Key Competencies for
Advancing in
Statistical Career Paths
Organized by Dr. M. Subbiah
Dept. of Statistics, SDNBV College Data Science Mentor
January 3, 2024 Pratiti Analytics LLP
Theme: Navigating Statistical Roles
01 Introduction Importance of statistical skills in today's data-driven world
02 Academic to Industry Foundations and Expectations
03 Cultures of Statistics Statistical Thinking and Learning
04 Conclusion Field Dynamics, Need for sustained learning and development
05 Q&A PURPOSE
01
Introduction
Importance of statistical skills in today's data-driven world
Statistics in Data World
A fundamental tool in the data world
Enabling Data Analysts and Scientists to
○ derive meaningful insights
○ make informed decisions
○ build reliable models for various applications
○ make sense of large and complex datasets
Data Analysts
Data Collection
Data Cleaning and Preprocessing
Data Analysis
Descriptive and Exploratory Analysis
Data Visualization
Business Intelligence
Data Mining
Quality Assurance
Data Scientists
A pivotal role in extracting valuable insights and knowledge from large
and complex datasets
Contribution to
○ informed decision-making
○ leverage their expertise in data analysis, statistics, machine learning, and
domain-specific knowledge.
○ extracting actionable insights from data
○ driving organizational success
Data Scientists
Problem Formulation Machine Learning Deployment
Data Collection and Cleaning Optimization and Iteration
Exploratory Data Analysis Data Interpretation and Communication
Feature Engineering Big Data and Advanced Technologies
Model Development Cross-functional Collaboration
Predictive Analytics Ethics and Privacy
Transitioning Careers
Data Analyst Data Scientist
Focus Focus
often involved in reporting and visualization Emphasize building and deploying machine learning
to support business decision-making models to make predictions about future events
Scope of Work Scope of Work
Typically work on well-defined problems with Tackle complex, open-ended problems that may involve
a clear business context. unstructured data.
Support decision-making by providing Engage in end-to-end data science processes, from
insights based on historical data. problem formulation to model deployment
Skill Set Skill Set
Proficient in statistical analysis and reporting Possess advanced skills in machine learning, statistics,
tools and statistical analysis. and programming in languages like Python or R and
Use tools for analysis and visualization, frameworks
Programming Languages like R / Python Strong in data preprocessing, feature engineering, and
model development.
Scope
Industries Organizations
• Manufacturing • Indian Statistical Service (ISS)
• Logistics • Indian Council of Medical Research (ICMR)
• Finance • Indian Council of Social Science Research
• Insurance (ICSSR)
• Medical • National Informatics Centre (NIC)
• Health care • Central Board of Taxes
• Health Economics • National Health Systems Resource Centre
• Pharma (NHSRC)
• Media and Journalism • Indian Council of Agricultural Research (ICAR)
• CRO • Reserve Bank of India (RBI)
• Entertainment • Central Statistical office (CSO)
• E-Commerce • National Sample Survey Organization (NSSO)
• Education • Private / Government HEIs
• Sports and Well Being
• Publications
Scope
International Organizations
• World Bank
• International Monetary Fund (IMF)
• Eurostat
• Organization for Economic Co-operation and Development (OECD)
• International Labour Organization (ILO)
• World Health Organization (WHO)
• United Nations Development Programme (UNDP)
• International Telecommunication Union (ITU)
• United Nations Children's Fund (UNICEF)
• International Rice Research Institute (IRRI)
Roles & Impact
1. Statistician 1. Child, Maternal and Families’ Health
2. Research Statistician 2. Countries' Economic Policies
3. Research Associate 3. Development and Analysis of Statistical
4. Quantitative Analyst Indicators
5. Biostatistician 4. Economic Development
6. Econometrician 5. Education
7. Actuary 6. Epidemiology
8. Statistical Programmer 7. Formulating Health Policies
9. Data Scientist 8. Global Health Statistics
10. Survey Statistician 9. Global ICT Trends
11. Operations Research Analyst 10. Health Research
12. Financial Analyst 11. Health Trends
13. Quality Assurance Statistician 12. Labor and Employment Issues
14. Predictive Modeller 13. Organization's Research and Publications
15. Clinical Statistician 14. Poverty Reduction
16. Statistical Consultant 15. Social Progress
17. Academic Statistician 16. Social, and Environmental
17. Sustainable Development and Governance
Preperation is the Key
A picture is worth a thousand words
Situation Analysis Strategy Implementation
Dew Factor – Evening Toss won Performance (Head and
after 7 Chose to Bowl Cummins)
Decision – Team Double Phase Wicket Bowling Changes with
Double Phase Wicket Part time bowlers (Not
able to capitalize)
Courtesy: https://www.youtube.com/watch?v=pCNU0hYTArs
02
Academic to Industry
Foundations and Expectations
Academic
Foundations Articulation
Mathematics Language for Uncertainty
Languages Three Eyes of Statistics
Employament
Computation Application
Calculator to Computer Eye Opener
Menus to Coding
Target
30% 50% 20%
Nature Theory Practice
Outcome
Situation
Competency Analysis
Able to understand
the problem
01 03 Pick the right tools
and techniques
Strategy Implemetation
Theoretical
understanding
02 04 Coding / Programming
and Story Telling
03
Cultures of Statistics
Statistical Thinking and Learning
What, When, How & Why?
