UCL ICH Statistics Courses

When data is collected from 2 binary responses on the same individuals or items then the data may be displayed as a 2x2 table. It is often of interest to determine whether there is some form of association between the two variables. The chi-squared test is most commonly used by researchers to do this. However, there are many other statistics that may be applied to a 2x2 table to provide more useful information. The aim of this workshop is to guide the researcher through alternatives to decide on the most appropriate investigation of their data.

*Cost for external participants. Workbooks, refreshments and lunch vouchers for non-ICH participants (12.45pm - 1.45pm) are included in the price.

This course aims to familiarise the audience with the concepts of Bayesian analysis. Its basic principles (e.g. Bayes theorem) will be explained and how these are embedded in the medical science and research. By the end of this workshop, students should be able to understand and critically evaluate published research that has used Bayesian analysis as well as be confident to implement Bayesian concepts in their own research.

*Cost for external participants. Workbooks, refreshments and lunch vouchers for non-ICH participants (12.45pm - 1.45pm) are included in the price.

The ability to critically appraise published research is a key skill for all researchers. In this workshop we discuss the importance of defining the research question and the collection and evaluation of research to answer the question as posed. Checklists are introduced and their implementation in the evaluation of individual papers illustrated. Participants will be given papers to evaluate in a practical session.

*Cost for external participants. Workbooks, refreshments and lunch vouchers for non-ICH participants (12.45pm - 1.45pm) are included in the price.

This is a course run by Katie Harron. For more information about Katie please visit the "About us" tab of the homepage. 

This short course is designed to give participants a practical introduction to data linkage and is aimed at researchers either intending to use data linkage themselves or to analyse linked data. Examples of the uses of data linkage, data preparation, methods for linkage (including deterministic and probabilistic approaches) and issues for the analysis of linked data are covered. The main focus of this course will be health data, although the concepts will apply to many other areas.

This course includes a practical example involving data to be linked, to enable participants to put theory into practice. Participants will need to bring their own laptops with software Link Plus installed on them. This is developed by Centers for Disease Control and Prevention (CDC) and is freely available from  http://www.cdc.gov/cancer/npcr/tools/registryplus/lp_tech_info.htm

On completing the course, participants will

• Understand the background and theory of data linkage methods
• Perform deterministic and probabilistic linkage
• Evaluate the success of data linkage
• Appropriately report analysis based on linked data

*Cost for external participants. Workbooks and refreshments are included in the price.

Missing data are very common in research studies. By ignoring missing values invalid conclusions may be drawn. This workshop tries to provide guidance on how to deal with missing values and the best ways of analysing a dataset that is incomplete.

Reasons for missing data, types of missing data (‘missing completely at random’, ‘missing at random’ and ‘missing not at random’), simple, ad-hoc analysis methods (‘analysis of completers’, ‘pairwise deletion’, ‘imputation of simple mean’, ‘imputation of regression mean’, ‘dummy variables adjustment’) and more complex methods (‘simple and multiple stochastic imputation’, ‘weighting methods’) are described and discussed.

On the 2nd, optional, day of the course, the theory is put in practice with the use of SPSS (v.17 or later) and real life datasets with particular emphasis on multiple imputation. Delegates are asked to bring their own laptops if they have SPSS installed on them and to share with others that do not have SPSS.  

*Cost for external participants. Workbooks, refreshments and lunch vouchers for non-ICH participants (12.45pm - 1.45pm) are included in the price.

** Second day is optional

Binary (proportion/percentages) outcomes are common in medical and scientific research. This course covers:

- odds ratios as a means of comparing binary outcomes between two groups
- how logistic regression allows for other factors within this comparison
- the basics of logistic regression
- model selection and goodness-of-fit with examples from a variety of applications
- interpretation of SPSS output
- discussion of extension to the analysis of ordinal outcomes

*Cost for external participants. Workbooks and refreshments are included in the price.

This is a NEW course run by George Michaelides. For more information about George please visit the "About us" tab of the homepage. 

The aim of this workshop is to provide an introduction to multilevel research design and data analysis using the R statistical environment. This workshop will enable participants to

(a) develop an understanding of multilevel research and appreciate the potential of multilevel modelling for theory development 

(b) better understand published research that uses multilevel techniques

(c) use the R statistical environment to analyse their own multilevel models (via libraries: nlme and lme4)

The workshop is intended for anyone interested in data analysis of multilevel data (e.g. school-classroom-student or hospital-ward-patient). Participants are not required to have any prior knowledge of multilevel modelling but some knowledge of basic statistical techniques (primarily linear regression) and the R statistical environment is required. Delegates will be expected to bring their own laptops with R installed (step-by-step installation guidelines will be sent to all participants 2 weeks before the course).

*Cost for external participants. Workbooks and refreshments are included in the price.

This is a NEW course run by Marco Geraci. For more information about Marco please visit the "About us" tab of the homepage. 

