Research Proposal Example 2

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

MSc Project Management Research Proposal

Predictive Analytics in Risk Management: How historical data and


probabilities are evolving risk management methodologies.
Contents
1. Title ..................................................................................................................................................... 3
2. Introduction ........................................................................................................................................ 3
3. Aims and Objectives ............................................................................................................................ 4
4. Literature Review ................................................................................................................................ 5
5. Methodology ....................................................................................................................................... 6
5.1. Data Collection: Interview ........................................................................................................... 6
5.2. Data Collection: Survey ................................................................................................................ 6
5.3. Analysis of Data: Interview .......................................................................................................... 6
5.4. Analysis of Data: Survey ............................................................................................................... 6
5.3. Project Timeframe ....................................................................................................................... 6
6. Conclusions and Recommendation..................................................................................................... 7
Appendices.............................................................................................................................................. 8
Appendix I. Survey Questions ............................................................................................................. 8
Appendix II. Interview Questions ........................................................................................................ 8
Appendix III. Activity Plan ................................................................................................................... 9
References ............................................................................................................................................ 12

2
1. Title
Predictive Analytics in Risk Management: How historical data and probabilities are evolving risk
management methodologies

2. Introduction
Technology today is a rapidly-changing, ever-expanding range of tools that are greatly improving the
projects and organisations who utilise it. Since the data explosion, there has been an increasing array
of innovative techniques and technological advancements that offer opportunity and rewards to those
who employ them successfully. Risk Management is a crucial stage in the Project Management
process. Both risk and project managers are relied upon for determining the likelihood of identified
risks occurring, what impact these risks turned issues may have, and what contingency plan should be
implemented to avoid it having a disastrous effect. Risk and project managers may draw on previous
experience or knowledge of subject areas to determine the likelihood and impact of these risks.
However, that can of course leave room for human error, for example through unintentional or
intentional bias in decision making. Therefore, risk management is one particular area that has seen
the introduction of technologies like artificial intelligence and predictive data analytics greatly improve
operations.

The aim of this research project is to examine the current technologies in place throughout risk
analysis processes, and how these are aiding risk managers and their methods of decision-making.
Alongside this, the impact human emotion and bias can have on the risk analysis process will also be
explored. Finally, what happens when the predicted analytics are correct and human intervention
affects the outcome, and what happens when the predictive analytics are wrong and what is the
subsequent impact from this. These two areas will be also be explored.

Through this research, including case studies, survey data analysis, and interview responses: the
impact that technology, such as predictive data analytics, has had on the risk management
methodologies and the future of risk analysis innovation will be thoroughly examined.

3
3. Aims and Objectives
The purpose of this research project is to examine the current technologies in place throughout risk
management, to examine the effect human emotion and bias can have on risk analysis, to research
what happens when either the analytics are wrong or the analytics are correct and ignored, and to
explore where this technology looks to be heading towards in the next ten years. This research will be
conducted throughout the following aims and objectives:

1. Examine the history of predictive analytics and risk analysis tools and technologies through
the Literature Review:
- Look at how data analytics and predictive analytics, have evolved since the Data
Explosion.
- Look at the advancements of the technologies applied to the risk management
industry over the last decade.
2. Investigate key incidents which involved the use of predictive analytics in risk through Case
Studies:
- Examine the L’Aquila 2009 Earthquake: where, based on data, a scientist flagged the
high risk of an earthquake yet was called an ‘alarmist’ – the predicted earthquake then
occurred and had a disastrous impact.
- Examine the Boeing 737 Max risk analysis procedures. Through analysis, the aircraft
was found to have a high likelihood of fatal crashes however the agency allowed it to
continue flying which led to a number of crashes killing hundreds of passengers and
crew.
- Examine the 2016 US Presidential Elections. Based on data, Hilary Clinton was the
predicted winner of the presidential race right up until the last minute. Study the
findings from this, and how the predicted data was so wrong.
3. Examine how human emotion and bias can affect decision-making using a survey, and
subsequent data analysis of the survey results:
- The survey will consist of two main questions: one with a known outcome and one
with an unknown outcome (see Appendix I. Survey Questions). Here, the survey
participant will be asked to select an answer from the options given.
- Along with this, the participant will also have the opportunity to provide their own
numerical answer to the question, based on their opinion.
- The survey results will then be compared against calculated probabilities for both of
these events using available data: historical data for the known outcome and gathered
data for the unknown outcome.
- The aim of this survey is to determine if previous knowledge (i.e. from the known
outcome) can affect someone’s decision-making for risk.
4. Finally, establish what technologies are already in place within industry – how they are
utilised, what they do exactly, and where they are likely to evolve in the next decade:
- This will be researched through use of an interview: which will be carried out with a
person within industry.
- The interviewee will have knowledge and experience of the current technologies used
for the risk analysis process within their industry.
- Interview participant(s) currently being sought, and will be confirmed at a later date.
This will ideally be someone within the IT or financial industry, who will have
experience working with both technology and risk.
- To avoid security or ethical issues, questions will not be too specific. Instead, it will be
more personal to the interviewee, e.g. Do you feel these technologies are effective?

