Participants randomized to the SimCoach intervention arm interacted with "Bill Ford," a simulated human (avatar) enacted in the SimCoach program. This white male avatar representing himself as an Army veteran who spoke to participants in a conversational manner. Participants interacted with him using a chat interface, where they could type responses to him. Bill Ford asked participants a series of questions that corresponded to validated PTSD or depression screening questionnaires. SimCoach then provided personalized recommendations for a symptom if the user reported experiencing the symptom at one of the two highest frequencies on the response scale. These recommendations consisted of a behavioral recommendation with an accompanying link to a website or online article on the topic.

Participants assigned to the Content Matched Control condition arm completed text-based versions of the PCL PTSD screening inventory and PHQ-9 depression screening inventory. These instruments used standard, validated language and did not use the modified conversational versions used in the SimCoach intervention condition. The same personalized recommendations provided to the SimCoach group were provided as conventional text and links. After receiving personalized recommendations, participants completed text versions of primary help-seeking, perceived barriers to seeking help, and secondary outcome (user experience) measures.

Participants randomized to the No-Treatment Control arm completed measures of the primary outcome (intentions to seek help) prior to being given a choice of SimCoach or the conventional screening administered in the other two arms of the study. After the help-seeking intention questionnaire was administered, participants were told that they would be asked some questions about any PTSD or depression symptoms that they might be having and were given the choice between chatting online with a virtual human or using an online form (options were presented in random order to prevent any influence of ordering on selection of the tool). After either interacting with SimCoach or filling out an online form, participants completed a questionnaire assessing user experience.

Participants were asked the following question: "How likely is it that you will seek help for issues such as stress, emotional, alcohol, drug, or family problems from the following people in the next month?" and responded using a Likert scale ranging from 1=extremely unlikely to 7=extremely likely.

Subjects were asked about their perceptions of both stigma-related and practical barriers to seeking help.

Participants responded to questions about their experience with the tool or the online form.