Quantile regression (QR) is a powerful statistical technique for the analysis of conditional quantile functions - models in which quantiles of the conditional distribution of the response variable (e.g., the median or the 5th centile) are modelled as functions of a set of predictors. QR is capable of handling complex effects on the location, scale and shape of a distribution. Unlike mean regression, no distributional assumption about the error term is needed. This introductory course to QR will enable participants to estimate and evaluate conditional quantile models. A tutorial to perform QR with R will be given in the afternoon. Participants are encouraged to bring their laptop with R (cran.r-project.org/) and the package quantreg (cran.r-project.org/web/packages/quantreg/index.html) pre-installed. 

At the end of this course participants should be able to:

- Define the quantiles of a continuous variable conditional on a set of predictors.

- Characterise the location, scale and shape of the distribution using quantile-based statistics.

- Estimate quantile regression coefficients and confidence intervals.

- Test statistical hypotheses on regression coefficients or related quantities, such as scale and shape indices.

- Interpret the results of a quantile regression analysis.

- Discuss conclusions and implications of findings.

- Perform QR analysis using common statistical software (R or Stata).

Familiarity with R, probability and regression analysis is expected from the delegates. Time permitting, Stata commands will be presented.

*Cost for external participants. Lunch for non-ICH participants (12.45pm - 1.45pm), workbooks and refreshments are included in the price.

This course is aimed to researchers that want to learn how to use the statistical software R in order to conduct statistical analysis. Delegates will be expected to bring their own laptops with R installed (step-by-step installation guidelines will be sent to all participants 2 weeks before the course). Moreover delegates will be expected to have a basic understanding of common statistical tests and concepts as these will not be taught during this workshop. The topics covered using R will include:

- Input of data sets
- Summary measures
- Graphical displays
- Statistical tests (parametric and non-parametric)
- Add-ons and packages in R
- Help in R And time permitting (notes will be given regardless):

- Regression analysis
- Sample size and power calculations

*Cost for external participants. Workbooks, refreshments and lunch vouchers for non-ICH participants (12.45pm - 1.45pm) are included in the price.

This workshop introduces various forms of regression and their usage:

- multiple linear regression (discussed in detail)
- interpretation of SPSS outputs
- choice of model
- assessing adequacy of model fit
- identification of outliers The course will be of use to users of alternative statistical packages as the issues covered are of generally applicable.

*Cost for external participants. Workbooks and refreshments are included in the price.

This is a basic course in medical statistics that covers the following topics:

- introduction, types of data and graphical displays
- summarising data
- reliability and validity
- confidence intervals and p-values
- hypothesis testing
- proportions
- non-parametric tests
- categoric data
- confounding factors

**Cost for external participants; £650 early-bird fee if registered one month or more in advance. Workbooks, online material, refreshments (and lunch for non-ICH participants; 12.45pm – 1.45pm) for every day are included in the price.

The concepts of sample size estimation are introduced. Both precision and the traditional power calculations are covered. Sample size estimation for:

- means
- proportions/ percentages
- time to event data
- rates
- measures of agreement
- centiles Excel sheets to perform the calculations are provided on a take-home CD.
This workshop will take place in a computer room where applicants will have the opportunity to use the computers provided in order to gain full benefit from the workshop practical exercises that utilize the excel sheets.

*Cost for external participants. Workbooks, CD and use of local computers are included in the price.

This is a practical course that will introduce delegates to SPSS and its uses for statistical analysis through hands-on experience. The course is suitable for individuals who are new to SPSS or who feel they would benefit from a refresher course, and will cover the following topics:

• Getting started with SPSS files
• Data entry: manually and from excel
• Editing and organising datasets
• SPSS options and functions
• Descriptive statistics
• Graphical displays
• Parametric significance tests
• Nonparametric significance tests

The course will take place in a cluster room, with access to a computer and version 21 of SPSS throughout the day. Delegates are welcome to bring their own laptops if preferred, but everyone wishing to bring their own computer should ensure the software is licenced before attending. Where possible, we recommend using a recent version of SPSS (e.g., 19-21) for maximum compatibility with the notes provided during the course.

*Cost for external participants. Workbooks and refreshments are included in the price.

This course is aimed at professionals who want to be able to understand the fundamental principles of statistics and to be able to conduct their own analyses. The latter will be explored via the use of the free statistical software R. Topics that will be covered are as follows:

- overview of quantitative research study designs
- types of data
- graphical displays
- summarising data
- confidence intervals
- hypothesis testing (parametric and non-parametric tests)
- p-values
- analysis of variance
- regression analysis (time permitting – notes will be provided regardless) Delegates will be expected to bring their own laptops with R installed (step-by-step installation guidelines will be sent to all participants 2 weeks before the course).

This course is a combination of the Introduction to Research Methods and Statistics and Introduction to R for Statistical analysis courses. Individuals are advised to attend either this 3 day course or the 5 day and 1 day courses.

*Cost for external participants. Workbooks, refreshments and lunch vouchers for non-ICH participants (12.45pm - 1.45pm) are included in the price.

This workshop gives an introduction to time to event (survival) data for the non-statistician, covering:

- use of Kaplan-Meier plots
- life tables
- Cox regression analyses
- interpretation of SPSS output

*Cost for external participants. Workbooks and refreshments are included in the price.