

S

---

statistical theor, and aims to develop a formal

proposal in the coming academic year. In addition.

we are discussing with Mathematics wa

y

s to

coordinate the teaching of probabilit

y with the

overall objective of maximizing learning

opportunities for both under

g

raduate and graduate

students. We also recognize a need for further

courses

at

the M. Sc. level. We intend to address this

deficiency in large part through new pro

g rams in

biostatistics and actuarial science (Recommenda-

tions S and 8).

Recommendation 4: It is recommended that the

Department review its requirements for the Masters

degree and in particular the expectations for the

itlosters thesis to he sure that the

y

are realistic and

facilitate timely completion of degrees. The

Department may also wish to consider an option of a

course based Masters degree, which is becoming

increasingly common in Statistics programs.

Response: The Department is committed to keeping

completion times low.

A

case by case re-evaluation

of the students cited in the self-study document

shows that after discounting semesters on leave and

co-op semesters, the mean completion time drops to

6.5 semesters with 19 of 35 complting in 6 or fewer

semesters. Among the remaining students, all but

3

took 8 or fewer. Factors leading to longer stays in

the program include part-time status, taking many

courses in a second discipline while pursuing our

degree, and informal work terms. The revised

distribution of completion times is shown on the

graph on the following page.

We intend to take steps to ensure that M. Sc.

projects are kept manageable. In particular, we shall

review the scope of completed projects each year to

ensure that we are not making excessive demands on

the students. We aim to set projects for students that

they should be able to complete in one semester.

We regard the projects as very important

components of our program since they focus on

communication skills, both written and oral. We

therefore believe that a course-work-only option

would weaken the preparation of our graduates for

the work force and are reluctant to move in that

direction. We note, however, that for students with

English as a second language, the process of project

writing can take more time, potentially delaying

graduation.

swdents to further their statistical training in the

S

context of specific, interdisciplinar

y

field. To this

end. we plan to continue contributin

g to the

Mana g

ement and Systems Science and

Environmental Science pro

g

rams. and to develop a

role in similar. emergin

g

programs at the Surrey

Campus.

• Furthermore, even though the actuarial science

pro g

rams cun-entiv attract outstandin

g

students

without proactive recruiting efforts, we would use

these opportunities to promote these programs as

well.

Recomniendaijo,, 2: Ii is recommended that the

Departments of Statistics and A ctuarial Science and

Mathematics and the Business School consider

appointing a Director of the MSSC program for a

four or five rear term.

Response:

We agree that the program would benefit

from a longer-term commitment from a formal

Director than the existing job-sharing arrangement

with the Department of Mathematics permits. In

•

addition, we believe that it is time that the program

be reviewed in light of recent curriculum.

developments in the contributing disciplines,

particularly includin

g

programming initiatives at the

Surrey campus. We propose to open discussions

with the other academic units contributing to the

MSSC program. There are five such units across

four faculties. We also need to conduct this review

while simultaneously developing plans for our own

department's potential role at the Surrey campus

(Recommendation

15).

Although it is not possible to

state a firm completion date for such a complex task,

we would like to complete the review process in

time for the autumn 2006 deadline for calendar

revisions.

RecoiwnendatIon 3: It is recommended that the

Deoarnnent review its graduate offt'rings in the

theor y of statistics with

-

a view to adding at least one

additional three credit course at the Ph. D. level in

this area. The Department should also discuss with

!vfathe,nattcs how best to provide some course

availabilit y

in advanced probability theory.

•

Response:

The Department stron

g

l y

endorses the

creation of an additional, doctoral-level course in

---

We also anticipate that planned expansion ot' our

graduate

course offerings will enhance students

opportunities to compIee their course requirements

without

delay.

Recommendation 5: 11 is

recwn,nended that the

Department proceed with its plans to develop a

cfaste,-s Program in Biosiatistics. In so doing, it will

be important to develop opportunities fbr student

practicums and funded research assistantships in

application areas.

Response:

We intend to develop this program as an

option within our existing M. Sc. program. (Please

see Recommendation 6.) We shall begin assembling

a proposal in the fall of 2005, with an expected date

for a completed proposal to be forwarded for

consideration by the Faculty of Science Graduatc

Studies Committee of April, 2006.

Recommendation 6:

The

pros and cons of

developing a separate degree program in

Biostatistics should he care

full y

weighed. There ,nay

be advantage in maintaining the single degree

pro grani with

an

option in Biostatistics.

Response:

The Department has considered this issue

and agrees with the implied recommendation that we

create an option in Biostatistics within the existing

degree program.

Recommendation 7: The Departmnent should

consider the possibilit y

of introducing an option in

Biostatistics within the Statistics Ph. D. program.

Response:

The Department will consider this option

as it proceeds to develop the M. Sc. option in

biostatistics and other graduate programming

changes. Under current regulations, we foresee a

ver y

limited impact to such an option at the Ph. D.

level where students are already strongly focused on

their dissertation research in a tightly focused area.

Recommendation 8: It is recommended that a

Masters Program in A ctuarial Science be a longer-

term project to be lid/v mounted in three or four

y

ears time. If the opportunit

y should arise, an

additional appointment in ,4ctzczrja1 Science would

S'

the critical mass to make such a program

successful.

A shorter-term priority should be the

design of actuarial courses

of

in the existing

graduateprogram .

Response: The Department agrees with the

recommendation. In the short term, graduate

students interested in Actuarial Science will register

in the Statistics program. The requirements of the

.

---

related fields: and possibly a term project. In the

fuwre, most existin

g

actuarial courses are expected

to he offered ever y

year and new elective courses in

the areas of propert

y

-casualt

y insurance.

demograph y

and pension mathematics should be

offered. Consequentl

y

, this Diploma could be

completed

in

four or five semesters without

additional course offerings.

Reco

p

mnejzdi,jjo,, 10: The Department should

cont!z

tic

to

develop its p/arts to contribute to the new

curriculum requirements Similar/v the University

should consider the value of a general education in

statistical literac

y

for all its undergraduates.

Response: The Department agrees with this

recommendation, including the proposal that the

University consider the value of a general education

in statistical literacy for all undergraduates.

