In the vast landscape of statistical techniques, it can feel like navigating a maze to choose the right method for analysing your data. The journey starts with identifying a promising statistical tool, then delving into applying it— a quest that can require diving into books, scouring the internet, or parsing through help files. This endeavor may be slow and somewhat tedious, but it’s all in the name of progress, provided you didn’t embark on the wrong path from the start. The key to not wandering aimlessly? Consulting with a statistician or a knowledgeable colleague, although not everyone has that luxury.

Imagine if, instead, you could tune into a lively, yet straightforward, live video stream where experts like Dr. Alain F. Zuur and Dr. Elena N. Ieno break down the complexities of statistical techniques into digestible, non-technical terms. Learning about a new statistical method, understanding its foundational concepts, and seeing it applied could streamline your analytical journey significantly.

Periodically, we will offer complimentary live Zoom sessions where these experts illuminate the core concepts of specific statistical methods. During these sessions, you are invited to engage via a chatbox, making the daunting task of navigating the statistical maze a little more straightforward and a bit more enjoyable.

Here are the upcoming free live Zoom sessions we have planned. For more details on each session, please refer to the descriptions provided below the form.

Data exploration and visualisation.  

• Date and time: 22 April 2024; 14.00-16.00 UK time.
• We will cover key tools for detecting outliers and collinearity in data. Additionally, we will explore methods to visualize relationships and dependencies. Furthermore, we will explain why creating histograms of the data can be a waste of time.
• These are the steps that you need to apply before analysing your data.

Analysing proportional data using beta and ordered beta GLM

• Date and time: 5 August 2024; 14.00-16.00 UK time.
• We will start by providing a concise explanation of the beta distribution and the ordered beta distribution. Following that, we will illustrate how both distributions can be applied within a Generalised Linear Model (GLM) to analyze proportional data, such as coverage. Additionally, we will demonstrate how the ordered beta distribution can be effectively used to analyzs proportional data that includes zeros and ones.

What is a Tweedie GLM?

• Date and time: 2 September 2024; 14.00-15.00 UK time.
• We will explain the concept of the Tweedie distribution and demonstrate the application of a Tweedie Generalised Linear Model (GLM) to continuous data (e.g. biomass data).

What is a linear mixed-effects model?

• Date and time: Monday 11 November 2024; 14.00-16.00 UK time.
• We will explore subjects like pseudo-replication, random effects, and mixed models, along with topics such as the number of clusters and the number of observations per cluster. Mixed models, also known as mixed-effects models, incorporate both fixed effects (covariates), and random effects. This approach allows for the analysis of complex data with multiple layers of variability, making it particularly useful in studies where data are collected from different groups or time points.
  

Fitting a regression model with spatial correlation

• Date and time: Monday 16 December 2024; 14.00-16.00 UK time.
• In this seminar, we will demonstrate the integration of a spatial correlation structure into a linear regression model using R-INLA. Please note that our focus will be strictly on the implementation process. Given the 2-hour time frame, we won't delve into the specifics of INLA itself. Nonetheless, this session will equip you with insights into both the numerical and graphical outputs generated by a regression model that incorporates spatial dependencies. Should you find the application of INLA relevant to your data, we encourage you to further explore INLA.

Zero-inflated GLM and GLMM for the analysis of count data.

• Date and time: Monday 17 February 2025; 14.00- 17.00 UK time.
• Outline: We start with a brief revision of the Poisson, negative binomial and Bernoulli distributions. We then discuss how these distributions are used in Poisson, negative binomial and Bernoulli generalised linear models (GLM) and generalised linear mixed-effects models (GLMM). We will use one case study to show how one can decide whether a GLM(M) or generalised additive mixed effects model (GAMM) can cope with a data set that contains an excessive number of zeros.

Multivariate GLM using GLLVM

• Date and time: Monday 14 April 2025; 14.00- 16.00 UK time.
• We will explain generalised linear latent variable models (GLLVM). A GLLVM is a GLM (or GLMM) in which multiple response variables are analysed simultaneously, while allowing for dependency between the response variables and also between the observations. 
• We will show a case study.

General information

• Starting April 2024, we will record our seminars and post the videos on our website along with the associated R code and a discussion board for seminar-related questions. For a one-time fee of 25 GBP, you will have unlimited access to these resources for all past, current, and future seminars, as long as we continue to organise these seminars. This fee helps cover our costs, as we incur charges each time a video is viewed.
• All online seminars are non-mathematical presentations.
• We will use Zoom webinar. Once 500 participants have joined the webinar, Zoom will automatically block the entrance. We will start the Zoom webinar at 13.45; just join early to avoid a closed door.
• We will email you the Zoom link for the meeting 1 day before the start of the seminar.