4
4. Literature Review
E.D. Solozhentsev (2012) discusses the risk management technologies and models throughout
economic and engineering systems in Risk Management Technologies: With Logic and Probabilistic
Models. He examines the components of these technologies, models of failure risk and models of bank
operational risk, and assessment and probabilities. He also explores examples of the technology
behind risk management and analysis, such as the analysis of insurance against fire of a dangerous
object.

Eric Siegel (2013) discusses the many uses of predictive analytics in Predictive Analytics: The power to
predict who will click, buy, lie, or die. Throughout this book, he examines the use of data and machine
learning, used to offer predictions and probabilities. He also explores examples of the use of predictive
analytics in risk, for example insurance companies, and the aviation and banking industries.

David E. Alexander (2010) examines the L’Aquila Earthquake of 2009, the background and risks
identified beforehand and the affect the earthquake had on the area at the time. He explores the
Disaster Risk Reduction (DRR), the risks identified early on, and the lessons learned from this disaster.

David Slotnick (2019) discusses the risks identified early in the analysis of the Boeing 737 Max, but
how the aircraft was allowed to continue flying. He examines the level of risk the aircraft manufacturer
faces, how the Boeing 737 Max was allowed to continue flying despite the risks, and the impact this
had when numerous crashes with this aircraft then occurred. Alongside this, Andy Pasztor and Andrew
Tangel (2019) again highlight the risks identified with the aircraft, with their initial analysis and review
of the situation in The Wall Street Journal.

James Markarian (2017) explores what went wrong with the predictive analytics behind the US
Presidential Elections of 2016. He explores the data science behind the analytics, the reliability of
these produced analytics, and lessons learned from these predictions being so wrong.

Ian Jordaan (2005) calculates the probability of the RMS Titanic colliding with an iceberg in his paper
Decisions under Uncertainty: Probabilistic Analysis for Engineering Decisions.

Finally, Mike Allen (2018) discusses the Axios interview with engineering giant Elon Musk, and the
likelihood of Musk’s company SpaceX launching an exploration to Mars in the near future.

5
5. Methodology
For the first part of this research project, a comprehensive literature review will be conducted to
examine the background of risk management, and the subsequent case studies within that area.

The primary research data for this project will be collected using the following tools: a survey and an
interview. Featured below are the methodologies for both the interview and the research: the data
collection and analysis behind both sections. Also featured is the proposed timeframe for this research
to be completed.

5.1. Data Collection: Interview


The data for the interview section (Objective 4) of this research project will be collected through a
one-to-one interview format, either in person or over the telephone which will be confirmed along
with a confirmed interview date nearer the time.

5.2. Data Collection: Survey


The data for the survey research (Objective 3) will be collected through the results of respondents
from an online survey. This survey will be created and distributed using SurveyMonkey and the
website’s appropriate links, i.e. via social media or messaging platforms. The respondents for this
survey will be random. There will be no specific group of people targeted for this research, as this will
allow the survey to gather a wide and general opinion (see Appendix I. Survey Questions).

For the data collection, the pilot survey will be using non-random sampling – meaning the pilot survey
will be sent to a select group of five individuals to review the survey questions and gather feedback.
Following this, the main survey will be using simple random sampling, and from this snowball sampling
will be implemented. This will allow the survey to gather a wide-ranged, general opinion to be
analysed later.

5.3. Analysis of Data: Interview


The data gathered from the interview will be analysed in a comprehensive report format. The report
will cover all questions asked and their subsequent answers, and visualise data if and when required
through diagrams or charts.

5.4. Analysis of Data: Survey


The data gathered from the survey will be analysed using descriptive statistics. Here, the survey
response data will be summarised using graphs or figures, i.e. histograms, pie charts, or scatter plots.
Summary statistics will also be implemented during the comparison stage of the research, when the
survey responses are compared against the earlier calculated statistics.

Also included in the survey data analysis will be inferential statistics, namely hypothesis testing using
continuous random variables from the survey responses in questions 2 and 4 (see Appendix I. Survey
Questions) – with the null hypothesis being ‘Previous knowledge of events affect the bias in human
decision-making’.

5.3. Project Timeframe


Research will take place over a two month (eight week) time period (see Appendix III. Activity Plan).
This will allow time to collect data from survey responses and thoroughly analysed it.

6
6. Conclusions and Recommendation
This research project will allow a thorough examination of the history of both predictive analytics and
risk management technologies, followed by the exploration of their current deployment and uses
through a comprehensive literature review.

The case studies will allow the exploration of predictive analytics in risk management throughout
different industries, and will examine what happens when:

1. The predicted risk analytics are correct but disregarded, and the subsequent impact this had.
2. The predicted risk analytics are incorrect, and the subsequent outcome from this.

The interview will allow an investigation into the current predictive analytic technologies used in risk
management in a real-world environment. This will build a better understand of why they are
deployed, how they are utilised, what they do, and if any improvements are planned.