Specifically, we propose to press ahead with

preliminary revisions to STAT 101, and with the

creation of a further course for consideration for

approval as both a designated breadth and

quantitative course. To date, we have received

breadth designation for STAT 100, and also for

EVSC 200 that is currently taught by R. Routledge

in our department.

Recommendation 11: It

is recommended that the

central administration allocate sonic base budget to

the

SCS

[Statistical Consulting Service] to increase

faculrv involvement in the area and to make the

service more accessible to researchers in all areas.

Response: The Department welcomes this

recommendation as it represents a conception of

what we feel the Consulting Service's role within

the academic community should be. We believe that

SFU would benefit substantially from resources that

would enable us to engage more freely in

consultations with researchers in other disciplines.

Our current, self-funding model does not foster

collaborative, applied research and education in

applied statistics as effectively as do other models.

For example, at the University of Waterloo, a faculty

member with statistical consulting background and

an interest in applications-based research was hired

to run their consulting service in lieu of standard

teaching requirements. Other models drawing more

broadly from the pool of expertise within the

pro g

ram will be modified to include

some

actuarial

courses and possibk

COUNCS

in other related fields.

The students cill he encoura

g

ed to

.

%^

ork on the

interface between Statistics and Actuarial Science or

the one between Finance and .ActLlaria! Science. The

project will be based on an actuarial problem.

Starting in the academic year 2005-2006, three

graduate courses in Actuarial Science will be offered

on a re

g

ular basis. The topics identified as most

appropriate at this time (and suggested course titles)

are: I) an introduction to financial and insurance risk

management (Stochastic analysis of insurance

portfolio); 2) an advanced course on risk theory and

ruin probabilities (Advanced actuarial models); and

3)

an advanced course on claim modeling or

credibilit

y

theory ,

(title to be decided). These courses

are intended

to

be offered as Selected Topics

(ST'AT-890) until it becomes appropriate to add

them as regular courses in our graduate program.

The Department has recently joined a group of

9

other universities in an application for funding

to

mount an EC/Canada cooperation program in higher

education and training in Actuarial Science. If

funded, this project could be extremely beneficial to

our graduate students.

In three to four years, the Department hopes to add

at least one more faculty member in Actuarial

Science and mount a complete Masters program in

this area.

Recommendation 9: It is recommended

that

the

Department and A ctuarial Science Group consider

the possibility

of

offering a post degree diploma in

.-l.iz sürial Science with appropriate sharing of

proceeds between the University and the

Deparrnz ent.

Response: The Department regularly receives

enquiries about the possibilities of stud

y ing in

Actuarial Science from local and international

students who have already completed an

undergraduate degree. A Post Baccalaureate

Diploma Program would definitely be an appropriate

way to meet the needs of this clientele.

The Department is considering replacing the existing

Certificate Program by a Post Baccalaureate

.

Diploma. The requirements would change to include

more

upper division courses in Actuarial science and

---

Department would also be desirable. With such a

level of support. our Consultin

g

Service could

expand the scope of its research-level consultation

work, provide more thorough assistance in applied

statistical training of graduate students in other

departments, and develop its existing initiatives in

outreach leaching for local industr, hospitals, and

resource management agencies.

Reconz,,,endat jg n, 12: It is recommended that the

Department and Facult

y

renovate the student space

to make it

tnorefiozctional

and

hospitable.

Response: We a

g

ree. Indeed during the Site visit, the

room was in particular turmoil in that an alarm

system was being installed in preparation for the

deliver y

of new computers. We have also replaced

the oldest chairs, and, at the request of the students,

are investigating replacing the rest of the chairs and

purchasing a rornd table for holding discussions,

reference books, and other amenities.

The computers are also now in place, though not at

individual student desks as recommended by the

Review Committee

:

. These desks are too small to

support even a flat screen monitor while still leaving

enough usable workspace for other tasks. We have

asked Facilities Management to provide us with an

estimate of the number of workstations that could be

installed in this room with sufficient workspace to

accommodate a computer, and the cost of the

associated renovations. We estimate that this would

likely result in a net loss of four workstations. If this

is indeed the case, then we shall not be able to make

these substantive improvements until we obtain

access to further space.

T

'deed, e'-en

• .vthcut these desirable renovations, we

have a chronic space shortage. Furthermore,

althou g

h it is particularly acute for our graduate

students, we are now pressed to the limit for faculty

offices, and have no space available to position a

workstation for the further secretarial position that

the committee has recommended. Nonetheless, we

recognize the chronic space shortage at the Burnaby

campus, and greatly appreciate the efforts of the

Dean of Science to help to alleviate this problem.

V'e look forward to whatever improvements can be

provided.

RecoInmeizLlaIioJz 13: Jr is reco,,:,, nfeJ thor the

Deparineizt.

perna .

.s

with

.tfo:he,;,a: j

cs, develop a

plan Jor lunger [Cr/u

:On1p1/tii:

support.

Response: Both departments propose to construct

this plan in conjunction with the new systems

administrator ho will be emplo

y

ed on an interim

basis as soon as we complete the recruiting process.

Recommendation 14: Jr is

recommended

that the

Department chaiige' its teaching load from four to

three courses per year for research active Jdculrv.

This chan ge

would also entail a review of

CUF....ent

policies for granting teaching credits in recognition

of other duties.

Response: We ver

y

much welcome this

recommendation, and shall begin implementing it in

the 2006-7 academic year. We have already begun

the review of our current policy for granting

teachin g

credit in recognition of other duties. In a

unanimous vote, the Department has accepted the

principle that reductions to the expected teaching

load of three courses per year for research-active

faculty members with regular appointments be made

only for the following: (i) major administrative tasks

such as chairing the departmental undergraduate and

graduate studies committees, and (ii) high-profile

external appointments of clear benefit to the

university such as serving on NSERC grant selection

committees or editing major academic journals.

We anticipate a phase-in period for this change.

Many of our faculty members are currently on

externally funded, modified contracts with reduced

teaching responsibilities. The resources needed to

implement this recommendation will become fully

available only as these modified contracts roll over

iiLo

regular appointments with full teaching

responsibilities. Hence, if we are not to rely too

heavily on sessional appointments, we shall have to

implement this recommendation incrementally. We

would clearly prefer to implement this

recommendation as soon as possible, but are

prepared to delay full implementation if necessary to

avoid compromising the integrity of our teaching

responsibilities.

