You are here
Introductory Medical Statistics - one-day virtual course
17 November, 2020
Registration is now closed.
Applicants can be put on a waiting list on request.
The programme is accessible further down this page.
Contact Magda Wheatley for further information.
- Click here to join our courses and events mailing list
- Online registration is by card only - this is our preference. NB: we encourage people to pay by card where possible. If a customer would like to receive an invoice from the College and pay through this route, a purchase order number must be provided to Imperial College. Course details will only be sent upon payment being received by the College. This process might be a lot longer and normally takes up to several weeks. Therefore, payment by card is much quicker and easier if you would like to avoid delay.
- Imperial College staff and students can pay by internal transfer; if so, you do not need to use the booking link. Please just supply your details and a project/grant code. Please contact Magda Wheatley about this.
Registration fees (VAT-exempt):
- MSc/PhD students: £110 - proof of student status may be required prior to registration
- Academic/NHS: £175
- Corporate/other: £225
* Please note that this course will NOT be recorded. *
Designed to introduce anyone who uses statistics in their work or research to the following:
- Basic epidemiological concepts (hierarchy of evidence and differences in study designs; confounding in observational studies vs. RCTs
- Descriptive statistics for quantitative, ordinal and qualitative data (mean, median and mode; standard deviation, percentiles and frequency distribution)
- Inferential statistics: estimating parameters in the population (confidence intervals)
- Testing a hypothesis (p-values; choosing a test; types of errors – false positive and false negative results; multiple testing)
- Correlation vs. simple linear regression to test relationships between quantitative variables (differences in aims and links between the two approaches; simple linear regression vs. ANOVA)
- Multiple linear regression to adjust for confounding; Interpretation of findings; examples of the impact of confounding on estimates of interest
- Different measures of risk (binary outcomes): relative measures of risk (odds ratio, relative risk, hazard ratio); absolute measures of risk (risk difference, NNT/NNH)
- Simple and multiple logistic regression (binary outcomes): interpretation of findings; examples of the impact of confounding on the estimates of interest
- Power and sample size calculations: why we need them and what parameters we need to perform them ; examples of sample size and power calculations for continuous and binary outcomes
The course will conclude with a practical session: revision and discussion of concepts presented in the course using real examples. Feedback with answers to questions through online voting.
Suitable for - Doctors, nurses, clinical research fellows and postgraduate students
Accreditation by the Royal College of Physicians currently sought
Feedback from most recent (two-day, face-to-face) course, in November 2019 -
"Absolutely great course! All speakers were good and helpful."
"All presentations excellent. Very good speakers; concepts explained very clearly."
"Enjoyed the course a lot. Very well prepared and presented."
"Excellent course – I will recommend to colleagues."
"Handout materials all very comprehensive and easy to understand. The booklet will be a very useful reference."
"The handout is arranged in a well-organised order. It is easy to follow the chapters taught."
"The practical was a great way of reinforcing concepts from the lectures! Really liked this/found it helpful!"
"Very helpful study days – thank you."
Credit for main photo: Carlos Muza on Unsplash
Images from our November 2019 course (credits: Diana van der Plaat):
James Potts introduces the first of his two practical Data Analysis sessions; Alex Adamson presents 'Power and sample size'; Winston Banya presents 'Basic concepts of survival analysis'; the paper critique session in progess.