skip to navigation skip to content
- Select training provider - (Bioinformatics)
Instructor-led course

Provided by: Bioinformatics


This course has 2 scheduled runs. To book a place, please choose your preferred date:


April 2024


[ Show past events ]

Events available

Analysing your own data: a workshop on statistics and experimental design (IN-PERSON)
NewSpecial£


Description

This week-long course is aimed at people with little or no experience using statistical analyses in research. It introduces participants to core concepts in statistics and experimental design, aimed at ensuring that the resulting data is able to address the research question using appropriate statistical methods.

The interactive course gives participants a hands-on, applied foundation in statistical data analysis and experimental design. Group exercises and discussions are combined with short lectures that introduce key theoretical concepts. Computational methods are used throughout the course, using the R programming language. Formative assessment exercises allow participants to test their understanding throughout the course and encourage questions and critical thinking.

By the end of the course participants will be able to critically evaluate and design effective research questions, linking experimental design concepts to subsequent statistical analyses. It will allow participants to make informed decisions on which statistical tests are most appropriate to their research questions. The course will provide a solid grounding for further development of applied statistical competencies.

As a follow-up of this course, we run an extra optional session on 25 April. This is an applied, hands-on session where you can bring your own data and we provide direct support to your analysis. This is exclusively available to participants on this course.


If you do not have a University of Cambridge Raven account please book or register your interest here.

Target audience
  • The course is aimed at people at postgraduate level who are involved in research.
  • Applicants are expected to have a working knowledge of R.
  • The course is open to Postdocs and Staff members from the University of Cambridge, Affiliated Institutions and other external Institutions or individuals
  • Further details regarding eligibility criteria are available here
  • Guidance on visiting Cambridge and finding accommodation is available here.

Bookings and fees

Please note that all participants attending this course will be charged a registration fee. The fees are as follows:

  • Industry participants (non-academic) pay £600
  • Members of the University of Cambridge and participants from other academic institutions pay £300

⚠ Bookings are only approved and confirmed once you have paid the fees in full.

Prerequisites
Topics covered

Statistics, Experimental Design, R

Objectives

During this course you will learn about:

  • One and two sample hypothesis tests
  • ANOVA
  • Simple linear Regression
  • ANCOVA
  • Linear Models
  • Model selection techniques
  • Power Analyses
  • Practices in experimental design that lead to high quality research
  • What to do with more advanced analysis techniques for experiments with unusual or complex designs
  • How to take power analysis into consideration in your experimental design
  • How to implement piloting in your experiments
Aims

After this course you should be able to:

  • Analyse datasets using standard statistical techniques
  • Know when each test is and is not appropriate
  • Link experimental design to your statistical analysis strategy
  • Formulate good research questions
  • Identify common design pitfalls, and how to avoid or mitigate them
  • Operationalise variables effectively
  • Identify and deal with confounding variables and pseudoreplication
Format

Presentations, demonstrations, discussions and practicals

Timetable

All days will run from 09:30 - 17:00.

There will be regular breaks, with coffee/tea and biscuits provided throughout the day.

A lunch break is scheduled around 12:30 - 13:30.

Further information on catering will be provided closer to the start of the course.

Day 1
We start with introducing the statistical inference framework, covering key statistical concepts. How do we use statistics in the context of hypothesis testing in research? We discuss how to choose and define variables and how to align the research question with the expected statistical analysis of the data.


Practicals focussing on simple hypothesis testing and dealing with categorical variables provide hands-on experience.

Day 2
We continue the course by expanding our knowledge on dealing with different types of variables. We introduce the concept of linear regression, which allows us to explore continuous predictor variables. We cover how to deal with a combination of continuous and categorical variables.

Day 3
After covering situations where we have to deal with more than two predictor variables, we introduce basic model comparison techniques. How do we decide what variables to include in our analysis?


We talk about confounding variables; what do we do if there are variables that are related to both the predictor and outcome variables? What ways are there to avoid this in our experimental setup? We also cover other experimental design considerations, such as dealing with pseudoreplication and independence.

Day 4
This day is mostly aimed at understanding statistical power: how much confidence can we have in our statistical analysis and how do we make sure that we have enough data to draw meaningful conclusions? We explore ways to calculate this and discuss experimental design considerations that help balance the need for sufficient data within practical experimental constraints.


We finish the day by putting into context all of our knowledge of the linear model framework that we built up the previous days. We then discuss ways in which we can extend this framework.

Day 5
The final day focuses on introducing common extensions to the linear model framework. This is achieved by worked examples on commonly encountered statistical techniques.


We finish the week with a capstone project. Here, participants are split into groups and use the opportunity to combine the various experimental design considerations with their acquired statistical knowledge. Using example data sets, participants work through the process of defining a research question, taking into account experimental design considerations and required statistical analysis. This is finished by performing the analysis, drawing conclusions and discussing the findings with the group.

Duration

5 days

Themes

Events available