Recommendation 15: The Department should

consider how

it can

best and

most

effectivel y be

involved in

the

Surre y

initiative.

If

if

is

to he is

im

.

olved in such initiatives as the program in

10.

---

With the Facult

y

of Health Sciences. We have

benefited tremendousl y

from Charmaine Dean's

pivotal role in the early deve!o

p

ment of the Faculty

as Associate Dean. We value

g reatl

y

the joint

appointment of Dr. Leilei Zen

g

between the two

units, and have recommended that another potential

member of the Faculty of Health Sciences be

appointed as an Associate Member of our

department. Carl Schwarz has also been chairing

their Tenure and Promotions Committee. In

addition, Jinko Graham, Joan Hu, Richard Lockhart.

and Brad McNenev have associate memberships in

Health Sciences, and Richard Lockhart has been

seconded to chair their Graduate Studies Committee

for the 2005-6 academic year.

We

intend to continue

making active use of these links to build a strong,

cooperative relationship that will be of benefit to

both units and the uriversitv as a whole.

Ci'.il..)1lS /

Sear..'Ii

01

kita

iJi;l!.'flg.

t/k'fl ii iceds

TO

clan pprorrare anpo!ntmcn[s and u'as to

W

en!egrai.

such appnh;nnw;its

'.iith the Depnrtnwnt on

the Bur'zahv campus.

Response: We believe that this is the most

important challenge facing the department in the

immediate future. It is essential that establish a

strong presence for our ke

y

stone discipline at the

new campus. Our immediate goal is to work with the

Department of Mathematics to support their new

program in operations research, and to provide

whatever introductory-level teaching is required for

other pro

g rams. We believe that we must do more.

01 primar

y concern, is the need to build a strong

group of faculty members who can develop a full

suite of educational and research activities. In

addition, we believe that. with strategic hiring, we

can build a group that

will

complement the research

interests of others at both the Surrey and Burnaby

campuses in statistics and related disciplines. In

particular, we foresee involvement in the operations

research program as a natural opportunity to

au g

ment our expertise in industrial statistics and, as

recommended by the review committee in a more

general context, in probability theory.

We are also encouraged by the favourable initial

response to our tentative proposal for a program in

data mining. We intend to give this particularly

careful consideration. We propose to discuss

detailed plans with Mathematics in the corning

months with the aim of generating a hiring plan

whose first phase would begin with advertisements

for one or two facuity positions in the fall of 2005,

and to develop a further plan for broader academic

consideration over the coming academic year.

Other Salient issues:

I. Interaction with Faculty

of

Health Sciences: In

Section 2 of the report, the committee refers to

expertise in the department in biostatistics as "very

valuable in

the k'velopinent

of

the new Facult y

of

Health Sciences."

.

Response: We concur that we have an important

role to pla

y

. The Department is indeed unanimous in

its concern that e maintain close, collaborative ties

.

2.

Supporl Staff: In Section 10 of the report, the

committee states. On/v two secretaries in a

Department of this si:e seems a low level of staffing

and we recommend that it be reviewed.)

Response: We appreciate this recognition of a

pressing need, and also the expression of support

from the Dean of Science, and look forward to a

timely solution. Nonetheless, we cannot implement

this recommendation until we find adequate space in

which to create a work station for the new appointee.

3.

International Programininq: In Section 10, the

report raises the potential f or "an international

program, especiall

y

in actuarial science where

excellent st!cients ,",iuld be found in Asia and

possibly even in the US.

Response: In due time, the Post Baccalaureate

Diploma and the graduate courses in Actuarial

Science could be used to offer an international

program in this field.

1/.

---

EXTERNAL REVIEW

THE DEPARTMENT OF STATISTICS AND ACTUARIAL SCIENCE

SIMON FRASER UNIVERSITY

April 20, 2005

Review team: Loveday Conquest, University of Washington, Seattle

Jose Garrido, Concordia University, Montreal

Jack Kalbfleisch (Chair), University of Michi

g an, Ann Arbor

Alton Harestad, (Internal Member), Simon Fraser University

Dates of Site Visit: March 2, 3 and 4, 2005

Report prepared by: L. Conquest, J. Garrido and J. Kalbfleisch

L

.

S

/?.

---

EXTERNAL REVIEW OF

THE

DEPARTMENT OF STATISTICS AND

ACTUARIAL SCIENCE AT SIMON FRASER UNIVERSITY

1. INTRODUCTION

The review took place on March 2, 3 and 4, 2005. This is the first review since the

Department of Statistics and Actuarial Science separated from the Department of

Mathematics and Statistics in 2001. The review team consisted of Loveday Conquest

(University of Washington). Jose Garrido (Concordia University) and Jack Kalbfleisch

(University of Michigan) as external members. Dr. Alton Harestad served as an internal

member of the Committee. The Committee had meetings with many individuals over the

three days, and we found the meetings to be very informative and interesting; we thank

all those involved for their full and frank discussions with the committee. We found the

self study document that the Department prepared to be very helpful to us. It was

remarkably objective and provided an excellent overview of the Department's strengths

and challenges as well as a clear picture of the environment in which it operates.

The Department of Statistics and Actuarial Science was formed in 2001 and over the past

four years has grown from a small group of only 7 or 8 to a group that will number

almost 20 next year. It has enjoyed exceptionally strong support from the Universit

y

and

from the Faculty of Science. It has made many excellent appointments and has built a

strong research and teaching unit. The Department has been exceptionally innovative in

taking advantage of opportunities for external funding in its recruiting efforts. It has

obtained support for three NSERC University Faculty Awards, a Tier 2 Canada Research

Chair, and most recently has made an excellent senior appointment to SFU's Shrum

Chair. This record of externally funded appointments is outstanding in the national

context of similar departments, and perhaps of any departments.

We were impressed that the Department has managed this period of remarkable growth

very well. It is very collegial and all members work well together. It has developed an

excellent core of younger faculty members who are well assimilated into all aspects of

the Department's activities, and are very involved in all aspects of its work..

