際際滷

際際滷Share a Scribd company logo

Dr. Lara Sucheston-Campbell - Building a working farm: Planning and planting for a successful analytics culture and team

Download as PPT, PDF
R
Rehgan Avon

2018 Women in Analytics Conference https://www.womeninanalytics.org/ Setting up a successful analytics team with the objective of asking and answering any hypothesis based questions is a bit like running a farm. Two fundamental principles of your (data) farm: its important the farm has a mission that everyone buys into and to which everyone can contribute andits imperative that your fields (of research) are always in slightly different states from planning to harvesting. This talk will focus on approaches to creating and maintaining an effective and successful analytics team, defined simply as people working together in a committed, interdependent way where team members share responsibility and hold themselves accountable for attaining the results. I will cover the elements of an effective team, from establishing an environment of psychological safety to structure and clarity of purpose, what an effective team member looks like, how to find and cultivate him or her and how to avoid the Apollo syndrome. Sustaining your farm while its growing is just as imperative as its growth, and thus we will also discuss the value of the daily stand-up and the use project retrospectives so next seasons crops are even better. This talk is for anyone interested in building a happy, productive analytic group and data farm of any size.

1 of 34
Download to read offline
Building a Working Farm: Planning and Planting for a Successful Analytics Team and Culture Lara Sucheston-Campbell, MS, PhD Associate Professor of Pharmacy and Veterinary Medicine Associate Director, Mathematical Biosciences Institute The Ohio State University
Resources Your farm Creating your team Creating and maintaining your team and culture 2
A brief introduction Im a Engineer (MS, Industrial Engineering- Operations Research) turned Genetic Epidemiologist (PhD - Genetic and Molecular Epidemiology) Finding & identifying genes related to getting cancer or treatment response. 2007-2016 Assistant then Associate Professor, University of Buffalo/Roswell Park Cancer Institute 2016 Associate Professor, The Ohio State University 3
Genetics and Epidemiology of Myeloid Malignancies (GEMM) Consortium Motivation: 5 year survival rate for Acute Myeloid Leukemia (AML) is 27% Little is known about AML risk factors, including the role of genetics. To overcome this barrier, we formed GEMM to unite investigators world-wide and pool data and resources. 4
Determining Influence of Susceptibility COnveying Variants Related to 1 Year survival after Blood or Marrow Transplant (DISCOVeRY-BMT) 5 Motivation: 1 year survival rate for leukemia patients treated with BMT is ~65% HLA Genetics is the biggest predictor of survival, the rest of the genome is underexplored To overcome this barrier, we formed DISCOVeRY- BMT to unite investigators world-wide and pool data and resources to find other important genes and improve survival 5
Congratulations you have a PhD! 6 Heres some (a lot of) money AndGO!
I did not get a PhD in this Cocktail conversations with successful female scientists Making the Right Moves: A Practical Guide to Scientific Management for Postdocs and New Faculty http://www.hhmi.org/developing-scientists/making-right-move Work Rules! Laszlo Block former SVP of People Operations at Google re:Work- a collection of practices, research, and ideas from Google An effort by Google to help share and push forward the practice and research of data-driven HR https://rework.withgoogle.com/about/ 7
Filter 8 -team size -dry lab appropriate -money -practical implementation -reality
Team definition 9 People working together in a committed way to achieve a common goal or mission. The work is interdependent and team members share responsibility and hold themselves accountable for attaining the results. MIT Information Services & Technology
Building a working farm 10 CC w/C. Ambrosone Harvest Growing Planting Planning
Why a farm? Easy to visualize and plan over time. Overall mission is implemented through a series of small/medium/large projects. Excellent framework for presenting results and conclusions (research or analytics). Growth happens through permutations of set resources. Economics. 11
What will you grow? Develop a clear vision Developing a shared team vision does not limit innovation. A vision statement provides a foundation for creativity for new directions. 12 HHMI
My vision My vision is that we will leverage genetics to identify individuals at risk for hematologic malignancies or disease and significantly impact patient survival after a diagnosis. 13
