Statistician and Scientist: A New Age in Collaboration
Presenter: Katya Ruggiero
When: Thursday, July 14, 2016 Time: 11:00 AM - 12:30 PM
Room: Lecture Theatre (Level 1)
Session Synopsis:
Before there was data there was�
...an important scientific question that needed answering: Might pigs fly? The question led to an idea: We should carry out an experiment to identify the molecular mechanisms that make pigs fly. And so it came to pass: The idea was enacted; an experiment was carried out; samples were taken from pigs across all the lands and sent to laboratories far and wide for analysis on big, expensive machines. There was a big cheer when, finally, the machines generated what everyone had been waiting for: DATA!!!! And everyone lived happily ever� �Uh-oh, what do we do with the data?� Every year many millions of dollars are spent across the globe in collecting data that cannot answer the questions for which they were intended. The problem is that, with the exception of simple experiments, knowing which statistical design is �best� is nearly impossible without expertise in the subject matter. But, imagine a portal into the mind of an expert guide in the statistical design of experiments, reasoning his/her way through the process of generating an objective-aligned data collection protocol. We will demonstrate our collaborative efforts towards creating this portal via our web-based personal assistant for designing statistically sound experiments.
Statistician and Scientist: A New Age in Collaboration
Presenter: Adam Butler
When: Thursday, July 14, 2016 Time: 11:00 AM - 12:30 PM
Room: Lecture Theatre (Level 1)
Session Synopsis:
Developing effective collaborations with scientists: experiences of an applied environmental statistician
Statistical methods and ideas play an absolutely central role within ecology and the environmental sciences, and these disciplines, in turn, present a wide range of interesting and challenging problems for research statisticians. The development of effective collaborations between statisticians and scientists is therefore critical. My job as an applied environmental statistician involves working at the interface between statistics and the environmental and ecological sciences - through statistical consultancy work (the provision of advice to scientists on the use of established methods), applied statistical research (the development of novel methods in collaboration with applied scientists), and, most commonly, through activities that lie somewhere inbetween these two roles. In this talk I will outline some of the general lessons that I have learnt from my experiences in working at this interface, and will illustrate these using specific case studies that relate to my own work. Key themes that I will focus on will be: 1) the importance (and difficulty) of clearly formulating the ultimate objectives of the work in a common language that is understandable to both scientist and statistician; 2) the value in understanding the pressures and constraints that scientists encounter within their work, and the impact that these can have upon the way that they use statistics; 3) possible strategies to use for managing academic disagreements, and for reaching compromises that are satisfactory for both statistician and scientist; 4) the need to recognise and communicate (but not overstate) the bounds of our own knowledge, and the knowledge of the wider statistical community; 5) the benefit from building long-term working relationships with scientists in particular areas, and in finding effective ways to maintain and develop these relationships. Many of the case studies that I will use relate to the analysis of data on the occurence, abundance, behaviour and phenology of bird populations, since this is the main area of application that I have worked in to date, but I will mention case studies that concern a range of other applications (including the spread of invasive species, and the analysis of trends in insect abundance). I will conclude the talk with a discussion of what I see as being the key challenges for the future in developing and maintaining good collaborative relationships with scientists.
Statistician and Scientist: A New Age in Collaboration
Presenter: Fred Van Eeuwijk
When: Thursday, July 14, 2016 Time: 11:00 AM - 12:30 PM
Room: Lecture Theatre (Level 1)
Session Synopsis:
Regularization Methods In Modelling Genotype To Phenotype Relations, A Synthesis Of Statistical And Biological Principles
Close collaboration between statisticians and biologists stimulates progress in statistical and biological science. New statistical methods are often developed to help answer new biological questions or deal with new biological data generating techniques. A highly interesting field from this point of view is that of plant breeding and genetics. Here, the strong increases in data volumes following from the application of high throughput genotyping and phenotyping techniques ask for new statistical methods, principally to deal with the n<