With a period of such remarkable growth and innovation, it is always a challenge to

balance the need for consolidation that naturally arises against :hc development of new

initiatives and growth opportunities. It is clearly important that the Department continue

to be innovative and look for new opportunities. It is also important however that there

be the time for the Department to consolidate its existing strengths and fully develop its

current activities. The Department has shown excellent judgment and taste in the past and

we expect it to continue to be a growing force in statistics and actuarial science in Canada

and beyond, as well as a leading department at SFU.

2. AN OVERVIEW OF RESEARCH

As mentioned in the Introduction, the Department has experienced tremendous growth

•

since its inception, in 2001. It has made exceptionally good appointments and it is

currently on a very steep trajectory; it is quickly becoming one of the very best

departments in Canada and with a growing international reputation.

/3.

---

The Department has good coverage of the areas of statistics and actuarial science. There

are very fine researchers working in the areas of experimental desi g

n, sampling theory.

Ba y

esian methods, the statistics of bioinfbrmatics and genetics. statistical education and

the theory of statistics. The individual and collective scholarship is impressive as is

indicated by the very substantial grant and contract support the Department enjoys.

The Department has developed an outstanding group in Biostatistics. which includes a

team of accomplished senior and mid-career researchers as well as a number of excellent

younger faculty. Its focus includes medical and health research, genetics and

bioinformatics and as well as more traditional areas of biometrics including ecological

and environmental applications. This expertise will be very valuable in the development

of the new Facult

y

of Health Sciences, and there are current plans for new (possibly joint)

appointments in Biostatistics associated with this initiative. There is great potential for

further development of this group.

The Department also has a smaller but excellent group in Industrial Statistics. This is a

second major application area and a promising one for further development. It is notable

that the journal

Technometrics is

currently edited in the Department.

With a third new appointment thisyear. the actuarial group will have rebuilt a very good

base on

which it can consolidate its present activities and take on new initiatives. The

current research expertise, in Insurance Portfolio and Risk Theory, provides a good

mixture of applied and theoretical research. Further development of the Actuarial Science

group should be a priority in the Department and the Faculty.

The Department has lots of strength in theory in various areas, but it has a definite

applied orientation. *This has great advantage and the Department is well positioned to

respond to current high priority areas. The theory of statistics and mathematical statistics

were traditional strengths of the Department, and there is still an excellent core in this

area If the opportunity arose, an additional appointment in statistical theory would be

useful. Likewise, expertise in the theory of probability and stochastic processes would

strengthen the actuarial group and help them expand into mathematical finance.

The Department is outward looking for significant parts of its research, teaching and

service activities. It has created links with many parts of the campus. Its collaborations

extend to other universities, the community, government, research centers and industry.

3.

UNDERGRADUATE PROGRAMS

The number of undergraduate majors in the combined programs has tripled since the mid-

1990s, a sign of substantial growth. Current students commented on the "great teaching"

and "great faculty 2ttitude" in the Department of Statistics and Actuarial Science (SAS).

Undergraduate statistics and actuarial science students generally have a good feeling

about the Department; they find that faculty are accessible, helpful and genuinely care

about their learning.

3.1

The Actuarial Science (AS) program is unique on the west coast of North America;

the undergraduate AS program draws very strong students. The career pathway of

actuarial sciences is well known. and students tend to self-select due to the rigorous

nature of coursework associated with the program. Many of the top performing students

---

in the Department are in AS. The Department is a good place to house such a pro

g

ram. as

there is a natural alliance with topics of statistical theory and methods. Re

g

arding the

business/finance courses associated with this program. AS students reported no

difficulties getting into those classes.

Since the previous external review, the Department has managed to reconstitute its

actuarial group. The new appointment, scheduled for this year, should provide the critical

mass and stability required to strengthen the existing Bachelor and Certificate programs.

In that respect. the Department can consider more frequent offerings of advanced

undergraduate actuarial courses, as well as an expansion into the areas of Property-

Casualty Insurance and Pension Mathematics.

The AS group is now well positioned to expand some of its undergraduate activities and

to be entrepreneurial about it. if it wishes. For instance, there is potential for an actuarial

post-Baccalaureate Diploma for students who already have a degree in another discipline

(e.g. mathematics, business) but are interested in a career change to Actuarial Sciences.

(Further discussion of this possibility can be found under Graduate Programs below.)

AS alumni were very satisfied with their undergraduate program and particularly with the

co-op experience. There was some concern voiced from AS students and alumni about

competing for co-op positions with students from the older, well-established AS

programs in the east, particularly with Waterloo and Manitoba. However, as a result of

successful co-op placements of AS students from SFU in firms in eastern Canada, it

seems that SFU opportunities are being established and growing. Increasingly, SFU's

prog

ram in Actuarial Sciences is recognized as one that is growing and producing good

W

sttudeud ' nts.

3.2 The Statistics part of SAS as an undergraduate major has experienced less success in

attracting strong students (a phenomenon assuredly not confined to SFU). In large

measure, this is because the career potential of a Statistics major is less clear to the

typical undergraduate, and it is consequently harder to attract strong students into the

majors programs. Some members of the department expressed some concern regarding

the mathematical strength of some of the Statistics majors and the admission

requirements into the major are currently being reviewed. In order to attract strong

students into its undergraduate statistics programs, the department is encouraged to

develop a recruitment plan. The committee was told that awards are given to the best

students in STAT 270; those students form a natural cohort from which to recruit

Statistics majors. The Statistical Society of Canada has recently developed an

accreditation process which could help to lend the same kind of certification cachet" to a

Statistics degree that a degree in Actuarial Sciences now holds. If students can be made

aware of the many career opportunities to which a Statistics degree can lead, this may

help attract strong-students in the same way that the AS major does. The faculty at SFU

has the advantage of having many young members who are doing interestin g work and

who may be able to serve as role models for undergraduates. Certainly areas of statistics

like biostatistics. environmetrics, statistical genetics, bioinforrnatics and industrial

statistics have the potential to be drawing cards.

Recommendation 1. It is recommended that the Department develop a plan for

recruitment of strong quantitative undergraduates into the statistics major program.

p

It.

---

Co-op placements for Statistics majors are somewhat more difficult to find than are

placements in AS. A point to note is that placements with federal a

g encies (e.g.

