2024: The FAR, Federal Acquisition Regulations, Part 31
Data4Good Conference 2018 Pro Bono OR and Statistics
1.
2. Data analysis for improving
operations and strategy
Amy Hughes, Pro Bono Manager, The OR Society
Amaka Nwagbara, Pro Bono Manager, Royal Statistical Society
3. Agenda
• Welcome
• Introduction to the Royal Statistical Society
• Introduction to the Operational Research Society
• Pro bono OR case studies
• Workshop: generating problems and responses
5. About me
• Degree in Mathematics and Statistics, as well as PhD in Applied Statistics
• Founder of Global Sports Statistics, an independent statistical consultancy
• 11 years professional experience with 7 years as a consultant
• Public engagement work includes appearances on BBC, The Guardian and various
other media outlets.
6. Royal Statistical Society
• The Society is one of the world’s leading organisations to promote the
importance of statistics and data in society.
• It is a global community of over 9,000 statisticians
and data scientists
• I have been a member since 2011.
“Our vision is a
world with data
at the heart of
understanding
and decision-
making.”
7. Statisticians for Society
• A pro bono scheme created by the Royal Statistical Society (RSS)
• It is a pilot scheme funded by the Big Lottery since 2017
• To provide data and statistical support to the third sector
• To encourage RSS members to offer their time and skills.
“I am the
current chair of
Statisticians for
Society Scoping
Committee.”
8. Criteria
The initiative aims to:
• support UK registered charities with a turnover less than £1 million
• actively match charities with RSS members at no cost. “The scheme is
open to socially
useful initiatives
of a similar size”
9. The Process – scoping phase
• A charity completes an enquiry form to request pro bono support
• The Statisticians for Society project manager contacts the charity to discuss the
project
• The project manager notifies the Scoping Committee of the request and sets up a
phone call for the appointed scoping volunteer and charity
• The scoping volunteer will then determine if a statistician can help with the project
10. Recruitment phase
• Following the call, the scoping volunteer completes a Scoping Form, providing
details of the project and skills required
• The form is sent to the charity to approve and sign
• After the signed form is returned, a call for participation email is sent to our
volunteers statisticians
• The charity is then matched with a volunteer statistician, or team of statisticians,
with the expertise for the project.
11. Project phase
• Before the project commences, the project manager will schedule a phone call
with the selected volunteer to discuss what’s involved
• The charity and volunteer will be asked to complete a Project Proposal Form,
detailing the work to be carried out
• Once the Project Proposal Form is returned, the work will commence
• The project manager will maintain contact with the volunteer and charity to
monitor how the work is progressing.
12. Since the launch…
• We have recruited more than 290 volunteer statisticians and data scientists for the
scheme
• We have begun working with 50 organisations across the UK
• So far, we have been successful in matching charities with volunteer statisticians.
13. Requests for support so far…
The areas of support commonly requested are:
• Data analysis
• Survey design
• Data collection
• Data visualisation
Charities also seek help with evaluation methods design, impact measurement,
research design and statistical guidance for future analysis.
14. Statisticians for Society
To find out more about Statisticians for Society, visit
www.rss.org.uk/Statisticians-for-Society
“Quote/pull out
box”
15. How Operational Research can
improve your organisation’s decision-
making
Facilitators: Graham Rand, Ian Seath and Amy Hughes
18. What is Operational Research?
• Sometimes known as Management Science, Operations Research, OR/MS or
MS/OR
• “The purpose of OR is to enable managers to make more effective decisions.”
• “OR is the use of scientific methods to help deal with complex problems faced by
industry, the public sector and other organisations.”
19. How does OR help managers make better
decisions?
By using ‘soft’ methods (looking at people and context) and ‘hard’ methods (mathematical
modelling and analysis) OR gives those who run organisations the power to make more
effective decisions, improve outcomes and build more productive systems based on:
• Using tried and tested decision tools and techniques
• Clearer agreement about what you’re trying to do
• More complete data
• Deeper insight into how different factors influence each other
• Consideration of all available options
• Careful predictions of outcomes and estimates of risk
20. What makes a suitable problem?
• Decision to be made
• Choices of action
• Significant effect
• not worthwhile for trivial decisions
• Suitable client
• willing to consider recommendations
• able to carry them out
“Quote/pull out
box”
21. What types of problem? Some examples:
“Quote/pull out
box”
Strategic "one-off" decisions
“Tactical" not strategic decisions
Extremely detailed issues
Providing good information
Relocating offices - to where should we move?
