Ecological Momentary Assessment: A New Method for Studying Affect In Adolescents with Autism

Background: Individuals with autism demonstrate difficulties in understanding and reporting on their own emotions (e.g., Losh & Capps, 2006; Rieffe, et al., 2007).  Most prior research has used laboratory measures, thus it is unclear how well individuals with autism are able to report on their emotions in more ecologically valid settings.  Ecological momentary assessment (EMA) is a non-invasive method of gathering real-time data that incorporates multiple daily assessments and allows for the detection of incremental variability in subject responses.  It is highly effective at capturing the affective experiences of typically developing children and adolescents (e.g., Larson, 1989; Silk, Steinberg, & Sheffield, 2003), but has not been utilized with individuals with autism. 

Objectives: To determine if EMA is an effective methodology for use with individuals with autism and, if so, to provide further insight into how adolescents with autism report on their own affective experiences. 

Methods:  Nineteen high-functioning adolescents with autism (M = 14.7 years) and 19 age- and IQ-matched controls (M = 14.1 years) were given answer-only cell phones and were called 40 times over 14 consecutive days and asked to rate their current levels of positive and negative affect using the PANAS-C.  Corresponding parent reports were collected for up to 10 of the 14 days.  In addition, prior to the EMA data collection phase, participants and their parents completed the Child Depression Inventory (CDI) as well as the Screen for Child Anxiety Related Emotional Disorders (SCARED).

Results: Adolescents with autism were just as likely to complete the protocol, averaged the same number of completed calls, and were willing to remain on the telephone for the same mean length of time as controls.  In addition, their use of the rating scale was not significantly different from controls, they endorsed experiencing some form of emotion just as frequently as controls did, and their general pattern of reporting for specific emotions was the same.  Overall, the adolescents with autism, in contrast to controls, reported higher intensity negative affect and more lability in positive and negative affect.  In comparison to parent report, there was some suggestion that the adolescents with autism, but not controls, were underreporting the intensity of their negative affect.  For both groups, lability of negative affect was related to self-reported depression symptoms, but not to parent reports of child depression or self- or parent-reported anxiety symptoms.

Conclusions: This is the first study to illustrate that EMA methodology can be effectively implemented with adolescents with autism.  The data indicates that, at least by adolescence, individuals with autism can perceive and report on varying levels of affect.  Although the accuracy of the intensity of affect they report is still unclear, data suggests they can introspect sufficiently to detect a change in their base level of affect.  Additionally, their ability to recognize fluctuations in negative affect on a day to day basis is related to their ability to report on mood symptoms.