Diversities of gifts, but the same spirit

1
Diversities of gifts, but the same spirit

• Peter Green
• RSS Presidential address
• 18 June 2003

2
A discipline of diversity
Public life
Social science Science Technology Medicine
Statistics
Business and industry
3
A discipline of diversity
4
A discipline of diversity
Interaction with the rest of the world is part of
the subject itself
5

• shelter and nourishment for

statistics

• a microcosm of

6
Connection or fragmentation?
What holds us together?
7
Connection or fragmentation?
Statistics in Society

• getting the correct denominator in workforce
  statistics
• computing DNA match probabilities
• assessing clinical effectiveness
• evaluating GM crop experiments

8
Connection or fragmentation?

• Heterogeneity of discipline
• intellectual strength
• structural weakness

9
medical
industrial
core
social
official
10
medical
industrial
core
social
official
11
medical
industrial
core
social
official
12
How the discipline develops

• Promoting our strengths should be a key priority
  for the discipline and for the Society

13
How the discipline develops demands of
applications

• Public policy
• evidence-based decision making
• performance measurement
• Legal system
• scientific evaluation of evidence
• Social science
• respect for quantification
• public archives, National Statistics

14
How the discipline develops demands of
applications

• Business
• data-mining
• Technology
• uncertainty in telecomms, images
• Science
• all scales Astronomy to Genomics
• quantum level?

15
How the discipline develops opportunity of
technology

• Sensors and instrumentation
• Data-logging capacity
• Communications
• Number-crunching

• transforming quantity and quality of data
• enabling highly computer-intensive analysis

16
How the discipline develops theoretical
innovation

• Relaxation of old philosophical quarrels
• Rehabilitation of Bayesian methods
• Key role of conditional inference
• graphical modelling
• Stochastic calculus
• martingales
• Point processes

17
The role of statistical modelling

• underpinning all parts of the discipline
• the most basic tabulation or summary involves
  conceptualisation
• what can vary?
• on what scale?
• depending on what?

18
The role of statistical modelling

• Discipline in creation of methodology
• Framework
• for study of foundations
• for expressing principles
• for provision of computational tools
• Use more to communicate ideas
• break down barriers between theory and practice?

19
Structured systems

• A framework for building models, especially
  probabilistic models, for empirical data

20
(No Transcript)
21
Contingency tables
Graphical models
22
Structured systems

• Key idea -
• understand complex system
• through global model
• built from small pieces
• comprehensible
• modular
• each with only a few variables

23
Modular structure

• Basis for
• understanding the real system
• capturing important characteristics statistically
• defining appropriate methods
• computation
• inference and interpretation

24
Conditional independence

• X and Z are conditionally independent given Y if,
  knowing Y, discovering Z tells you nothing more
  about X
• X ? Z ? Y

X
Y
Z
25
Conditional independence

• as seen in data.

Does survival depend on ante-natal care?
.... what if you know the clinic?
26
Conditional independence
ante
survival
clinic
survival and clinic are dependent
and ante and clinic are dependent
but survival and ante are CI given clinic
27
(No Transcript)
28
(No Transcript)
29
A natural directed graph from genetics
A
AB
A
O
O
Mendel
30
Model for lip cancer data
regression coefficient
covariate
random spatial effects
relative risks
observed counts
31
or non-
Bayesian
32
Bayesian structured modelling

• borrowing strength
• automatically integrates out all sources of
  uncertainty
• properly accounting for variability at all levels
• including, in principle, uncertainty in model
  itself

33
Bayesian structured modelling

• borrowing strength
• automatically integrates out all sources of
  uncertainty
• for example in forensic statistics with DNA
  probe data..

34
(thanks to J Mortera)
35
(No Transcript)
36
Bayesian structured modelling

• borrowing strength
• automatically integrates out all sources of
  uncertainty
• for example in modelling complex biomedical
  systems like ion channels..

37
Ion channelmodel
model indicator
transition rates
hidden state
Hodgson and Green, Proc Roy Soc Lond A, 1999
binary signal
levels variances
data
38
model indicator
C1
C2
C3
O1
O2
transition rates
hidden state
binary signal
levels variances
data











39
Structured systems success stories include...

• Genomics bioinformatics
• DNA protein sequencing, gene
  mapping, evolutionary genetics
• Spatial statistics
• image analysis, geographical
  epidemiology
• Temporal problems
• longitudinal data, financial time series, signal
  processing

40
The methodology gap

• Subgroups develop their own ideas and jargon
• Weaker communication between than within
• Little evidence in RSS journals
• But wide use of outdated and inappropriate
  statistical techniques in some areas

41
The methodology gap

• - the pressures
• pace of working life
• ? specialisation
• ? quick approximations
• training more focussed
• both theoretical and applied

42
The methodology gap

• RSS provides something for almost every
  specialism
• but how many of us exploit that?

43
(No Transcript)
44
(No Transcript)
45
(No Transcript)
46
Making more of methodology

• Relevance to applications is the main stimulus
  and justification
• But, for the sake of the vigour of the subject
  and cross-fertilisation between applications,
  there is a vital role for generic methodology
• not mathematical statistics
• not application-specific

47
Generic methodology

• The generalised likelihood ratio test
• Fisher scoring
• The practice of fitting dose-response
  relationships by MLE
• all existed before.

48
(No Transcript)
49
Generic methodology

• . but the generalised linear model framework did
  not just unify,
• it generated new application-specific technique
• it promoted good practice generally

50
A good methodology paper might cover all of ...

• underlying philosophical principles
• mathematical development
• statistical modelling of a real process
• computational implementation
• data analysis
• model criticism
• interpretation of inference and performance

51
But is statistics research reproducible?

• Could a competent and adequately-equipped reader
  obtain equivalent results?
• No, because of lack of detail, and time
  investment required
• Further, methodological work claims more general
  applicability
• what does reproducible mean then?

52
(No Transcript)
53
Reference implementation

• code that is warranted to give the authors
  intended answers in a moderately-sized problem.
  It need not be efficient but it should be
  available to anyone and everyone

• suggests use of an open-source language like
  .

54
(No Transcript)
55
long version of paper
data
figure
instructions
paper
56
(No Transcript)
57
(No Transcript)
58
The RSSs role in closing gaps

• We need to bridge some gaps for the future health
  of the discipline
• cf. Merger with IoS in 1993

59
RSS gap-closingpublications

• Publications Review
• Merging 4 Journal series into 3
• distinguished only by technicality of content
• New Magazine from 2004
• distributed to all members

60
RSS gap-closingmeetings

• Joint meetings between Sections
• e.g. Performance monitoring
• New general conference structure
• RSS 2004 in Manchester
• connecting practice with research

61
RSS gap-closing

• Education and training
• recruitment and supply of statisticians study
• LTSN
• accrediting university courses
• short courses
• Medals and prizes

62
Only connect.
63
.. an opportunity to demonstrate the unity in
diversity that characterises our activities P
Armitage, 1983
We still suffer from too much diversity and
inadequate bridges between the various parts of
our subject C Moser
64
Only connect! That was the whole of her sermon.
Only connect the prose and the passion, and both
will be exalted....
Only connect!
E. M. Forster, Howards End, chapter 22, 1910.