The mental health field has grown significantly in recent times, with people now being more aware of its importance. This growth can be seen, for instance, in the increased number of mental health applications and mood trackers available in mobile application stores. However, these mobile applications are largely directed towards a wide main-stream audience, and rarely seem to have support from professionals in the psychology or psychotherapy field to address a population with mental health issues (Caldeira, 2017). Mental illness refers to a wide array of mental disorders that affect the way people function psychologically, reflected in their thoughts, emotions, and actions (Thieme et al., 2015).

To track mood means to register different moments of a day or week regarding how a person is feeling; however, most mood trackers function on the basis of single words (emotions), which can be limited in terms of expression. Mood differs from emotions -- which are short-lived reactions to stimuli (Desmet, 2015) --, and can be composed of several different ones that vary throughout a time period. Mood is interesting to study and to record in the context of mental illness because it directly affects people’s well-being (Desmet, 2015).

In this paper we focus on mood trackers, and particularly on using images, rather than text -- hitherto the most prevalent modality of tracking mood --, to create a mood tracker for users currently going through a psychotherapy process. We decided to focus on images that could have several connotational meanings to assist users that may have a more difficult, yet more rewarding reflection of their own emotional state.

Images, in addition to having a literal (denotational) message, also have a symbolic (connotational) one. This connotational meaning usually comes from people’s pre-existing knowledge and life experiences (Joly, 2019), and is precisely why images with more possible connotations could have an important and lasting effect on how mood trackers can improve therapy, both by providing conversation starters between therapist and patient and by allowing the patient to use them as self-reflection tools.

With the aim of developing a new mood tracker in co-design with clinical therapists, to create opportunities for mental illness patients to reflect and think about psychological internal states, during the time between therapy sessions, we carried out two studies to explore images as tools for reflection.

The studies had 20 and 10 participants, respectively. In the first study we asked participants to look at 36 images and choose from a list of emotions which they most associated with each one. This allowed us to reduce the number of images used in the second study to 17 by eliminating those with less emotions associated with them; every image with an agreeability above 40% was cut out, because this meant that they had less connotational meanings. In the second study we opted to use the Geneva Emotion Wheel Version 2.0 - GEW (Sacharin, 2012), a well-known and well-established tool, to further search the connotational options of these images. We instructed participants to choose up to 3 pairs of emotions to associate with each image.

To analyse the results, we evaluated the number of pairs of emotions that had been selected for each of the 17 images. The results varied between 7 and 15. Since the GEW has a total of 20 pairs of emotions, if we were to take the same approach as before and eliminated the images that had below 40% agreement, we would only eliminate one of the images. Meaning that both studies have agreeability in the several connotational meanings of each of the images.

Results indicate that the images considered to represent more abstract “ideas” seem to have a bigger range of connotations associated with them. We believe that the use of images in the context of mood tracking applications can be beneficial towards a better progress within therapy.

This study is part of a project which will take an interaction design approach to mood trackers to better accommodate them into the context of therapy and help patients to deal with the time between therapy sessions, through a mobile application.

The main limitation of both studies is the small samples. We plan to conduct further research with larger samples of individuals undergoing therapy to understand whether the use of connotational images is useful as a tool of self-reflection and if it helps broaden the dialog with the therapist.