Statistics for the soul.
Text: Tim Schröder
People who experience bullying or sexual abuse during childhood often suffer from psychological problems later in life. Using statistical methods, Giusi Moffa aims to identify which symptoms lead to serious disorders during a person’s lifetime so that psychologists can offer help.
Sexual abuse and bullying are two of the most severe forms of trauma that a child or young person can experience. Unfortunately, both are alarmingly common. According to estimates by Statista, a statistics service provider, 14 percent of all children and young people in Switzerland today experience cyberbullying (as of 2023). Experts estimate that 11 percent of girls and 5 percent of boys are the victims of sexual abuse.
The consequences for a person’s physical and mental health can often continue into adulthood, ranging from persistent fatigue to depression, compulsive behavior, and even paranoia and hallucinations. Many of those affected require psychological help. Treatment can prove challenging because many victims of bullying and sexual violence can experience different symptoms at different points in their life.
One symptom can often evolve into another – for example, anxiety can evolve into depression. “It is extremely difficult for experts to determine which symptoms could lead to further psychological problems later in life. In many cases, the causal chain is unclear,” says Giusi Moffa, Professor of Statistics at the Department of Mathematics and Computer Science, University of Basel. “This makes it difficult for psychologists to find the right therapeutic approach. How can those affected be helped to improve their quality of life? And which symptoms should be treated to prevent even more serious problems from developing further down the line?”
Studies with 6,000 participants.
To provide experts with new ideas for future therapies, Giusi Moffa approached the problem from a mathematical perspective. She analyzed the statistical relationships between 20 different psychological aspects – from difficulty concentrating and insomnia through to paranoia – to form causal chains. The aim is to find out which symptoms are most likely to lead to specific consecutive symptoms later on. To do so, Giusi Moffa, who is also an honorary research associate at University College London, used two surveys of around 6,000 British people aged 16 and over. Participants were asked whether they had experienced bullying and sexual abuse and also about specific symptoms.
Exploring causal relationships.
Professor Moffa then used a specialized statistical method to analyze these data and link the different symptoms with arrows or lines – like a family tree – to create what experts call a graphical model. The statistical machinery delivered a causal diagram that shows which symptom may potentially lead to another.
The trick with this method is to repeat the procedure multiple times and link the different symptoms in a new way each time – as though you were linking the stations of an underground transport network. In total, Giusi Moffa ran the statistical analysis 10,000 times on the data from the British surveys. This created what you might call the largest common denominator of all causal diagrams – a graphical model that collates the results of all 10,000 runs and highlights the strongest links between the various symptoms.
Inspiration for new therapeutic approaches.
The two British surveys on bullying and sexual abuse already included the frequency distribution of the various symptoms – and, implicitly, information on the probabilities with which different symptoms occur together or one after the other.
Using the data to establish a meaningful causal relationship between the various symptoms is, however, another thing entirely; partly because there are a huge number of possible links between 20 psychological aspects. Only statistical processing can bring order to the chaos. The more often the statistical model identified a causal relationship between two symptoms during the 10,000 runs, the darker the line connecting them became on the summary causal diagram.
“Some of the causal relationships surprised us,” says Giusi Moffa. For example, a strong connection emerged between “worry” and “obsessions”. A clear chain was also revealed from “fatigue” to “difficulty concentrating” and “worry” and on to “hallucinations”. “These causal chains are not yet the key to successfully treating the psychological consequences of bullying and abuse,” says Giusi Moffa. “But they can help to develop new forms of therapy. It is quite conceivable that early treatment of fatigue can reduce the likelihood of hallucinations.”
The results provided no such pointers for other psychotic disorders such as paranoia. The causal diagrams show a direct link between the traumatic experience and paranoia. “This offers very few options for potential therapy. It would be different if the paranoia aspect were embedded deeper in the network between different causal chains,” says Giusi Moffa.
Nevertheless, she has high hopes for her study, the results of which have now been published in the journal Psychological Medicine. “We’ve advanced entirely novel methods to look at bullying and sexual abuse in combination, as well as examining how both forms of trauma influence each other in their subsequent symptoms. It would be amazing if we could now use the results to develop new therapeutic approaches.”
Giusi Moffa is a professor of Statistics and leads the research group for Statistical Science at the Department of Mathematics and Computer Science.
More articles in this issue of UNI NOVA (May 2024).