What will you grow? Implement the vision with a mission statement Paint with broad strokes, but also identify key measures of success. Provide reasoned and emotional justification. Tie it to the values and culture of your company. Be clear and honest. Create a future that distinguishes your program from competitors. 14 HHMI
15 Does having a good mission statement really work? Work Rules! (G); Wrzesniewski & Grant Team members who are inspired have increased productivity and happiness. Reframing tasks as opportunities to help others, can make occupations and tasks feel more significant and increase motivation.
Me Hires CODING Hire people better than you (at something) A bad hire is toxic and drains your teams energy and creativity. Set the bar high, dont compromise, be willing to wait! Hire someone who is better than you in some meaningful way. Youll end up with a much stronger team. 16 Work Rules! (G)
17 How to hire people better than you (at something) Give people a reason to join your team (mission) Everyone recruits Take a risk: Consider a trade-off by going outside of your immediate field A computer scientist or physicist interested in bioinformatics but without a strong biology background. Someone with a molecular biology background who has started self-training in R or Perl.
18 Take the curious Research presentation Journal Club Chalk talk Multiple interviews across disciplines Long term incentives (aside from $) Additional conferences, trainings and first author papers opportunities. How to hire people better than you (at something)
Definition of a Team 19 People working together in a committed way to achieve a common goal or mission. The work is interdependent and team members share responsibility and hold themselves accountable for attaining the results. MIT Information Services & Technology FUNCTIONAL & EFFECTIVE
Apollo Syndrome: sum of the parts < whole 20 Dr. Meredith Belbin discovered the Apollo Syndrome, a condition where teams of highly capable individuals, selected using ability and aptitude tests, can collectively, perform badly.
Apollo Teams 21 Teams selected using ability and aptitude tests. They were bound to win. The Apollo teams often finished near the bottom.
Why did they fail ? Excessive time in destructive debate. Difficulties making decisions, resulting in jobs often being omitted. Team members acted in silos, resulting in teams that were difficult to manage. Some teams recognized what was happening and over compensated by avoiding confrontation. 22
What does the Apollo syndrome teach us? 23 HOW A TEAM WORKS MATTERS MORE THAN WHO IS ON THE TEAM AS MUCH AS
24 Modified from Project Aristotle What Makes an Effective Team?
Making your team safe, eg lets make a Sharknado part 6 Frame the work as a learning problem, not an execution problem. I may not be understanding this correctly, I may miss something, this is why I hired people smarter than me. Ask a lot of questions about everything , eg model 25
Additional ways to help achieve an effective team culture 26
Trust Be transparent. Encourage people to think & act like owners. Give your team an uncomfortable amount of freedom. Expect a lot. but verify Adapted from Work Rules! (G))
A tale of two tails: best and worst Dont surprise people. Tell them if theyre low performers and help them learn or find new roles Put your best people under a microscope to find outand replicatewhat makes them succeed. Work Rules! (G))
Short-term one time incentives: the death of learning Self-Determination Theory (a theory of motivation) -Attaching incentives to tasks reduces intrinsic motivation when incentives are removed. Adapted from Deci & Ryan; Work Rules! (G)
Pay unfairly and reward failure The best people are worth more than average people. Reward failure, else you kill risk taking. Work Rules! (G))
Stand up meetings Short meetings (10 mins) to start your day. For my lab, these are only on Monday Go around the table: What are you working on? What are your goals for the week (or today?) Are there barriers to accomplishing these goals? How can I help? 31
You will do the same or a similar work again, what worked and what did not? 32 A look back: the role of the Retrospective
The take home Have a team mission Hire people better than you Splash time (and cash) to recruit Try something new Wait! How a team works is as important as who is on it Trust a scary amountbut verify Looks at your tails Long term not short term incentives Pair unfairly Reward failure 33
Thank you for listening! Questions? 34

More Related Content

Dr. Lara Sucheston-Campbell - Building a working farm: Planning and planting for a successful analytics culture and team