Environment Canada. Fisheries & Oceans) are restricted to Canadian citizens. In

addition. students whose first language is not English may have additional

communication barriers to overcome. Nonetheless, suitable co-op placements are being

found for Statistics majors. The faculty may be able to play an increased role here in

helping to identif

y

potential employers through collaborative contacts, and perhaps

through alumni.

3.3

Management and Systems Sciences (MSSC) appears to be a good program that

combines statistics, mathematics, computer science and business. The committee heard

relatively few comments on MSSC during the site visit and wonders if its potential is

being exploited to the fullest extent possible. It was noted that administrative

responsibilities for the MSSC program alternate between Statistics and Mathematics on a

biennial basis. Thus, neither department claims real ownership of MSSC. The committee

feels that the full potential of this program may be better realized with a more permanent

director. Such a person could more effectively exercise leadership, develop the program's

potential, and create a suitable home and identity for the students.

Recommendation 2. It is recommended that the Departments of Statistics and A ctuarial

Science and Mathematics and the Business School consider appointing a Director of the

MSSC program for afour or five year term.

3.4

Course Offerings and Recommendations for Coursework. The committee heard from

SAS majors the desire to have certain types of undergraduate courses offered

(particularly at the advanced levels—time series analysis was mentioned), or offered

more often than they are at present. An advanced course that is offered in alternate years

can leave a student with undesirable choices: take the course before having satisfied

prerequisites; wait to take the course two years later (thereby delaying graduation); or try

to learn the material as a reading course. If tightening up on admission requirements to

the major can be thought of a "weeding" process, then the undergraduate curriculum also

may need some "fertilizing" through upper level courses being available to students each

year. Having two new hires on board (one in AS, one in Statistics) will presumably help

address this issue.

Some statistics alumni suggested that the undei i4raduate program should include more

mathematics requirements (e.g., courses in real analysis) for statistics majors. For those

intending to pursue graduate school in a theoretical statistics program, more math

electives are recommended. This may best be handled as an advisin

g

issue for the

Department, rather than ending up with an overly restrictive major.

4. GRADUATE PROGRAMS

4. 1 General.

The graduate pro

g

rams offered by the Department are very healthy. There

has been a substantial growth this year to about

35

graduate students from a previous

baseline of about 24 students. Graduates of both the Masters and the Ph.D. programs are

I',.

---

able to find very good jobs in areas related to their specialties. Especiall

y

with the

. - increase in faculty complement. a further increase in graduate enrollment over the next

three or four years would seem appropriate.

We heard from the graduate students that the

y

would like to have more regular offerints

of the more senior courses. Some students felt that, especiall y

in the masters program,

there was sometimes not the opportunit y

to take some courses of interest. In the first

year. the y

lacked the prerequisites and in the second year. the course was not offered.

One notable omission in the curriculum was a second Ph.D. level course in the theory of

statistics, and the committee feels that it is very important to remedy that deficiency. It

was also noted that there were no advanced courses in probability theory. It would be

worthwhile to explore with Mathematics how some course work in this important area

could be included in the graduate program in both Departments.

Recommendation 3. It is recommended that the Department should review its graduate

offerings in the theory of statistics with a view to adding at least one additional course at

the Ph.D. level in this area. The Department should also discuss with Mathematics how

best to provide some course availability in advanced probability theory.

Completion times for the graduate degrees are generally good, but there was a rather long

tail in the historical data on time to complete the Masters degree. Some of this may be

due to co-op work terms or interrupted study, but it also seems clear that some students

are taking a long time to complete the degree. Especially with a thesis based masters

degree, there is danger that the expectations may sometimes be too great. The committee

.

feels that a masters program should not normally take in excess of six academic terms for

completion and recommends that the Department review the requirements of its masters

program and takes measures to assure that most students would complete in a two-year

time frame. Completion times for the Ph.D. degree seemed to be more or less at an

average level. Again, there were some very long completion times and the Department

should look to see how best to reduce these.

Recommendation 4. It is recommended that the Department review its requirements for

the Masters degree and in particular the expectations for the Masters thesis to be sure

that they are realistic and facilitate timely completion of degrees. The Department may

also wish to consider an option for a course based Masters degree. which is becoming

increasingly common in Statistics programs.

4.2 Proposed Masters program in Biostatistics.

Biostatistics is a high demand area that

has grown rapidly over the past two or three decades. There are many strong programs in

this area in Canada and the US, and graduates at both the masters and Ph.D. levels are in

high demand in Universities, industry, government and research enterprises. In the light

of this and the strengths of the Department, the proposed masters program in Biostatistics

seems very natural. The faculty contains an excellent core of individuals working in

various areas of biostatistics, biometrics and environmetrics and certainly has the internal

expertise to mount the program; the creation of the new Faculty of Health Sciences

makes this an even more natural initiative. It is currently planned to hire two

biostatisticians connected with that initiative and this will add further to the expertise.

One disadvantage is the lack of a Medical School at SFU; interactions with the Medical

School provide a very natural source of student research support and practical experience

p..

---

for many programs in Biostatistics. Nonetheless. there are examples in both Canada and

the US of strong biostatistical programs at universities that have no medical school. The

biostatistics group at SFU has a number of good sources of practicum experience throul

its connections with the British Columbia Cancer Agenc

y , local hospitals. Department of

Fisheries and Oceans, the Ministry of Forests, and the development of the new Faculty of

Health Sciences should help to provide additional opportunities. Further development of

these connections and seeking new opportunities seems especially important.

It was not clear to the committee that there would be particular advantage in creating a

new degree program in Biostatistics. It appeared to us that most of the advantage could

be obtained by introducing a Biostatistics option within the existing masters program.

This latter approach could be undertaken with approvals within the University only. Well

developed and packaged, this would enable the Department to market a Biostatistics

program to prospective students and to further build its teaching and research activities in

this area. It is possible that we are missing the advantages of developing a new program

with Ministerial approval; it is clear that a strong proposal to do so could be developed

and, if this is the direction undertaken, we would expect the application to be strong and

to gain approval.

