Each of Army STARRS’ five main components was designed to capture a snapshot of soldiers’ experiences in different jobs and at different phases of Army service. For example, data from the New Soldier Study (NSS) help investigators learn more about the men and women who are just entering the Army. Data from the All Army Study (AAS) component help the research team learn about established active duty soldiers. The term “active duty” applies to soldiers who are working full time in active military service, including activated Reserve Component and Army National Guard. By the time data collection ended in March 2013, the Army STARRS research team had surveyed more than 40,000 established active duty soldiers.
To access volunteers representing a wide range of Army experiences, the Army STARRS investigators randomly selected Army units to participate in the AAS. These units were at all stages of active duty service (routine in-garrison, pre-deployment training and preparation, deployed operations, post-deployment reset, etc.) and were stationed in the United States and around the world. More than 10,000 AAS participants were soldiers deployed to Afghanistan who were surveyed as they traveled through Kuwait on military leave.
Soldiers who participated in the study were asked to complete a survey that included questions about their health and experiences. It covered events across all phases of Army service including training, combat, and non-combat experiences. The soldiers were asked for permission to link their questionnaire responses to their administrative data (medical records, legal records, personnel records, etc.), which were supplied by the Army.
A very important part of Army STARRS is privacy. In the AAS, like all of the study components, soldiers’ personal information remains strictly confidential and individual soldiers will not be identifiable (see Confidentiality is Key, below).
Every research study faces challenges and opportunities to develop new methods. This is also true with Army STARRS. Listed below are some of the more interesting aspects of launching AAS.
A Representative Sample
In order to capture an accurate picture of Army life, researchers needed data from a sample of soldiers who represented the Army as a whole. To accomplish this, the Army STARRS identified a number of Army units that, through the random selection process, represented the Army at all stages of the deployment cycle (pre-deployment, deployment, and post-deployment) as well as units that did not deploy. The project team worked closely with the various U.S. Army Commands to create schedules that ensured soldiers in the selected units were able to hear a briefing on Army STARRS and have an opportunity to participate in the study. Developing this representative sample of the Army required a great deal of effort by both the research team and the Army.
Researchers needed a snap shot of all phases of Army service, which meant surveying soldiers around the world. This required coordination between the research team and Army Commands, Army Service Component Commands and Direct Reporting Units including AMC, ARCENT, CIDC, EUSA, FORSCOM, IMCOM, INSCOM, MEDCOM, ODUSA, TRADOC, USAREUR, USARPAC, USARSO, and USASOC 1ST Army NETCOM/9TH Signal.
Paper-and-Pencil or Laptop?
Adapting to the constraints of schedules, global locations, and resources, the Army STARRS team developed two types of AAS questionnaire: a paper-and-pencil version (also known as “survey-in-a-box”) and a computerized version. The research team chose the version best suited to each soldier’s environment and coordinated with the Army to get the materials to the Army STARRS participants.
Confidentiality is Key
Confidentiality is vital to every part of Army STARRS and this includes AAS. No individual or personally-identifying information (PII) will ever be reported in the findings of the Army STARRS research project. To ensure this, the research team removed all PII that could identify a soldier from responses and other study materials. The Army removes all PII from administrative records before sending them to Army STARRS.
In addition, each soldier’s questionnaire responses and other data are combined with information from other soldiers. Researchers are analyzing the combined information and reporting on their findings. Questionnaire answers are not used to evaluate a soldier’s abilities and they are not shared with anyone in the Army. The only exception occurred during the data collection phase, when, if a participant showed imminent danger of self-harm or harm to someone else, the individual was referred to the Army chaplain-on-call for a confidential follow-up consultation.
Researchers asked soldiers for permission to link their questionnaire responses with administrative data from the Army. However, to maintain privacy, soldiers never put their names on their questionnaires—each volunteer was issued a random survey ID. The many types of administrative data provided by the Army don’t contain PII either; the Army replaces the PII with codes. The challenge for the research team was linking the Army’s various records for each individual to the individual’s questionnaire responses. The team solved the problem by developing a computer program that uses the Army STARRS survey IDs and the Army’s codes to link data without adding identifying information.
Because the Army STARRS questionnaire addressed some very personal topics, the Army and the research team considered the needs of any soldier who had a strong emotional response to the survey questions. Researchers and the Army worked to develop a number of safety measures that protected soldiers and their privacy. Resource cards with phone numbers for programs such as Military OneSource, the Suicide Prevention Lifeline, Wounded Soldier and Family Hotline, and others, were distributed to every soldier during Army STARRS briefings. Following each survey session, every participant filled out a card indicating whether or not s/he would like to meet confidentially with a chaplain. Each card was sealed in an envelope and handed to a member of the civilian survey administration team. The civilian team delivered the sealed envelopes to the chaplain-on-call. In the event that a participant taking the computerized questionnaire demonstrated imminent danger of self-harm or harm to someone else, the computer sent a private message to the civilian survey team who contacted the chaplain-on-call. The chaplain reached out to the soldier privately. This is the only situation where information was shared with a member of the military. Communications with a chaplain are confidential.