Interaction microanalysis of foster care research using THEME

Abstract

Environmental stress is a key element to the understanding of the psychopathology of children in foster care. Such children often present a wide range of symptoms from anxiety to depression, including abnormal behaviors in their interactions with adults that can be related to experience suffered in their family of origin (e.g., abandonment, abuse, etc.). Foster care should provide a safe environment, both to protect children from abuse and to help them build a well-adjusted developmental trajectory. The relationships with the family of origin may also be maintained. How do children in foster care behave in relation to caregivers given the differences between the families they grow up in? This study focuses on three adult-child relationships: those with a foster carer, a mother and a father. Each adult-child interaction was recorded several times in a day-to-day environment. On each occasion the instruction was given to behave naturally while interacting with a child. No additional material was supplied. Our observations concern the verbal and non-verbal comportment of a 4-year-old foster child named Julia when entering the study, with her caregivers. Once the principal elements had been coded (behaviors, verbalizations), a sequential behavioral patterns analysis was performed using the THEME© program. For this purpose, a 2-min interaction was chosen from the third video of an event which appeared particularly representative of the relationship between Julia and her different caregivers. According to whom Julia was with, the results reveal very different interactive processes. We observe, for example, that with the foster carer the interaction patterns were primarily focused on play objects, whereas they involved more collaborative activity with the father and distraction/avoidance behaviors with the mother. The study identifies the use of disengaging and self-exciting behaviors in all types of interaction. Those emotion regulation strategies are particularly developed during parent-child sessions, showing pathological processes of relationship.

Keywords: clinical psychology; emotion regulation; foster care; interaction analysis; pattern detection.

Figure 1

Interaction initiated by Julia in %.

Figure 2

Julia-Foster Carer interaction pattern. ju: Julia, fc: foster carer. The X-axis shows the number of behavioral units occurring within the selected 2-min interaction (129 acts overall). Green lines highlight the disengagement behaviors and blue lines are related to self-excitement attitudes. The numbers appearing in the left side of the figure are related to an identification number of each node and is not related with the appearing occurrence of each behavior.

Figure 3

Monte Carlo test results. Shows the results of T-shuffling (blue) and T-rotation (red) for TPA of Julia and her foster carer interactions where 282 patterns were found. The difference between real and randomized data is important and no are found in the randomized data with the longer patterns.

Figure 4

Julia-Mother interaction pattern. ju: Julia, me: mother. The X-axis shows the number of behavioral units occurring within the selected 2-min interaction (111 acts overall). Green lines highlight the disengagement behaviors and blue lines are related to self-excitement attitudes. The numbers appearing in the left side of the figure are related to an identification number of each node and is not related with the appearing occurrence of each behavior.

Figure 5

Monte Carlo test results. Shows the results of T-shuffling (blue) and T-rotation (red) for TPA of Julia and her mother interactions where 293 patterns were found. The difference between the real and randomized data is great and with longer patterns, none are found in the randomized data.

Figure 6

Julia-Father interaction pattern. ju : Julia, pe : father. The X-axis shows the number of behavioral units occurring within the selected 2-min interaction (114 acts overall). Green lines highlight the disengagement behaviors and blue lines are related to self-excitement attitudes. The numbers appearing in the left side of the figure are related to an identification number of each node and is not related with the appearing occurrence of each behavior.

Figure 7

Monte Carlo test results. Shows the results of T-shuffling (blue) and T-rotation (red) for TPA of Julia and her father interactions where 169 patterns were found. The difference between real and randomized data is great and no pattern are found in the randomized data when pattern length increase.