In our discussions with those in Biostatistics, the possibility of a Biostatistics option at

both the Masters and the Ph.D. level was raised. Many students are seeking both levels of

training, and it may be easier to recruit strong students into the Masters program if they

knew there was a Ph.D. option to follow. The introduction of a Ph.D. program in this

area would also recognize the substantial faculty expertise and the capability that the

R

D

e

e

c

pa

om

rtm

m

e

e

nt

nda

ha

ti

s

on

in

5:

re

I

s

t

e

i

a

s

r

r

c

e

h

c

s

om

upe

m

r

e

vi

nde

sion

d t

i

ha

n

t

thi

the

s a

D

re

e

a

pa

.

rtm ent proceed w ith its plans to

9

develop a Masters Program in Biostatistics. In so doing, it will be important to develop

opportunities f or student praclicums and f unded research assistantships in application

areas.

Recommendation 6: The pros and cons of developing a separate degree program in

Biostatistics should be carefully weighed. There may be advantage in maintaining the

single degree program with an option in Bioslatistics.

Recommendation 7: The Department should consider the possibility of introducing a

op/ion

in

Biostatistics i ithin the Statistics Ph.D. program.

4.3 Proposed Masters Program in Actuarial Science. The development of a proposed

Masters program in Actuarial Science seems a natural project for the Department to

undertake. This project has great potential, and there is certainly a market for a program

in this area.

At the moment, however, the Department has only two faculty members in Actuarial

Science and is currently seeking to add a third member at the Assistant Professor level.

There is time needed for members of this group to develop further their research

programs and to consolidate the undergraduate program.. It would also seem important to

have some time to build additional link's with the business school and other related areas.

The committee felt that some time and the addition of at least one more faculty member

in this area would be important before mounting a graduate program.

/,.

---

At present, there is the opportunity for students to work on the interface between statistics

and actuarial science at both the Masters and the Ph.D. level, and to develop their thesis

work in an area of Actuarial Science. This seems to work reasonably well and provides

an opportunity to develop additional strength in graduate studies in Actuarial Science.

The first step toward a separate graduate program in Actuarial Science ma

y

be to develop

one or two graduate courses to be offered on a regular basis.

Recommendation 8: It is recommended that a Masters Program in A ctuarial Science be a

longer terin project to be mounted in three or four years time. I/the opportunit

y

should

arise, an additional appointment in A ctuarial Science would give the critical mass to

make such a program successful. A shorter term priority should be the design of

actuarial courses ofjè red in the existing graduate program.

4.4 A Post Degree Diploma in Actuarial Science. In discussion with the actuarial science

group, the idea of a Post Degree Diploma arose. This would provide an opportunity for

graduates in other areas to take a one or two year program in Actuarial Science to

develop their knowledge and skills in that area. Many individuals with various

undergraduate degrees see great career opportunities in Actuarial Science and the market

for such a diploma could be substantial. This program could utilize various

undergraduate courses and involve rather modest additional course offerings along with a

project. Such a program could be financially beneficial.

Recommendation 9: It is recommended that the Department and the A ctuarial Science

Group consider the possibility of offering a post degree diploma in A ctuarial Science

with appropriate sharing ofproceeds between the University and the Department.

5. SERVICE TEACHING

Both undergraduate and graduate service teaching are given high priority in the

Department. Faculty in SAS see this as an important responsibility. An introduction to

statistical thinking, regardless of one's major, is a very important ingredient of a general

education. Tutorial support for the service-teaching role is provided in a somewhat

unusual manner. The Department operates a drop-in facility or Workshop where students

can come to get help with lecture material and assignments. The facility has a number of

computers that run statistical software packages and is staffed by a Director and several

Teaching Assistants, who also grade assignments for the various couses. It was our

impression that this Workshop works well in providing a resource for students to seek

help.

The Department has a key role to play in SFU's new initiatives for the Quantitative,

Writing, and Breadth requirements as part of the undergraduate curriculum. It is natural

to develop quantitative skills in the context of statistical applications; in addition, there

may be room for certain statistics courses for non-majors to fulfill part of the breadth

requirement. The Department is revising the STAT 100 course into a 'Statistics for

Everyday Life" course that would meet the quantitative requirement for humanities

majors. There is also the intention to review and revise STAT 201 (Statistics for

S

Biologists) and STAT

203

(Statistics for Social Scientists) by the spring of 2007. The

Department is already offering one service course (STAT 270) at the Surrey campus,

1'?

---

with plans to add at least one more. As the Surrey campus grows. the Department plans to

continue to ascertain the statistics needs of Surrey students.

Recommendation 10: The Depcirrnent should

CO/IuiflhEC

to develop its plans to contribute

to the new curriculum requireuzents. Similarly, the University

should cons:der the value

of a general education in statistical literacy for all its undergraduates.

6. CONSULTING SERVICE

The Statistical Consulting Service (SCS) serves SFU in teaching. research, and service

roles. In a previous review, it was recommended that the SCS become self-supporting by

developing appropriate revenue streams. SCS has accomplished that, primarily through

taking on contracts for statistical consulting services from organizations outside the

university. The current level of outside contracting along with some Departmental funds

for teaching a graduate consulting course is sufficient to fund the salary and benefits of

the director position. Sometimes a company or agency in need of a statistical consulting

service will contact a Departmental faculty member, and then be referred to the SCS.

Extension of thSCS into the area of actuarial sciences is a distinct possibility with the

growth of that core group in the department. Such an extension would benefit graduate

students with a particular interest in actuarial science who currently have statistical

consulting course requirements.

Graduate students in Statistics are required to take a course in statistical consulting,

STAT 811-812, which is taught by the SCS director and supported by Departmental

funds. Engaging in statistical consulting on real problems and interacting with real clients

provides excellent training. Students learn how to work with clients, and to progress from

initially broad inquiries to more focused questions that can be answered with suitable

statistical design and analysis. Skills gained in the process include thinking on one's feet,

communicating statistical concepts to non-statisticians, and writing statistical

recommendations for clients in language they can understand.

The SCS also makes a substantial contribution to scientific research at SFU. From the

point of view of statistics, there is always the possibility that a consulting project will

develop into a co-authored journal paper or thesis topic for a statistics student. More

broadly, however, such services can help to raise thc quality of research in a variety of

disciplines that depend on good statistical analysis and design for valid research.

There is the potential for the SCS to be a much stronger resource for SFU than it is now.

