Early Career Success - Part III - PowerPoint Presentation

Early Career Success Factors for

Statisticians in Business and

Industry

Robert N. Rodriguez

Senior Director, Statistical R & D

SAS Institute

Pre-JSM Diversity Workshop

Joint Statistical Meetings

August 1, 2009

Outline

Personal perspective at SAS

Opportunities for statisticians in business and

industry

Factors for early success

About SAS

Leader in statistical software used by universities,

business, and government

Founded in 1976

Continuous reinvestment in research and development,

including 22% of revenue in 2008

11,000 employees, 400 offices globally

Over 45,000 customer sites in 110 countries

SAS Research & Development

1000+ software developers in Cary, Beijing, Pune, ...

Integrated development environment

Millions of lines of C and Java code

Systems for building, documenting, and delivering

software

SAS Campus

Cary, North Carolina

Advanced Analytics Division

Over 100 Ph.D. specialists in statistics, operations

research, numerical analysis, …

Software products

SAS/STAT, SAS/ETS, SAS/QC, SAS/OR, SAS/IML,

Enterprise Miner, Forecast Server, …

Used by statisticians, researchers, data miners, …

Analytical components for software solutions

Data Flood

Data-Based Decisions

Customer Perspective

Learning About Customer

Problems and Data

Statistical Needs in Corporate

Environments: The Five D’s

1. Data planning Design of surveys, experiments, clinical trials, …

2. Data access and management Disparate data sources and poor data quality undermine analysis

Databases, data warehouses (controlled by IT, not analysts)

3. Data preparation Getting the data into analysis-ready form (“80% of the effort”)

4. Data analysis and modeling

5. Delivery of analytical results User interfaces, graphics, web reports, FDA submissions, …

What’s Involved in Producing

Statistical Software?

1. Listening to customers

2. Keeping up with advances in

statistical methodology

3. Designing, writing, testing code

4. Writing user documentation

5. Providing technical support and

training

6. Consulting with customers

7. Presenting to customersStatistical software testers

Cheryl LeSaint and Yu Liang

Growth Opportunities for

Statisticians

1. Development of analytical solutions

Integrated solutions for business problems

Developed by interdisciplinary teams industry experience

software development skills

expertise in statistics, data mining, operations research

Examples fraud detection for banks

credit scoring

customer retention and marketing automation

credit, market, and operational risk analysis

web analytics

warranty analysis

Growth Opportunities for

Statisticians (cont’d)

2. Consulting in financial and retail industries

Ability to formulate a business problem with a statistical

model

Examples

survival models for customer lifetime value

predictive model for repayment behavior

forecasting demand for store items

experimental design for direct marketing

Early Success Factors:

Undergraduate Preparation

How much math do I need for graduate work in statistics?

calculus, linear algebra

statistics is not a branch of mathematics, but you need to be

mathematically prepared

What should I major in?

math, statistics, biology, computer science, physics,

engineering, economics, psychology, …

What else should I take?

probability and statistics--enough to understand how modern

statistics is used to analyze data and solve problems

computer programming

technical writing

Early Success Factors:

Undergraduate Preparation

Explore statistical careers through internships,

StatFests, and summer opportunities

North Carolina State

Summer Institute for

Biostatistics Training

Field trip to SAS

Early Career Success Factors:

Graduate Training

Apply to the right program in statistics or biostatistics

Where would you like to work when you finish?

Are you interested in academic research and teaching?

Are you interested in business or government?

Talk to faculty and alumni—they’ll be glad to advise you!

Consider an in-demand area of statistics

Survey design and analysis

Econometric modeling

Statistical computing

Become a student membership of ASA!

Internship opportunities in December Amstat News

Early Career Success Factors

Interviewing

Research the organization in advance

Ask perceptive questions

Are the staff absorbed in their work?

What brings them back to work year after year?

What is the least satisfying aspect of their work?

What is the most rewarding aspect of their work?

If you are excited about what they are doing, they

will be excited about you!

Early Career Success Factors:

Where to Start Your Career

Excellence attracts excellence, so look for

group of flourishing statisticians

statisticians valued as problem formulators/solvers

statisticians collaborating with others

senior statisticians serving as mentors and leaders

statisticians active professionally (members of ASA)

Traps to avoid

isolation from other statisticians

limited understanding of what statisticians do (“just run the

reports”)

lack of support for professional activity

Early Career Success Factors:

Becoming a Prized Professional

Work on your writing skills

Careful motivation

Clear conclusions

Learn to give presentations that

anticipate and meet audience needs

Develop special computing skills

Management of large data sets

Advanced statistical programming

Become active in the ASA

Keep learning

Give back to the profession!

Effective Writing

by H. J. Tichy

A best buy at $2.57

Early Career Success Factors for

Statisticians in Business and

Industry

Robert N. Rodriguez

Senior Director, Statistical R & D

SAS Institute

Pre-JSM Diversity Workshop

Joint Statistical Meetings

August 1, 2009

What Drives Statistical Software

Development?

Customer Problems Recent Development Directions

Complex dataHighly flexible models, Bayesian models, methods for model selection and validation

Missing data Multiple imputation

Messy data Outlier detection, robust methods

Planned dataSurvey methods, sample size computation, design of experiments

Unexplored data Graphical methods

Massive dataScalable algorithms, parallel processing, distributed computing

What We Look for in

Statistical Software Developers

Ph.D. in statistics, biostatistics, applied math, …

Specialization in an area of modern statistics

In-depth knowledge of computational techniques

Professional programming skills (hard to find!)

Ability to write large, complex programs in C (not the same as

writing programs in SAS, Matlab, S-PLUS, or R)

Developed through on-the-job mentoring

Motivation

Challenged by creating software that moves new methods into

practice and helps customers solve problems

What We Look for in

Statistical Software Testers

M.S. or Ph.D. in statistics, biostatistics, …

Graduate coursework in several target areas

Knowledge of applications and computational methods

Skills

Ability to verify computations through validation programs

written in SAS, SAS/IML, SAS macro

Ability to communicate effectively with other testers and

developers

Motivation

Challenged by setting and meeting high standards of

accuracy and performance that exceed customer expectations

Where Do Statisticians Contribute

at SAS?

Software development

50+ developers for statistics and operations research

Software testing

20+ testers

Documentation

Technical support

15+ statisticians

Education

12 statisticians

Marketing and consulting