Page 47 - JSOM Fall 2023

for posttraumatic stress disorder (PTSD) in military personnel completed a survey on a computer tablet. Participants self-
and veterans. 12–14 reported their background information (e.g., age, education)
and responded to the EOD CMS-T items as well as mental
Another step toward psychometric scale validation is to health measures.
demonstrate that the candidate metric can differentiate be-
tween population subgroups. For instance, it is plausible that Mental Health Measures
more experienced individuals from the EOD community may Anxiety symptoms, depressive symptoms, posttraumatic stress
utilize strategies differently than their less experienced coun- symptoms, and perceived stress were evaluated with the Gen-

terparts. Given that some scientists have observed different eralized Anxiety Disorder-7 scale (Cronbach α = 0.88, total
7

strategy “fingerprints” across athlete subgroups (e.g., elite vs. possible score = 21), Patient Health Questionnaire-8(α = 0.85,
non-elite, more vs. less successful), the development of strategy total possible score = 27), PTSD Checklist for DSM-5 (α = 0.95,
fingerprints for various military occupational specialties is a total possible score = 80), and the 10-item Perceived Stress
sensible yet underutilized conceptual approach. 4,5 Scale (α = 0.86, total possible score = 40), respectively. 15–18
This report describes the development and validation of the U.S. Data Analysis
Navy EOD CMS-T, a population-specific measure of strategy Candidate CMS-T items were screened using established crite-
use in EOD training environments. We expected that several the- ria. Specifically, each candidate item was required to demon-
oretically relevant, internally stable factors would emerge, and strate an absence of floor or ceiling effects, an acceptable range
we hypothesized that strategy use would relate to mental health. (≥3), and adequate variability (standard deviation (SD) ≥0.6).
We also expected that there would be differences in the way that Then, candidate items were loaded into an initial principal axis
the strategies of automaticity, “psyching up/down,” and internal factor model with orthogonal (Varimax) rotation and Kaiser
19
dialogue would be used, as compared to the context of sport. normalization. Criteria for initial factor identification in-
Lastly, we explored if the CMS-T could differentiate between cluded eigenvalue >1.0 and substantial contribution (≥5.0%)
the following: EOD accessions, students, and technicians; EOD to the total latent construct (i.e., combat mindset). Next, initial
enlisted personnel and officers; and the qualification levels of criteria for item retention included decisive loading on a single
basic, senior, and master EOD warfare specialist. factor (>0.5), with low cross loading (≤0.4). Within factors,
each item was further expected to demonstrate an essential
contribution, as defined by the following: a substantial cor-
Methods
rected item-total correlation (≥0.5) and an observable adverse
EOD CMS-T Scale Development impact on Cronbach alpha upon item removal (≥10% reduc-
Scale items were developed by a working group composed of tion). Each factor was considered internally reliable if Cron-
eight SMEs from EOD Training and Evaluation Unit 1 (San bach alpha met or exceeded 0.70. Because Cronbach alpha
20
Diego, California), six Naval Health Research Center (NHRC) underestimates the reliability of 2-item subscales, we used
scientists, and one expert in scale development and psycho- Spearman-Brown coefficients for the 2-item subscale. After
21
metrics (SporeData Inc.). All SMEs were active duty Service- item removal, factor analytic models were iteratively repeated
members with extensive training, operational, and deployment until a final, optimized solution was derived. Once the opti-
experience, who were currently serving in training, evaluation, mized subscales were established, they were labeled based on
and/or leadership roles. Facilitated by the first author—who item commonality and theoretical precedence. Subscale scores
has doctoral-level training in sport and performance psychol- were calculated as means (i.e., sum score of subscale items di-
ogy—the working group developed 30 candidate items with vided by the total number of subscale items).
the goal of fusing theoretical and practical relevance. Building
upon the sport/performance framework, each item was ini- Lastly, convergent validity was established two ways. First,
tially drafted by the scientists. Next, the working group iter- bivariate associations with mental health symptoms were
atively developed each item until the EOD SMEs appraised it evaluated. Second, we explored differences between EOD
as having face-validity. All items are based on a 5-point Likert- accessions, advanced students, and basic, senior, and master
type frequency scale from 1 (never) to 5 (always). Examples of EOD technicians, and between EOD enlisted personnel and
items include “I am able to control distracting thoughts during officers. Selected group differences were evaluated statistically
training evolutions,’’ and ‘‘I practice a way to relax prior to the with ANOVA. For groups of three (e.g., basic, senior, master
start of training evolutions.” EOD technician), statistical modeling was limited to presumed
“maximally different groups” (e.g., basic vs. master, EOD ac-
Participants cession vs. EOD technician). For each comparison, theoreti-
The NHRC Institutional Review Board approved this study cally relevant variables (e.g., education, military experience)
protocol (NHRC.2018.0019). Participants (N = 164) were were first evaluated as candidate covariates following stan-
recruited from units within EOD Group 1 (San Diego, Cali- dardized selection criteria, and were then included in the cor-

fornia) and from the Center for Explosive Ordnance Disposal responding analysis of variance if criteria were met. 22
and Diving (CEODD) at Great Lakes, Illinois. This sample
comprised EOD accessions (new recruits) at CEODD (n = 77), Results
EOD advanced students in the U.S. Navy Tactical Training
Course (n = 24), and EOD technicians (n = 63) with warfare Detailed participant characteristics are provided in Table 1.
designations of basic, senior, or master EOD technician.
The rotated factor matrix is displayed in Table 2, and sub-
Procedure scale characteristics are shown in Table 3. As shown, five
This was a cross-sectional study, and participants provided in- internally reliable subscales were defined and named: relax-
formed consent. During a single in-person session, participants ation (5 items), attentional-emotional control (5 items), goal

Psychological Strategies in EOD Technicians | 45