Currently only those who have access to budgeted funds can avail themselves of the SCS

for a long-term or intensive consulting arrangement. The SCS offers brief advice via

online statistical consulting (Ask Dr. Stats); it also provides pro

bono

consulting,

necessarily for a limited time, for graduate students and faculty. Currentl y

only the

Director is funded and there is little incentive for SAS faculty to increase their

involvement with SCS activities. It is the committee's impression that the Departmetn

would like to develop this facility further.

In view of the university-wide role the SCS already plays, we recommend that the SFU

administration allocate some central funding to the SCS. This could pay for some faculty

release time to increase and regularize SAS faculty involvement in the SCS and perhaps

---

S

fund the involvement of an advanced statistics graduate student in the SCS as an

additional consultant. Involvement of Statistics faculty on a regular basis would bring

much more expertise to bear on the SCSs work and integrate the service more into the

mainstream of the Departments activities. This funding would be an investment that

would have substantial dividends in increased research quality in many disciplines.

Reconnnendation II: Iris recommended that the central administration allocate some

base budget to the SCS to increase faculty involvement in the area and to make the

service more accessible to researchers in all areas.

7. SPACE

The Department is housed in satisfactory, though certainly not elaborate, quarters. It is

becoming very crowded with very little common space and no extra offices for visitors.

for example. New appointments will add more pressures and accommodations will need

to be found. As the Department continues to grow, providing some increase in space

should be a priority of the Faculty.

There was great variation in the quality of space for graduate students which seemed to

be based on their research area and the ability of their supervisors to pay for upgraded

facilities. Committee members felt that the quality of space for many of the graduate

students was the worst that they had seen anywhere. One of the main areas houses

almost 20 students in a converted classroom. It has very old and run down workstations

S

with a few computers available for student use; we heard time and again complaints

about broken chairs and a generally inhospitable environment. No students were

occupying the space when we visited and we thought we understood wh

y

. The committee

strongly recommends that there be an investment made to convert this space into

something more attractive and more functional. This is an issue of quality of life for

students and should be quickly addressed. Nowadays, most statistics departments would

provide students with a large enough working service to have a computer on each desk.

Recommendation 12: If is recommended that the Department and Faculty renovate the

student space to make it more functional and hospitable.

7. COMPUTING

The issue of computing came up time and again in conversations that we had with

faculty. Several felt that the level of support for computing in the Department, the

School and the University was not meeting their needs. The Department currently

contracts with the Netware Support Group in the Faculty of Applied Science for

computing and software support. There was a definite feeling, however, that while this

arrangement met some of the basic needs, it could not respond adequately or in a timely

manner to specific issues as they arose.

Members of the Department were pleased that the Dean of Science has agreed to fund, at

least for one or two years, a position for a computing technician that would be shared

with Mathematics. There is a strong hope that this will provide the needed support, at

least on the shorter term. The review committee really does not have the expertise

C21.

---

needed to comment in an' detail on these issues.

it

is clear, however, that there has been

a problem in computing support and that it has, at least in some instances, interfered with

efticient use of time and resources. This issue was also raised in the 1993 External

Review (see their Recommendation 7).

There is now at least a one year period during which SAS together with the Department

of Mathematics can consider how best to use this new position. This should also provide

the time to prepare more clearly the case for this kind of computing support and to

develop a long-term plan. Some portion of the budget for computing support should come

from research funds and some from Faculty or Department operating funds

;

building a

workable model for longer term computing support should be a high priority of the

Department.

Recommendation 13.' It is recommended that the Department. perhaps with Mathematics,

develop a plan for longer term computing support along %i'ith appropriate budget models.

8. WORKLOAD TEACHING LOADS

Issues of workload, and in particular teaching loads, were raised in several interviews. At

the moment. SAS operates with a nominal four course teaching load, but with a fairly

extensive schedule of teaching releases for various internal and external service

contributions as well as some releases for graduate supervision. In statistics departments

in Canada, faculty members who are active in research and in supervising graduate

students would normally have a three course teaching load. In addition, SFU's

Department of Mathematics has changed the normal teaching load from four to three

courses per year. The committee recommends that the Department establish three courses

per year as the normal teaching load of research active faculty members.

As is the case in many departments in Canada that have been successful in attracting

UFAs, specially funded chairs, or Canada Research Chairs, there are a number of faculty

who have special arrangements for reduced teaching and service responsibilities so that

they can devote more of their time to developing their research program. Such differences

can cause some difficulties in the Department as colleagues with similar expectations on

research are given very different assignments in teaching and service. SAS seems to

have handled these differences very well, and there is a strong esprit de corps with a

general appreciation of the considerable value to the Department as a whole of these

specially funded positions. As well, the faculty members who hold these awards are very

willing to help out where they can and are undoubtedly going to be very strong

Departmental citizens in the future.

Even among those who are not on such special awards, it is relatively rare that SAS

faculty have taughi four courses in a given year. Some senior faculty members have

received release time from teaching in order for take on major administrative

responsibilities or to engage full-time in funded research. The current algorithm for

teaching credit is based on 12 credits a year and gives credit for many, but not all, major

responsibilities outside the classroom. The committee feels that it would be better for the

norm to be 9 credits per year with the Department recognizing that there are a variety of

administrative and supervisory tasks that must be collectively shouldered in an equitable

t

I 19

---

manner. Such a change would require that the algorithm for granting teaching credits be

. -

replaced by a new, perhaps less generous, set of guidelines developed through

Departmental discussions. This is-entirel

y

feasible, given the highly collegial nature of

relationships within the Department.

Recommendation 14: fits recommended that the Department change its teaching load

from four to three courses per year/br research active facult

y

. This change would also

entail a review of current policies for granting teaching credits in recognition of other

duties.

9. SURREY

The role that Statistics and Actuarial Science should play is not completely clear. It is

clear, however, that there will need to be a presence of SAS at the Surrey campus even if

only through a service role. The Department currently offers Stat 270 at Surrey, and it

seems likely that additional service courses will be needed to support various activities

there.

