Poster #13 - Evaluating the Ugandan Parent Rated Observation of Communication, Emotion, and Social Skills with Rasch Analysis

• In Event: Poster Session 19

Abstract

Background: Numerous instruments for identifying early behavioral markers of autism spectrum disorder exist in Western societies, but none have been developed specifically for use in African countries such as Uganda. The objective of this study was to adapt a pre-existing tool, the Parent-Rated Observation of Communication, Emotion, and Social Skills (PROCESS), and determine its psychometric integrity and scale properties for use in Uganda.
Methods: Following translation and cultural adaptation of the PROCESS, the Ugandan version of the PROCESS (or U-PROCESS) was deployed to collect data from a community sample of two hundred and forty-six (n=246) parents of children aged 12-36 months, attending immunization clinics at two health facilities in Kampala, Uganda. To investigate the measurement properties of the U-PROCESS in this context, Rasch analysis was conducted including response category analysis, fits statistics, reliability, and tests of unidimensionality and local independence of items. Corresponding analyses on the psychometric properties of the original version of the PROCESS were conducted in a convenience sample of n = 1306 children drawn from an urban midwestern city in the US.
Results: Overall, the PROCESS was found to be feasible to administer in both US and Uganda. While the PROCESS seems to function well in the US sample, several issues emerged that implicate the utility of exporting the instrument to the Ugandan context. The response category ordering of the U-PROCESS did not perform as intended in this sample, though a vast majority of the items (22 out of 25) fitted the Rasch model. Item separation and reliability indices were found to be adequate, but this was not the case for person separation and reliability indices. The criterion of local independence was met, but the U-PROCESS did not meet the assumption of unidimensionality.
Conclusion: Evidence that the majority of items on the U-PROCESS function adequately and provide reliable data is a critical first step in establishing the validity of this instrument in an African sample. Our findings suggest that the U-PROCESS should be further refined to better enhance its measurement characteristics and therefore its clinical utility in this setting.

Keywords: Autism, early identification, PROCESS, Uganda

Authors

• Lana Hantzsch, University of Minnesota - Twin Cities
  Presenting Author

• Emmanuel Bonney, University of Minnesota - Twin Cities

• Sooyeon Sung, University of Minnesota - Twin Cities

• Isabel M Smith, Dalhousie University

• Lonnie Zwaigenbaum, University of Alberta

• Catherine Abbo

• Jed T. Elison, University of Minnesota - Twin Cities