Formal and computational modelling to advance health psychology theories: Conceptual and practical considerations
James Allen and Olga Perski have expertise in formal and computational modelling, communicating these techniques to non-specialists and development through stakeholder involvement. We have developed a formal model of smoking relapse (https://osf.io/preprints/osf/wx8s5 _ v1) and a scoping review of modelling practices for theory improvement (https://doi.org/10.1080/17437199.2024.2400977).
A key barrier to progress in health psychology is the lack of specificity in our theories (the ‘theory crisis’). Formal and computational modelling (representing a theory using a mathematical framework and implementing it using computer code) is a promising way forward, providing novel ways to think about behaviour change. However, few health psychologists have formal modelling skills and these are difficult to learn without an introduction. Here, we will introduce key formal modelling concepts and a relevant guiding framework for their development. Participants will, through working together to formalise a popular theory, gain hands-on experience of formal modelling.
Activities throughout the workshop will consist of a mix of expert talks, facilitated discussion and hands-on implementation of some simple computational models. Before lunch we will introduce the key motivations and concepts for the day, before leading a whole group discussion to select our theory to model and develop maps of key concepts and how they are linked. In the afternoon we will present the key requirements for a formal model, before providing hands-on instruction on how to begin modelling our chosen theory based on the discussed linked concepts. Finally, we will facilitate discussion to ensure that this work is continued after the workshop has concluded.
Morning – formalisation of selected model
Talk: Intro to the theory crisis
Talk: Intro to formal modelling and framework
Discussion: Theory selection and systems mapping activity. Core theory components, and how they are linked, will be specified through small- and wider-group discussion.
Afternoon hands-on development of subset of theory
Talk: Mathematical formalism examples.
Activity: Implement model subset in R and generate outputs.
Discussion: Future plans for continued collaboration.
Participants will understand the issues with current psychological theories, and be familiar with available solutions from formal modelling to a point where they can be applied in their own work. Through worked examples, participants will create a formal model of a subset of a collaboratively agreed upon theory. Having fully developed a model of a subset of the theory in the workshop, this will form the basis for continued collaboration.
Up to 20 researchers and practitioners with an interest in improving psychological theories. No prior knowledge of formal modelling is required.
There is no conflict of interest. All materials and software introduced are open source.
How to apply single-case studies for clinical questions in health psychology.
Marij Zuidersma, University of Groningen, University Medical Center Groningen, the Netherlands
Marij Zuidersma, University of Groningen, University Medical Center Groningen, the Netherlands
Harriëtte Riese, University of Groningen, University Medical Center Groningen, the Netherlands
Scientific evidence is often derived from group-based (or ‘nomothetic’) studies, yielding results that apply to the ‘average’ person. Yet, we are typically interested in answers to clinical questions that apply to individual persons. Due to great differences between individuals as well as the variation within individuals in complex temporal relations between symptoms and behaviours, results of group-based studies do not apply to individuals. Consequently, they often yield inconsistent findings and small effect-sizes. The single-case design (SCD) is increasingly used for making inferences about individual persons regarding the nature, onset, and treatment of their symptoms. This workshop will educate participants about the relevant aspects of SCD.
After this workshop, participants can:
Describe advantages and disadvantages of SCD.
Design and execute an SCD.
Apply visual analysis to analyse data from a single-case experimental design (SCED) and apply ARIMA time series analysis to analyse data from a single-case observational design (SCOD).
Find relevant SCD-resources.
Session 1: Interactive presentation: SCD: why, when, what and how? Participants will be actively engaged by providing them with assignments.
Session 2*: Practical assignment: participants analyse SCOD-data with Autoregressive integrated moving average (ARIMA) modelling
Session 3*: Practical assignment: participants analyse SCED-data with visual analysis.
Session 4: Presentation about important aspects of SCD, including analysis techniques, pooling of results of multiple participants, limitations, guidelines for appraisal and reporting, ethics, and an overview of useful resources.
*For sessions 2&3 it is necessary that participants bring their own laptop with R installed on it with them.
Researchers, clinicians and anyone else interested in learning more about SCD.
Maximum number of participants:
Short statement about expertise facilitator 1: Marij Zuidersma
Marij Zuidersma is a senior researcher at the University Centre Psychiatry at the University Medical Centre Groningen, University of Groningen, the Netherlands. She has expertise on SCD methodology. She designed, executed and analysed work from different SCD’s, including both SCOD and SCED. She supervises other researchers in performing SCD, and gives a course, lectures and workshops about SCD. She collaborates with different local and international research groups that are experts on SCD, including the board of the International Collaborative Network for N-of-1 Trials and Single-Case Designs (ICN).
Short statement about expertise facilitator 1: Harriëtte Riese
Harriëtte Riese is an associate professor at the University Centre Psychiatry at the University Medical Centre Groningen, University of Groningen, the Netherlands. She is an expert in monitoring in daily life with an interest in clinical implementation of innovations obtained from her research. Her expertise lies in advancing personalized mental health care through SCOD tools like the eDiary PETRA. She has given lectures at a SCD course.