There are also a number of potential places where Statistics and Actuarial Science could

be substantiall

y

involved. One proposal, about which we heard relatively little, is to

mount an interdisciplinary program in data mining which would combine basic training

in statistics and computer science with more advanced work on statistical methods of

prediction and classification and computer science work on data base systems, artificial

. intelligence and machine learning. This could be a very good program and the skeleton

outline given in the self study document might be further developed to give an excellent

program for some students in Computer Science or Statistics.

A more immediate possibility arises with the proposed new program in Operations

Research (OR) that is being coordinated through the Department of Mathematics. Any

such program would absolutely need to have some fairly substantial involvement of

statistical methods as well as stochastic modeling. Certainly statistical methods to the

level of regression methodology and courses on stochastic processes and their

applications would be essential. It is important that there be discussion with Mathematics

as to how these needs will be met.

There may be a real opportunity for the Department to become involved in this area and,

at a minimum, it should provide some basic support for the initiative as it unfolds. There

may be a larger role to play in this initiative and there could be substantial value to an

appointment of someone in stochastic and statistical aspects of OR. The issue of isolation

of such an appointment at the Surrey campus is a real concern and the Department needs

to think through how it can best contribute to this University initiative.

Recommendation 15: The Department should consider how it can best and most

effectively be involved in the Surrey initiative. If it is to be involved in such initiatives as

the program in operations research or data minIng, then it needs to plan appropriate

appointments and

%I;ayS

to integrate such appointments with the Department on the

Burnaby campus.

0

0.

---

10. FOR THE

RECORD:

• The Department

staff

seemed to he operating very effectively and had a very

positive outlook. There may he a need for an increase in the staffin

g

levels in the

Department. One secretary currently provides support for the Chair as well as

graduate studies. Only two secretaries in a Department of this size seems a low

level of staffing and we recommend that it be reviewed.

• The library support and facilities appeared to be ver

y

good. We heard no concerns

about library facilities and were impressed with the information we received in

our interviews.

• The split from Mathematics seems to have been accomplished in a very amicable

way and there seems to be very

g

ood relations between the two Departments.

Such collegial relationships are not always seen when such divisions take place.

The division has been very well handled administratively with excellent

leadership on all sides.

• There are concerns expressed in the document that the Department is left to find

resources to pay benefits for faculty members who are funded from external

sources. We understand that this charge to the Department was a mistake and that

steps are being taken to correct this. It seems obvious that such charges are not in

the interest of promoting initiative at the Department level.

• The Alumni were extremely positive about the Department and their educational

experience (except for the quality of space and computing facilities).

• Leadership in the Department is deep and very strong. Several senior members

have played outstanding leadership roles in establishing the Department and in

seeking out external opportunities for development. Several of the younger

members of the Department also have great potential in this respect. The

University and the Faculty have been willing to invest heavily in backing these

leaders and this has and will continue to pay great dividends.

• The committee heard comments from several faculty members that more and

more work has been devolved to the Department and faculty level with no more,

and often fewer, resources. This is of course a common, but also justified

complaint; such additional responsibilities do increase the stress of the faculty

position, and leave less time for the main responsibilities of research and teaching.

• There would be potential for an internationalprogram, especially in actuarial

science where excellent students could be found in Asia and possibly even in the

US. With good incentives in revenue sharing from the University, this may be an

option worth investigating.

External Revieft' Committee for the Department of Statistics and Actuarial Science

---

0 -

Summary of recommendations

Recommendation I. It is recommended that the Department develop a plan for

recruitment of strong quantitative undergraduates into the statistics nuyor program.

Recommendation 2. It is recommended that the Departments

of

Statistics and A ctuarial

Science and Mathematics and the Business School consider appointing a Director of the

MSSC program for

afbur

or

five

y

ear term.

Recommendation 3. It is recommended that the Department review its graduate offerings

in the theory

of

statistics with a view to adding at least one additional three credit course

at the Ph. D. level in this area. The Department should also discuss with Mathematics

how best to provide some course availability in advanced proba bility theory.

Recomnzendation 4. It is recommended that the Department review its requirenientsjbr

the Masters degree and in particular the expectations for the Masters thesis to be sure

that they are realistic and facilitate timely completion

of

degrees. The Department may

also wish to consider an option

of a

course based Masters degree, which is becoming

increasingly common in Statistics programs.

Recommendation 5: 1/ is recommended that the Department proceed with its plans to

develop a Masters Program in Biostatistics. In so doing. it will be important to develop

.

opportunities for student practicumns andfunded research assistantships in application

areas.

Recommendation 6: The pros and cons of developing a separate degree program in

Biostatistics should be carefully weighed. There maybe advantage in maintaining the

single degree program with an option in Biostatistics.

Recommendation 7: The Department should consider the possibility

of introducing an

option in Biostatistics within the Statistics Ph.D. program.

Recommendation 8: It is recommended that a Masters Program in A ctuarial Science be a

longer-term project to be fully

mounted in three or four years' time. If the opportunity

e!

uld arise an additional appointment in .4 tuarial Science w'nild give ti:: critica

l

l mass

to make such a program successful. A shorter-term priority should be the design of

actuarial courses of

in the existing graduate program.

Recommendation 9: It is recommended that the Department and the A ctuarial Science

Group consider the possibilit

y

of offering a post degree diploma in A ctuarial Science

with appropriate sharing ofproceeds between the University and the Department.

Recommendation JO: The Department should continue to develop its plans to contribute

to the new curriculum requirements. Similarly, the University should consider the value

of a general education in statistical literacy for all its undergraduates.

Recommendation 11: It is recommended that the central administration allocate some

is

base budget to the SCS to increase faculty involvement in the area and to make the

service more accessible to researchers in all areas.

---

Recommendation 12: It is recommended that the Department and Facult

y

renovate the

student space to make it more functional and hospitable.

Re con unendation 13: It is recommended that the Department, perhaps iiiih Mathematics,

develop a plan for longer term computing support.

Recommendation 14: It is recommended that the Department change its teaching load

from four to three courses per year/6r research active faculty. This change would also

entail a review of current policies for granting teaching credits in recognition of other

ditties.

Recommendation 15: The Department should consider how it can best and most

effectivel y

be involved in the Surre

y

initiative. If it is to be involved in such initiatives as

the program in operations research or data mining, then it needs to plan appropriate

appointments and hays to integrate such appointments with the Department on the

Burnaby campus.

.