Relaxation program sessions had 3 phases: initial dialogue (2 min); a main section (25 min); final ritual (3 min). During the main section, participants listened and observed the therapist, who described and demonstrated all the exercises of the session, which were focused on Jacques Choque Method (Choque, 1994).

Loose Parts Play program sessions had 3 phases: initial dialogue (2 min); main section (25 min); final ritual (3 min). During the main section, participants were allowed to play freely with any materials (loose parts) available in the playground.

In the combined program, the sessions had 4 phases: initial dialogue (3 min); loose parts play moment (20 min); relaxation exercises moment (5 min); final ritual (2 min).

No intervention. Maintained their usual routines.

Sessions focused on relaxations activities (Jacques Choque Method).

Emotion discrimination was measured through the Emotion-discrimination Task. This measurement consists in two performance tasks of increasing difficulty. In the task 1 (neutral condition) children were tested on their ability to discriminate between cars and flowers, and faces with hats versus faces with glasses. This neutral condition is not scored, since is used to check if children understood how to sort different cards. In task 2, children had a sheet in front of them with a sample drawing of one category in the top left corner (e.g. an angry face) and a drawing of the other category in the top right corner (e.g. a sad face). Then, children have six cards (with the same emotion) and one by one they have to place each card in the correct category after a demonstration. The cards that were placed correctly were counted with 1 point, with a maximum of 3 points.

Emotion identification was measured through the Emotion-identification Task. The task consists of eight drawings of facial emotion expressions, two for each emotion (happiness, sadness, fear, and anger). Then the researcher show two sheets with four drawings of facial emotion expressions on each sheet and ask to the children: "Who looks happy?", and children have to point to the drawing with the correct facial expression. Next, the researcher ask if "Is there anybody else who looks happy?". This procedure is repeated for anger, sadness, and fear. The number of emotions correctly identified was recorded, with a maximum score of 2 per emotion.

Emotion recognition was measured thought the subscale Others' Emotion Recognition (6 items) of the Portuguese version of the Emotion Expression Questionnaire. This Subscale indicates the extent to which the child can recognise the parents' and others' emotions. Parents can rate to what degree each item is true on a 5-point scale (0=(almost) never, 1=rarely, 2=sometimes, 3=often, 4=(almost) always).