Finally, the survey will allow the analysis of the decision-making process in risk analysis and
management. It will provide an in-depth study of human decision-making vs. calculated probabilities
or predictions, and examine if human decision-making can be impacted by previous knowledge or
bias.

Overall, this research project aims to explore the use of predictive analytics in risk management, and
if they are improving or impeding the industry. It will examine the growth of these technologies,
studying how it has evolved the industry so far and how it will advance over the next decade.

7
Appendices
Appendix I. Survey Questions
The following four questions will be featured within the survey for this research project:

This survey has been created with the aim of exploring the average decision-making process in risk
analysis. Through this survey, the aim is to research if previous knowledge affects bias in critical
thinking. There will be four questions in which respondents will be asked to give their own opinion.
This survey should take no longer than 10 minutes to complete.

1. How likely do you think it was for the RMS Titanic to collide with an iceberg? Select one
answer.
a. Very Unlikely: 0 – 10 % Chance
b. Unlikely: 10 – 20% Chance
c. Somewhat Unlikely: 20 – 30% Chance
d. Somewhat Likely: 30 – 40% Chance
e. Likely: 40 – 50% Chance
f. Very Likely: Over 50% Chance
2. In your opinion, what do you think was the likelihood (chance) of the RMS Titanic colliding
with an iceberg? Please answer in percentage, e.g. 50%
a. Answer will be provided in numerical format by the respondent
3. How likely do you think it is for man to go to Mars in the next 50 years? Select one answer.
a. Very Unlikely: 0 – 10 % Chance
b. Unlikely: 10 – 20% Chance
c. Somewhat Unlikely: 20 – 30% Chance
d. Somewhat Likely: 30 – 40% Chance
e. Likely: 40 – 50% Chance
f. Very Likely: Over 50% Chance
4. In your opinion, what do you think is the likelihood (chance) of man going to Mars within the
next 50 years? Please answer in percentage, e.g. 50%
a. Answer will be provided in numerical format by the respondent

Appendix II. Interview Questions


The following questions will be asked during an interview with an industry worker. Through this
interview, the aim is to examine what current risk analysis and management technologies are in place
– how they are utilised, what they do exactly, and where they are likely to evolve in the next decade.

1. What is your experience with technology and/or risk management?


2. Have you worked with predictive analytics?
3. Do you currently use any risk analysis technologies in your place of work?
4. Do you apply a more qualitative or quantitative approach to risk management? Or a mixture
of both?
a. Which do you think is more affective in the risk management process?
5. Do you find these technologies useful/effective during risk analysis/management procedures?
6. In your opinion, by using these technologies, what is the aim?
7. Do you feel predictive analytics are beneficial to the risk analysis process?
8. Do you see these technologies growing in the next decade?
a. If so, what exactly do you think will change?
9. Would you make any changes to the current risk analysis processes?
10. Any other comments to add?

8
Appendix III. Activity Plan
An activity plan (or Gantt Chart) has been created for the lifespan of this research project. Here, tasks are identified in the left column, and are marked with
their planned completion months – please note some tasks may stretch across multiple months. Milestone tasks are clearly highlighted in red, and have been
added to a separate Milestone Table on the right of the Activity Plan.
10
11
References
Andy Pasztor and Andrew Tangel. (2019). Interal FAA Review Saw High Risk of 737 MAX
Crashes. Available: https://www.wsj.com/articles/internal-faa-review-saw-high-risk-of-
737-max-crashes-11576069202. Last accessed 17th March 2020.

David E. Alexander. (2010). The L'Aquila Earthquake of 6 April 2009 and Italian
Government Policy on Disaster Response. Available:
https://www.tandfonline.com/doi/full/10.1080/19390459.2010.511450. Last accessed
17th March 2020.

David Slotnick. (2019). The FAA said the Boeing 737 Max had a high risk of crashing, but
let the plane continue flying anyway. Available:
https://www.businessinsider.com/boeing-737-max-faa-analysis-likely-to-crash-2019-
12?r=US&IR=T. Last accessed 17th March 2020.

E.D. Solozhentsev (2012). Risk Management Technologies: With Logic and Probabilistic
Models [online]. United States: Springer. Available from: Springer Link [accessed 17
March 2020].

Eric Siegler (2013). Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or
Die. United States: John Wiley & Sons.

Ian Jordaan. (2005). Decisions under Uncertainty: Probabilistic Analysis for Engineering
Decisions. Cambridge University Press: United Kingdom. p40-43.

James Markarian. (2017). What The Election Taught Us About Predictive


Analytics. Available: https://www.forbes.com/sites/forbestechcouncil/2017/02/08/what-
the-election-taught-us-about-predictive-analytics/#5b4cf0fc6532. Last accessed 17th
March 2020.

Mike Allen. (2018). Elon Musk: There’s a 70% chance that I personally go to Mars.
Available: https://www.cnbc.com/2018/11/26/elon-musk-says-there-is-a-70percent-
probability-he-will-go-to-mars.html. Last accessed 17th March 2020.

You might also like