All raters provided informed consent and were given standardized instructions on how to complete the CAARS 2 and, where applicable, the other instruments used for the validity studies (see chapter 9, Validity for study details). All raters were compensated for taking part in the data collection process (see Standardization Phase in chapter 6, Development).

Most respondents completed the CAARS 2 remotely (via an email link). Those with limited access to an internet-enabled computer, or those needing more assistance (e.g., difficulty navigating a computer, accessing the link), completed the CAARS 2 locally at an administrator’s office.

Individuals were eligible for inclusion in Self-Report data collection if they lived in the U.S. or Canada, were at least 18 years old, and indicated that they were able to read English well or very well. Observers had to meet these same criteria, and in addition, had to indicate that they knew the person being rated for at least one month and that they knew them very well. Although there were no restrictions placed on the nature of the relationship between the observers and the person being rated (see Demographic Characteristics of the Observers in appendix J), observers were required to have interacted with the person being rated daily, weekly, or monthly in the past three months (either in person, or via telephone, text message, or video chat), and they had to be able to answer questions about the rated person’s gender, age, race/ethnicity, state or province of residence, and education level (EL).

All data were cleaned between data collection and analysis. Cases were removed from analyses when there were concerns about validity of the data collected due to (a) insufficient adherence to instructions (e.g., rated themselves on the Observer version); (b) incorrect responses to at least two out of the three items where instructions were provided on how to respond (e.g., responded with a selection other than “Never” despite being instructed to “Please respond Never for this item”; (c) omissions exceeding 3% of the total items in the survey (note that 3% was selected as a threshold for omitted responses as nearly complete information from all participants was desired in order to properly evaluate performance of items and scales); (d) excessive identical consecutive responses; (e) evidence that the assessment was not completed in a single sitting (i.e., time to completion in excess of 24 hours); or (f) an excessively fast time to completion (determined via analysis of the distribution of response time and outliers). Applying these data-cleaning criteria resulted in the removal of 12.9% of cases for Self-Report and 13.4% of cases for Observer.