The Feedback Instrument for Rescue Force Development \u2013 Group work (FIRE-G)<\/strong> is a standardized, psychometrically validated evaluation instrument designed to measure the perceived quality of collaborative learning and peer interactions during group work exercises in emergency service training. Developed as a specialized modular extension to the broader Feedback Instrument for Rescue Force Development<\/em> (FIRE; Schulte & Thielsch, 2019; Schulte et al., 2019), the FIRE-G addresses the specific pedagogical dynamics of cooperative problem-solving among rescue forces, including firefighters, incident commanders, disaster response personnel, and emergency medical service leaders. The scale operationalizes the unidimensional construct of learning-conducive peer interactions<\/em> (cooperative learning<\/a> dynamics) using an ultra-brief three-item format administered via paper-and-pencil or computer-assisted systems. The instrument was developed and validated across multiple empirical samples totaling over 600 emergency personnel at the Institute of the North Rhine-Westphalia Fire Brigade (Institut der Feuerwehr Nordrhein-Westfalen<\/em>; IdF NRW). In the primary validation cohort (N<\/em> = 375), confirmatory factor analysis (CFA<\/a>) demonstrated excellent structural fit for an essentially tau-equivalent unidimensional model (χ²(2) = 2.95, p<\/em> = .229, CFI = .998, TLI = .997, RMSEA = .04, SRMR = .04). Internal consistency reliability proved robust across cohorts, yielding Cronbach’s α = .83 and McDonald’s ω = .83 in the primary validation study, and McDonald’s ω = .77 in an initial exploratory cohort (N<\/em> = 240). Convergent and discriminant construct validity analyses confirmed substantial alignment with general peer group appraisal (r<\/em> = .53 with the FIRE Group<\/em> scale) and functional divergence from practical tactical exercises (FIRE-E), instructor demeanor, cognitive entry prerequisites, and momentary emotional bias. With an average administration time of approximately one minute, the FIRE-G provides educational practitioners, occupational psychologists, and emergency training organizations with an efficient, reliable, and construct-clean metric for didactic course optimization and pedagogical research<\/span>.<\/p>\n FIRE-G, rescue forces, firefighter training, group work evaluation, cooperative learning, didactic evaluation, emergency services, psychometrics, peer interaction, instructional quality, scale validation<\/p>\n The Feedback Instrument for Rescue Force Development – Group work (FIRE-G) was developed through an institutional research partnership between the Department of Organizational and Business Psychology<\/strong> at the University of Münster<\/strong> (Westfälische Wilhelms-Universität Münster<\/a>) and the Institute of the North Rhine-Westphalia Fire Brigade<\/strong> (Institut der Feuerwehr Nordrhein-Westfalen<\/em>, IdF NRW) in Münster, Germany.<\/p>\n The vocational demands placed upon rescue forces—such as professional and volunteer firefighters, technical relief specialists, and incident command officers—require a sophisticated integration of technical expertise, physical sciences, tactical doctrine, emergency medicine, and statutory legal frameworks. When deployed during critical incidents, commanding officers operate in dynamic, volatile, high-stakes environments where they must make rapid, consequential tactical decisions under severe time constraints. In such high-tempo crisis operational scenarios, commanders cannot readily pause to consult field manuals, legal compendiums, or external subject-matter experts. Consequently, theoretical classroom education and pedagogical preparation within modern fire academies and crisis training centers represent a vital, life-critical component of emergency force readiness.<\/p>\n Within adult vocational instruction and leadership development, group work stands as a core instructional method (teaching method<\/a>) alongside plenary lecture instruction, independent study, and paired exercises (Drumm & Scholz, 2007). In emergency services leadership courses, collaborative group work exercises are routinely implemented to train officers in analyzing operational case studies, dissecting fire safety legislation, interpreting complex building codes, and formulating multi-agency response strategies. Despite the widespread use of small-group assignments, evaluating the efficacy and didactic quality of group work has historically suffered from methodological conflation. In traditional end-of-course student evaluations of teaching (SET), items assessing small-group activities are frequently intermixed with instructor facilitation, administrative clarity, or general physical logistics.<\/p>\n The primary purpose of the FIRE-G<\/strong> is to isolate and measure the educational quality of small-group collaborative processes from the perspective of course participants. Rather than evaluating whether the lecturer was engaging or whether the classroom supplies were adequate, the FIRE-G assesses the degree to which peer-to-peer interactions generated a collaborative learning climate characterized by reciprocal knowledge construction, competent and equitable contribution, and active dialogue. By decoupling instructional group interactions from direct lecturer behaviors, the scale fills an essential gap in instructional feedback systems.<\/p>\n In applied training environments, the FIRE-G serves multiple operational functions:<\/p>\n The psychological construct captured by the FIRE-G is learning-conducive peer interaction<\/strong> (lernförderliche Interaktionen<\/em>) within small-group instructional settings. Rooted in social-constructivist and cooperative learning theories, this construct conceptualizes collaborative learning not merely as passive co-presence or logistical group division, but as active sociocognitive engagement among adult learners.<\/p>\n Learning-conducive interactions encompass three core interrelated operational facets:<\/p>\n Adult learners in emergency leadership courses bring diverse experiential backgrounds from municipal career departments, volunteer fire stations, industrial plants, and emergency medical agencies. When an instructional task functions effectively, participants engage in cognitive elaboration: they articulate tacit knowledge, explain tactical reasoning to peers, justify operational decisions, and encounter alternative problem-solving strategies (Renkl, 2015). This active sociocognitive processing prompts cognitive conflict, exposes individual knowledge gaps, and leads to refined conceptual schemas. The FIRE-G directly measures this subjective gain through item reflections on how much the participant acquired knowledge directly from fellow trainees (Item 1).