Uploaded January 2025 | Updated September 2026, 1 week ago
Despite having acquired valuable skills such as teamwork, leadership, work ethic and discipline during their service, military veterans face significant challenges when transitioning to the civilian workforce, which are quite evident during the employment interview, the most widely used selection tool that all job applicants have to pass.
To address these challenges, the goal of the VetTrain project, sponsored by the U.S. National Science Foundation, was two-fold.
First, it sought to gain a deeper understanding of both interviewers' and veteran interviewees' perspectives about the latter’s interviewing performance.
Second, it designed new assistive, personalized AI technologies to help military veterans succeed in civilian job interviews.
The first part of VetTrain examined the perceptions of interviewers and veteran interviewees regarding the strengths and weaknesses of veterans in civilian job interviews.
Three strengths emerged from the study: communicating soft skills, professionalism, and confidence.
On the other hand, ineffective translation of relevant technical skills, excessive use of military jargon and nervousness proved to be weaknesses widely observed.
Whereas the employment interview is likely to be stressful for almost everyone, veteran interviewees who display signs of nervousness or anxiety may be unjustifiably rated more negatively during the interview due to stereotypical association with psychological issues.
With this in mind, the second part of VetTrain conducted mock interviews between military veterans and interviewers from industry sectors that typically hire veterans, such as transportation, construction, and engineering.
This allowed us to elicit naturalistic stress responses experienced by military veterans during the interviews and then examine self-reported and bio-behavioral measures of stress, such as heart rate and sweat activity.
This also resulted in a better understanding of unique verbal communication gaps, such as ineffective translation of relevant military experience and technical skills, over-explaining of responses, and excessive use of military jargon, that disadvantage veterans from succeeding in the civilian job interview.
Findings indicated large inter-individual variability in stress responses and verbal communication patterns among participants, highlighting the need for personalized interview training strategies.
To address these challenges, the VetTrain project developed artificial intelligence technologies that provide personalized feedback during interview preparation—for example mock interviews—, helping veterans understand and manage their stress responses.
In addition, we designed new natural language processing methods that can pinpoint exact turns in the dialog that effectively serve or hurt the job interview outcome.
We also addressed the design of user interfaces for AI-based job interview training systems, and investigated how people are influenced by given descriptions of AI models in a system.
And finally, the project examined user’s perceptions of and trust in the designed AI technologies.
Findings from this research finally suggest that employers should use structured instead of unstructured interviews.
The training of interviewers that is required for structured interviews will also provide employers the opportunity to familiarize their interviewers with the unique potential challenges with interviewing veterans, thus giving them the chance to effectively present their skills.
Integrating and using these strategies and technologies benefits both employers and veterans; employers by placing them in a position to hire individuals with desirable and high skill sets, and veterans by facilitating their smooth and successful transition to the civilian workforce.
Despite having acquired valuable skills such as teamwork, leadership, work ethic and discipline during their service, military veterans face significant challenges when transitioning to the civilian workforce, which are quite evident during the employment interview, the most widely used selection tool that all job applicants have to pass.
To address these challenges, the goal of the VetTrain project, sponsored by the U.S. National Science Foundation, was two-fold.
First, it sought to gain a deeper understanding of both interviewers' and veteran interviewees' perspectives about the latter’s interviewing performance.
Second, it designed new assistive, personalized AI technologies to help military veterans succeed in civilian job interviews.
The first part of VetTrain examined the perceptions of interviewers and veteran interviewees regarding the strengths and weaknesses of veterans in civilian job interviews.
Three strengths emerged from the study: communicating soft skills, professionalism, and confidence.
On the other hand, ineffective translation of relevant technical skills, excessive use of military jargon and nervousness proved to be weaknesses widely observed.
Whereas the employment interview is likely to be stressful for almost everyone, veteran interviewees who display signs of nervousness or anxiety may be unjustifiably rated more negatively during the interview due to stereotypical association with psychological issues.
With this in mind, the second part of VetTrain conducted mock interviews between military veterans and interviewers from industry sectors that typically hire veterans, such as transportation, construction, and engineering.
This allowed us to elicit naturalistic stress responses experienced by military veterans during the interviews and then examine self-reported and bio-behavioral measures of stress, such as heart rate and sweat activity.
This also resulted in a better understanding of unique verbal communication gaps, such as ineffective translation of relevant military experience and technical skills, over-explaining of responses, and excessive use of military jargon, that disadvantage veterans from succeeding in the civilian job interview.
Findings indicated large inter-individual variability in stress responses and verbal communication patterns among participants, highlighting the need for personalized interview training strategies.
To address these challenges, the VetTrain project developed artificial intelligence technologies that provide personalized feedback during interview preparation—for example mock interviews—, helping veterans understand and manage their stress responses.
In addition, we designed new natural language processing methods that can pinpoint exact turns in the dialog that effectively serve or hurt the job interview outcome.
We also addressed the design of user interfaces for AI-based job interview training systems, and investigated how people are influenced by given descriptions of AI models in a system.
And finally, the project examined user’s perceptions of and trust in the designed AI technologies.
Findings from this research finally suggest that employers should use structured instead of unstructured interviews.
The training of interviewers that is required for structured interviews will also provide employers the opportunity to familiarize their interviewers with the unique potential challenges with interviewing veterans, thus giving them the chance to effectively present their skills.
Integrating and using these strategies and technologies benefits both employers and veterans; employers by placing them in a position to hire individuals with desirable and high skill sets, and veterans by facilitating their smooth and successful transition to the civilian workforce.