  • 1. Building a Working Farm: Planning and Planting for a Successful Analytics Team and Culture Lara Sucheston-Campbell, MS, PhD Associate Professor of Pharmacy and Veterinary Medicine Associate Director, Mathematical Biosciences Institute The Ohio State University
  • 2. Resources Your farm Creating your team Creating and maintaining your team and culture 2
  • 3. A brief introduction Im a Engineer (MS, Industrial Engineering- Operations Research) turned Genetic Epidemiologist (PhD - Genetic and Molecular Epidemiology) Finding & identifying genes related to getting cancer or treatment response. 2007-2016 Assistant then Associate Professor, University of Buffalo/Roswell Park Cancer Institute 2016 Associate Professor, The Ohio State University 3
  • 4. Genetics and Epidemiology of Myeloid Malignancies (GEMM) Consortium Motivation: 5 year survival rate for Acute Myeloid Leukemia (AML) is 27% Little is known about AML risk factors, including the role of genetics. To overcome this barrier, we formed GEMM to unite investigators world-wide and pool data and resources. 4
  • 5. Determining Influence of Susceptibility COnveying Variants Related to 1 Year survival after Blood or Marrow Transplant (DISCOVeRY-BMT) 5 Motivation: 1 year survival rate for leukemia patients treated with BMT is ~65% HLA Genetics is the biggest predictor of survival, the rest of the genome is underexplored To overcome this barrier, we formed DISCOVeRY- BMT to unite investigators world-wide and pool data and resources to find other important genes and improve survival 5
  • 6. Congratulations you have a PhD! 6 Heres some (a lot of) money AndGO!
  • 7. I did not get a PhD in this Cocktail conversations with successful female scientists Making the Right Moves: A Practical Guide to Scientific Management for Postdocs and New Faculty http://www.hhmi.org/developing-scientists/making-right-move Work Rules! Laszlo Block former SVP of People Operations at Google re:Work- a collection of practices, research, and ideas from Google An effort by Google to help share and push forward the practice and research of data-driven HR https://rework.withgoogle.com/about/ 7
  • 8. Filter 8 -team size -dry lab appropriate -money -practical implementation -reality
  • 9. Team definition 9 People working together in a committed way to achieve a common goal or mission. The work is interdependent and team members share responsibility and hold themselves accountable for attaining the results. MIT Information Services & Technology
  • 10. Building a working farm 10 CC w/C. Ambrosone Harvest Growing Planting Planning
  • 11. Why a farm? Easy to visualize and plan over time. Overall mission is implemented through a series of small/medium/large projects. Excellent framework for presenting results and conclusions (research or analytics). Growth happens through permutations of set resources. Economics. 11
  • 12. What will you grow? Develop a clear vision Developing a shared team vision does not limit innovation. A vision statement provides a foundation for creativity for new directions. 12 HHMI
  • 13. My vision My vision is that we will leverage genetics to identify individuals at risk for hematologic malignancies or disease and significantly impact patient survival after a diagnosis. 13
  • 14. What will you grow? Implement the vision with a mission statement Paint with broad strokes, but also identify key measures of success. Provide reasoned and emotional justification. Tie it to the values and culture of your company. Be clear and honest. Create a future that distinguishes your program from competitors. 14 HHMI
  • 15. 15 Does having a good mission statement really work? Work Rules! (G); Wrzesniewski & Grant Team members who are inspired have increased productivity and happiness. Reframing tasks as opportunities to help others, can make occupations and tasks feel more significant and increase motivation.
  • 16. Me Hires CODING Hire people better than you (at something) A bad hire is toxic and drains your teams energy and creativity. Set the bar high, dont compromise, be willing to wait! Hire someone who is better than you in some meaningful way. Youll end up with a much stronger team. 16 Work Rules! (G)
