(PS4-2) A Validity Check of a Novel EMA Application for Young Children

The clinical assessment of children and adolescents has long been limited by informant discrepancies and the inability to interpret and integrate the discrepancies of child’s self-report and parent, teacher, and clinician reports (De Los Reyes et al., 2021). No single informant seems to be able to predict treatment outcomes above and beyond the combination of all informants (Wergeland et al., 2023). It is hypothesized that this is caused by the lack of contextual and dynamic considerations in these reports (De Los Reyes et al., 2021). Ecological momentary assessments (EMA) are a potential solution to collecting dynamic, context-sensitive self-report data that may provide insight into why informants consistently differ and how we should integrate these reports. This study piloted a novel EMA application for young children (4-10 years old) to test the usability and feasibility. The original authors determined whether children reported behaviors faithfully as compared to their parents, if they could respond consistently, and if they used the entire scale for each emotion (Milojevich et al., 2023). Our analysis served to answer the question: “Do their answers make sense?”. We hypothesized that children’s responses would be coherent: when they had encountered a negative behavior that day, they would feel mad, sad, and/or tired and not happy or excited. 

Since we were only interested in the presence of an emotion, and not the degree, we converted the likert emotion scales to a binary measure with 1 = 0 and 2, 3, 4, 5 = 1. Incidence comparisons were calculated using a Poisson distribution of the presence:absence of an emotion relative to the presence:absence of a negative behavior.  

Contrary to our hypothesis, children reported negative-only emotions in the presence of a negative behavior only 8% of the time. To investigate whether children were generally reporting more negative emotions in the presence of a negative behavior, we compared incidences of negative-only and mixed (positive and negative) emotions between the presence and absence of a negative behavior. Reported negative emotions were significantly different (Incidence Rate Ratio = 1.76, 95% CI [1.25, 2.45], p< 0.0001) between instances where a negative behavior was present and was not present (61% and 32% respectively). 

These results suggest that children can successfully report having negative emotions in response to having experienced a negative event; however, their specificity is muddied by an overreporting of positive emotions. This may be due to the lack of temporal proximity between the event and the time of the survey. It is possible that children sometimes reported an event that had occurred hours before and thus had time to emotionally stabilize since. A simple solution is to change the prompt to “In the past hour…”. This would ensure a temporal link between the emotion and behavior, improving ecological validity. 

Finding new methods of assessing young children’s experiences as they vary across contexts and time is valuable for assessment and may provide the information necessary to understand the full picture of informant discrepancy. Checks such as this one are necessary to validate EMA for young children before it is implemented in clinical settings.