Self-reports play a central role in psychotherapy. During therapy sessions or through clinical questionnaires, patients are asked to reflect on their experiences and feelings over the previous days or weeks, and to describe how these relate to one another. Questions might include: ‘When do your symptoms occur?’ or ‘What affects your mood?’ Their responses often form the basis for diagnosis and help guide treatment decisions.

However, memories do not always reflect what people actually experience in everyday life. To help prevent erroneous conclusions in clinical practice, Dr Julia Jeannine Schmid, senior assistant at the Faculty of Behavioural Sciences and Psychology, and her research team investigated how accurately retrospective assessments capture the relationships between behaviour, emotions and symptoms.

The study involved 280 participants: 118 individuals with major depressive disorder, 47 with social phobia and 115 controls. Over the course of one week, participants used their smartphones to record their current experiences, emotions and symptoms six times a day. At the end of the week, they were asked to assess retrospectively how strongly pairs of these experiences had been correlated.

Retrospective assessments overestimate connections

Comparing the real-time recordings with the retrospective assessments revealed clear differences. Looking back, participants consistently reported stronger links between symptoms and emotions than were evident in their day-to-day experiences. For example, they perceived poor sleep quality and negative mood to be much more closely connected than the real-time data suggested.

This tendency was most pronounced among people with depression. They rated the relationship between depressive symptoms and health-related behaviour as moderate, whereas the real-time data showed no meaningful association.

Beliefs shape memories

The researchers conclude that retrospective assessments are influenced not only by people's actual experiences but also by their beliefs about how behaviour, emotions and symptoms are related. For example, someone may assume that stress inevitably leads to unhappiness, even if this pattern is not reflected in their everyday experiences.

The findings suggest that retrospective self-reports and real-time measurements capture different aspects of psychological experience. Retrospective assessments tend to reflect people's subjective beliefs, whereas smartphone-based real-time data provide a more accurate picture of how behaviour, emotions and symptoms co-occur in everyday life.

Implications for psychotherapy

Taken together, the findings highlight the value of combining retrospective self-reports and smartphone-based real-time measurements. Rather than treating the two approaches as interchangeable, clinicians can draw on the complementary strengths of each to tailor psychotherapeutic treatment more effectively to individual patients.

Julia J. Schmid et al.
Beliefs about relationships between symptoms and experiences in daily life: Evidence from experience sampling and self-report in individuals with or without major depression or social phobia
Behaviour Research and Therapy, 2026
Open Access Article