Aim - minimum cost, minimum travelling time for staff (important to get right and costly to
reverse)
Output – recommendations
Producing a volunteer rota
This is all about processes - not one-off decisions
Output - recommendations and possibly also a computer system for future situations
Organising appointments - matching providers and clients, at specific times
Aims – to make a good match, minimum wasted time
Output - for use by technical person – maybe the developer
Keeping track of performance
Aim– collecting measurements and using them to highlight changes
Output - providing a computer system as an aid for the client to make future decisions
22. Hard or Soft OR?
Soft OR - qualitative, subjective
• less prescriptive, more
descriptive
• help stakeholders see the issues
more clearly
• facilitation rather than
recommendation
• usually the start of any project
“Quote/pull out
box”
Hard OR - quantitative, objective
• measure what you can
• may have to ignore what you can't
• make specific recommendations
Many projects
combine the two
approaches
23. So what does an OR problem-solving approach
consist of?
“Quote/pul
l out box”
• Problem structuring, formulation (always)
• Data gathering/analysis (sometimes)
• Helping managers to understand (this may be all that's required)
• Producing computer systems (quite often)
• Most OR involves MODELLING (quantitative or qualitative)
25. OR for the Third Sector
“Quote/pul
l out box”
• Some of the problems third sector organisations are faced with:
• ‘We know we’re doing a good job – but how can we prove it?’
We felt that OR could help and set up Pro Bono OR
• ‘We have lots of different options for the future but it’s impossible to decide which to
choose in such uncertain times.’
• ‘We’re under huge pressure to do more with less, and we don’t know how we’re going to
do it.’
• ‘It’s hard to stay objective when we’re faced with such emotionally charged decisions.’
26. Pro Bono OR
Set up as a service to third sector organisations by The OR Society
‘provides some of the head to your organisation’s heart ’
Some organisations we have helped:
27. What are the aims?
• Help Third Sector organisations improve
efficiency
• Promote OR in the Third Sector
• Promote effective use of OR
• Give OR analysts an opportunity to practise
in a wider arena and widen their skills
28. Some comments from the clients:
• “We’ve benefited hugely from your work and support in all areas of the project, and from
an organisational perspective you’ve enabled us to take a highly professional approach to
increasing the efficiency of our charity”
• “The work is already supporting our planning and development for next year and
allowing us to focus our thoughts and decisions on the places of most importance for our
organisation”
• “Resource planning has been a stumbling block for years. As a new CEO I know how
important it is to motivate staff and a key part of that was re-thinking how we plan our
services.”
29. What happens next?
• Expression of interest.
• Registration form.
• Speak with OR specialist to discuss if there is a potential project.
• Project scope sent to volunteers.
• Organisation selects volunteer.
• Project proposal drawn up.
• Project commences.
• Feedback sought.
30. A Case Study in
Operational Research
“It’s a dog’s life”
Ian Seath
31. Can we answer the following questions?
• How many dogs there are “in the system”
in the UK and how do they move
between different parts of the system?
• How many dogs are relinquished and
why? Which type of dog?
• Can we forecast the likely impact of
different interventions?
The RSPCA is the UK’s leading animal welfare
charity.
32. Key project stages
• Review the available data
• Design and build a model
• Develop recommendations for
possible uses and future development
of the model
33. What data would you want to collect and
where would you find it?
34. Literature Review
• Literature Review of
relevant papers and
reports
• ~50 different sources
• ~100 data points and
times series
• Assessment of data
quality
37. A Stocks and Flows Model of UK dogs
• A Stocks and Flows Model helps us to
understand the “system” in which dogs
exist; where they are and how they move
between different parts of the system
• Decide what four “stocks” of dogs you
would want to include in a model, based
on the overall population data from the
literature review
Stock 1
Stock 4
Stock 2
Stock 3
38. A Stocks and Flows Model of UK dogs
• Here is a model answer that
defines four main stocks of dogs
• What do you think are the inflows
and outflows that lead to increases
and reductions in each of the
stocks?