  • 17. 17 How to hire people better than you (at something) Give people a reason to join your team (mission) Everyone recruits Take a risk: Consider a trade-off by going outside of your immediate field A computer scientist or physicist interested in bioinformatics but without a strong biology background. Someone with a molecular biology background who has started self-training in R or Perl.
  • 18. 18 Take the curious Research presentation Journal Club Chalk talk Multiple interviews across disciplines Long term incentives (aside from $) Additional conferences, trainings and first author papers opportunities. How to hire people better than you (at something)
  • 19. Definition of a Team 19 People working together in a committed way to achieve a common goal or mission. The work is interdependent and team members share responsibility and hold themselves accountable for attaining the results. MIT Information Services & Technology FUNCTIONAL & EFFECTIVE
  • 20. Apollo Syndrome: sum of the parts < whole 20 Dr. Meredith Belbin discovered the Apollo Syndrome, a condition where teams of highly capable individuals, selected using ability and aptitude tests, can collectively, perform badly.
  • 21. Apollo Teams 21 Teams selected using ability and aptitude tests. They were bound to win. The Apollo teams often finished near the bottom.
  • 22. Why did they fail ? Excessive time in destructive debate. Difficulties making decisions, resulting in jobs often being omitted. Team members acted in silos, resulting in teams that were difficult to manage. Some teams recognized what was happening and over compensated by avoiding confrontation. 22
  • 23. What does the Apollo syndrome teach us? 23 HOW A TEAM WORKS MATTERS MORE THAN WHO IS ON THE TEAM AS MUCH AS
  • 24. 24 Modified from Project Aristotle What Makes an Effective Team?
  • 25. Making your team safe, eg lets make a Sharknado part 6 Frame the work as a learning problem, not an execution problem. I may not be understanding this correctly, I may miss something, this is why I hired people smarter than me. Ask a lot of questions about everything , eg model 25
  • 26. Additional ways to help achieve an effective team culture 26
  • 27. Trust Be transparent. Encourage people to think & act like owners. Give your team an uncomfortable amount of freedom. Expect a lot. but verify Adapted from Work Rules! (G))
  • 28. A tale of two tails: best and worst Dont surprise people. Tell them if theyre low performers and help them learn or find new roles Put your best people under a microscope to find outand replicatewhat makes them succeed. Work Rules! (G))
  • 29. Short-term one time incentives: the death of learning Self-Determination Theory (a theory of motivation) -Attaching incentives to tasks reduces intrinsic motivation when incentives are removed. Adapted from Deci & Ryan; Work Rules! (G)
  • 30. Pay unfairly and reward failure The best people are worth more than average people. Reward failure, else you kill risk taking. Work Rules! (G))
  • 31. Stand up meetings Short meetings (10 mins) to start your day. For my lab, these are only on Monday Go around the table: What are you working on? What are your goals for the week (or today?) Are there barriers to accomplishing these goals? How can I help? 31
  • 32. You will do the same or a similar work again, what worked and what did not? 32 A look back: the role of the Retrospective
  • 33. The take home Have a team mission Hire people better than you Splash time (and cash) to recruit Try something new Wait! How a team works is as important as who is on it Trust a scary amountbut verify Looks at your tails Long term not short term incentives Pair unfairly Reward failure 33

Editor's Notes

  • #5: Consortium: 9 studies w/~6000 AML cases /~20000 controls (growing). Genome-wide data are available on everyone, ~9 million single nucleotide polymorphisms (SNPs). Major data collection centers include Ohio State (Data Coordinating Center), Newcastle (UK) and Northwell
  • #6: Consortium: 151 centers , ~3500 leukemia cases and their unrelated donors Genome-wide data are available on everyone The project is in collaboration with the National Marrow Donor Program and the Center for International Blood and Marrow Transplant
  • #13: This is from the HHMI
  • #25: Data used to determine : 3 success metrics , 180 teams , 201+ interviews a, 250 inputs (questions in to 2 categories: what are the team characteristics (eg skills, background etc) and what are team dynamics (what are norms, goals and conflicts do they socialize outside of work) and 35 stat models.
  • #28: .