Effective Statistical Consultancy

Description

Overview: This class provides students with an introduction to statistical consultancy and the opportunity to apply the statistical analyses covered in the other MSc courses to real-life problems. The class will teach students how to engage with professionals working in business, industry and the public sector, apply their statistical knowledge in different situations and effectively interpret and communicate the results of a statistical analysis to non-statisticians working in these organisations. The class will cover clinical trial design and analysis, critical appraisal of scientific papers and of statistics in the media, and will involve interactive discussion sessions on fundamental issues in interpretation and communication of statistics. The problems covered in this class will come from real consultancy projects. The class will be taught as a 10 week block in semester 2. There are 20 contact hours and the teaching will be on the basis of lectures and interactive tutorials. There will be 10 hours of lectures (1 hour per week) with 2 hour tutorials every second week. Lectures will be the main means of conveying and explaining the material with one lecture possibly involving an external speaker. Tutorials will be a mixture of practical workshops and discussion groups. Support will be available from lecturer at these times. Students will be required to work on the assessed group project outside of the allocated contact time and to coordinate the work within their group. Syllabus: 1. Fundamentals of statistical consultancy 2. Preparing to consult 3. Study design including clinical trial design 4. Sample size calculations 5. Quoting, SAPs, SOPs and CRFs 6. Dealing with data 7. Reproducible Analyses: Coding data and preparing an analysis data set 8. Visiting consultant statistician 9. Reporting data – table and figures 10. Medical research ethics Learning Outcomes: The class provides opportunities for students to develop and demonstrate knowledge, understanding and skills in the following areas: i) Knowledge Based Outcomes: On completion of the class students should be able to: - identify the appropriate statistical analysis for different types of data and processes - know the importance of using reproducible code in all statistical analyses and in having a clear pipeline from raw data to analysis data set - undertake the appropriate statistical analysis for different types of data, including producing confidence intervals - produce sample size calculations for different research studies - interpret statistical analysis results for different types of data and in different contexts - understand the importance of defining clear timelines and deliverables ii) Skills Outcomes: - On completion of this class students should be able to demonstrate: - the ability to analyse real-life data and interpret the results from such analyses - the ability to write clear reproducible code for an analysis - the ability to communicate the results of analyses to non-statisticians - skills in writing statistical reports and giving statistical presentations - skills in team working - competency in using the statistical software packages such as R. Assessment: Assessment for this class will be based on coursework (75%, comprising a one hour class test (25%) and a group project (50%)) and an individual presentation (25%) on aspects of the project. The group project will involve practical sessions in which students meet with a ‘client’, design a study, analyse the data and present the results as a formal report, submitted by the group. Individual reporting and interpretation of the group project will be assessed via an individual oral presentation. Peer review will also be included in the group assessment process. The class test is scheduled for Week 7 and the group project is distributed in week 6 with submission at the end of week 10. The individual presentations are of duration 5 minutes for each student and will take place on the Wednesday afternoon in week 11.