• N.B. there may be more than
one inflow and/or outflow for
each stock
Companion
dogs
Working
dogs
In Rescue
Strays
39. A Stocks and Flows Model of UK dogs
Companion
dogs (99.8%)
Working
dogs
In Rescue
Strays
• Births
• Imports
• Deaths
• Exports
• Births
• Imports
• Deaths
• Deaths• Deaths • Births
• Births
• Imports
41. Factors affecting the flows
Canine
birth &
death
rates
Political
Economic
Social
Technology
Legal
Environment
• If you were the RSPCA, what
interventions would you want to
make to improve the system?
42. The Model
Stray dogs
Dogs in welfare
Owned dogs
Owned births
Owned
deaths
Stray births
Stray deaths
Welfare
births
Welfare deaths
Owned dogs to welfare
Welfare dogs to owned
Owned dogs to stray
Stray dogs to owned
Stray dogs to welfare
Imported owned dogs
Dogs imported to
welfare
Owned dog
exports
Number of dogs
microchipped
Owner education
level re dogs
Welfare life expectancy
Owned dog life
expectancy
Stray dog life
expectancy
Space available in
welfare
Euthanasia policy
Emmigration
Number of dog
breeders
Number of puppy
farms
Funding for welfare
Welfare capacity
Demand for
dogs in UK
Number of
owned dogs
having puppies
Demand for dogs -
non-UK
Disposable income
Insurance rates
Number of
dogs lost
Number of
dogs
abandoned
Number of dogs
neutered
9.4m (23)
2009 LB: 12,340
2010 LB:10,630
est 1,250 (300
to 2130)
111,000 per
year (13)
Strays back to owner -
53,280 per year (13)
Rehomed - 9,990
per year (13)
Put to sleep/die 8,880
per year (13)
Working dogs
27,750 per year (13)
Entering welfare:
129,743 per year (1)
0.5% of welfare in
flows (27, scaled)
11.4% of outflow died (27)
LB 7142 euthanised (1)
Assume 0
Welfare dogs reunited with
owner - 12.5% of those
entering = 16,218 (1). 7.7%
of those entering (27)
Rehomed - 80.9%
of outflows (27)
69.5% of welfare
inflows (27, scaled)
30% of welfare inflows (27,
scaled) - this includes LA and
public straying dogs
11.4% of dogs per
year? (36)
25% of dogs per
year? (est from
24, 41)
12 years (34)
4% of microchipped are from Ireland,
3% from other EU, 2% from outside
EU - needs scaling up by 2% (24)
50,000 dogs imported
from Ireland per year (31)
Working dogs
to owned
greyhounds - 3,910
per year (33)
Working
dog births
Working
dog deaths
Working dog life
expectancy
Requirements for
working dogs
Working dogs to
welfareAt least 2500
police dogs
(45)
43. How the RSPCA has been able to use
the results
• Helping to scope the direction of their
campaign to tackle the illegal trade in
puppies
• Using the data to inform their
response to DEFRA’s
consultation on overhauling
existing pet vending legislation
44. How the RSPCA has been able to use
the results?
• Helping to scope the direction of their
campaign to tackle the illegal trade in
puppies
• Using the data to inform their
response to DEFRA’s
consultation on overhauling
existing pet vending legislation
45. WORKSHOP
What are the main issues you are
facing?
Facilitators: Graham Rand, Ian Seath, Rob Mastrodomenico,
Amy Hughes and Amaka Nwagbara
46. What are the main issues you are facing?
• Strategic
• Any big decisions coming up?
• Tactical
• Any processes you are concerned about?
• Improving information
• Any issues about information?
47. In groups
We have 30 minutes. Discuss:
• Issues that concern you
• How OR may be able to help
• Barriers to solving problem e.g.
• Different stakeholders (trustees, paid staff, volunteers) with different
aims/priorities