Uncertainty Information Mathematics I
What is the probability of What is the probability of What is the formula for
getting a head in single getting a head in single finding Arithmetic mean
toss of a coin? toss of a fair coin? for ungrouped data?
Statistics Mathematics II Models
Why that is the formula? What is the difference Why t-test in Linear Models?
When do we use percentile? between Finite / Infnite / Is it a small-sample test?
How percentile is Infinity? Approximate / How do we decide a test?
computed? Asymptotic?
Approach - Trio
Mathematics
M Abstraction
S
Generalization
C
Academic Statistics
Ladder S Intuitions
Questions
R
Relations
Insights E
Computation P
C Coding
Programming
Percentages
40% 60%
Mathematics Statistics
• Understand Theory* • Choose the tight culture
• Ability to conceptualize • Identify the right Questions
• Identify the constraints • Do a firm SWOT analysis
• Choose the right • Know the ways to interpret
Approach • Pick the right tool for better
*5 Functions - Eg story-telling
Inference From a Model
Model equation is
◦ bweight = -4865.2 + 206.6 * nutrition and R2 is 0.54
Interpretation
◦ Coefficient of " nutrition ": one unit increase in nutrition will increase
the mean "bweight" by 206.6 g
◦ Plus sign indicates increment
◦ Constant / Intercept: when no nutrition, bweight is -4865.2
Inference From Models
Model equation is
◦ bweight = 3130 + 206.6 * (nutrition - 38.687) and R2 is 0.54
Interpretation
◦ Coefficient of " nutrition ": one unit increase in nutrition will increase the
mean "bweight" by 206.6 g
◦ Plus sign indicates increment
◦ Constant / Intercept: Average "bweight" when the " nutrition " is equal
to its average value
04
Conclusion
Field Dynamics,
Need for sustained learning
and development
Propeller
Read
Books* are the Treasure
Train
Whatever You do You Learn
Socialize
Join the right community
*Introduction to Mathematical Statistics. Hogg, Joseph W. McKean, Allen T. Craig https
://minerva.it.manchester.ac.uk/~saralees/statbook2.pdf
Our goal - Read
Books* are the
Treasure
*Introduction to Mathematical Statistics. Hogg, Joseph, Allen T. Craig
https://handoutset.com/wp-content/uploads/2022/05/Introduction-to-Mathematical-Statistics-Robert-V.-
Hogg-Allen-Craig.pdf
Our goal - Train
Recent Advances of Statistical Cultures
Statistical Modeling: The Two Cultures in 2001
http://www2.math.uu.se/~thulin/mm/breiman.pdf
Statistical Modeling: The Three Cultures in 2023
Whatever You do You
https://hdsr.mitpress.mit.edu/pub/uo4hjcx6/release/1#:~:text=Table%202%2
Learn 0shows%20an%20overview,and%20ML%20for%20theory%20prediction
.
The Perils of P-Values (2020)
https://link.springer.com/chapter/10.1007/978-3-030-43714-5_5
The ASA Statement on p-Values: Context, Process, and Purpose (2016)
https://www.tandfonline.com/doi/full/10.1080/00031305.2016.1154108
The debate about p-values(2015)
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4858512/
Programming Skills
Data Analysis Projects
Application Areas – Economics / Ecology / Psychology / Language /
Medical / Health / Sports / Finance etc
Our goal - Socialize
WWW - data Sources - free
https://archive.ics.uci.edu/
Application Areas –
Join the right
Economics
community Ecology
Psychology
Language
Medical
Health
Sports
Finance…………………………………….
Publish
https://github.com/
https://rpubs.com/
https://streamlit.io/
No LinkedIn kind
Holistic Approach
Level 4
Prediction for
new unseen
data from the
Level 2 learning
Level 5
Validate
intuitions, AI Techniques,
assumptions Technologies,
Level 3 and Tools
through data
Explain
theory testing
/ building
or
Level 1 causal
Describe Data, relations
Insights from Data
Job Market
Visual
Analyst / Data AI / Data Legal
Analyst / Data Product Compliance /
Scientist Developers Data Privacy
Business Data Engineers
Developers / / Architects /
Analysts / Infrastructure
Marketing & Engineers
Sales
Ethics – Future Professionals
The man who moves a mountain begins by carrying away
small stones - Confucius
05
Q&A Purpose
Respect Nature
Thank You
[email protected]
9952726863
I am still learning - Michelangelo