<\/p>\n A persistent vulnerability in collaborative group work is the emergence of dysfunctional group dynamics, specifically social loafing<\/a> or free-riding (Karau & Williams, 1993), wherein individual group members withhold effort when personal accountability is diffuse. A related phenomenon is the sucker effect<\/em> (Gimpel-Effekt<\/em>; Schnake, 1991), wherein highly motivated members reduce their own contributions to avoid being exploited by unengaged peers. In an optimal collaborative group setting, participants participate proportionally in alignment with their specialized individual competencies, skills, and background knowledge. The FIRE-G operationalizes this cooperative norm through Item 2, assessing the degree to which every member actively contributed according to their individual competence.<\/p>\n Effective small-group work requires an interactive architecture that stimulates multidirectional discourse rather than parallel, isolated individual work. Group assignments must prompt dialogue, debate, collaborative negotiation, and shared consensus building regarding operational tactics and statutory mandates. The third facet of the construct captures the extent to which the group task catalyzed vibrant interpersonal exchange and discourse among the trainees (Item 3).<\/p>\n Importantly, during the psychometric refinement of the FIRE-G, the authors deliberately excluded items reflecting external administrative or instructor variables (such as clarity of task instructions, adequacy of instructor coaching during group exercises, and plenary debriefing). By removing these confounding logistical aspects, the construct was purified into an unconfounded assessment of internal group communicative synergy and sociocognitive learning exchange.<\/p>\n The theoretical architecture of the FIRE-G rests upon established principles of educational psychology, social psychology of groups, and adult vocational learning paradigms.<\/p>\n According to social constructivist<\/a> educational theory (e.g., Vygotsky, 1978; Renkl, 2015), high-level learning occurs when individuals actively construct cognitive schemas through mediated interaction with their social environment. Within classroom settings, group work serves as one of four fundamental social arrangements, alongside individual work, dyadic partner work, and frontal plenary lecturing (Drumm & Scholz, 2007). In small-group settings, adult learners do not merely absorb information passively; they are compelled to explicate their line of thought, explain abstract operational rationales to peers, debate ambiguous edge cases, and justify tactical maneuvers. This process of verbal explication forces the learner to organize internal mental models systematically. As highlighted by Renkl (2015), when peers present conflicting perspectives or identify flaws in an argument, the learner experiences cognitive dissonance, which drives deeper information processing and long-term retention.<\/p>\n In high-reliability training environments such as the fire service, group exercises follow structured pedagogical trajectories. Pluntke (2013) conceptualizes instructional group work across three distinct, consecutive phases:<\/p>\n Because the second phase represents the autonomous arena where trainees interact directly with one another, the pedagogical success of the exercise hinges entirely on group-level dynamics during this window.<\/p>\n The theoretical rationale for developing the FIRE-G also incorporates classic social-psychological models of collective effort. As demonstrated by Karau and Williams (1993) in their Collective Effort Model (CEM), individuals routinely exhibit social loafing when individual outputs are pooled and personal accountability diminishes. In training cohorts with wide age and experience discrepancies, social loafing can quickly trigger the sucker effect (Schnake, 1991), degrading the exercise into cynical disengagement. Pedagogical researchers emphasize that instructors can mitigate these dysfunctional phenomena by designing challenging, interdependent group tasks, keeping group sizes modest, and ensuring that assignments require the explicit mobilization of differentiated member competencies (Liden et al., 2004). The FIRE-G captures whether the resulting group execution successfully achieved this cooperative equilibrium.<\/p>\n The psychometric validity of the FIRE-G has been systematically demonstrated through empirical investigations assessing content validity, internal structural validity, and external construct validity (convergent and discriminant evidence).<\/p>\n Content validity was evaluated in Study I<\/strong> (N<\/em> = 33), comprising seven senior fire academy instructors (86% male, mean age = 38.9 years, SD<\/em> = 6.0) and 26 officer trainees enrolled in a certified fire incident commander course (Gruppenführer<\/em>, 96% male, mean age = 30.5 years, SD<\/em> = 6.9) at the Institute of the North Rhine-Westphalia Fire Brigade (IdF NRW). All participants conducted a rigorous review of candidate items evaluating instructional relevance and comprehensibility. With the exception of one initial item (G_5: “Trainees were interested in a good group work outcome”), which was rated important by only 76% of respondents and was subsequently revised into a first-person format, all proposed group work items achieved approval ratings exceeding 80% for instructional relevance. Furthermore, no item was deemed incomprehensible by more than 3% of the expert panel. Expert evaluations affirmed that the scale captures the essential facets of peer-driven collaborative learning<\/span>.<\/p>\n Factorial validity was established across two independent cohorts. In Study II<\/strong> (N<\/em> = 240), an exploratory factor analysis (EFA) demonstrated strong single-factor saturation for the interaction-focused items, with factor loadings ranging from .81 to .84. In Study III<\/strong> (N<\/em> = 375), confirmatory factor analysis (CFA) formally tested the single-factor structural representation. Using maximum likelihood estimation with robust standard errors (MLR), an essentially tau-equivalent unidimensional model yielded an excellent fit to the empirical data (χ²(2) = 2.95, p<\/em> = .229, CFI = .998, TLI = .997, RMSEA = .04, 90% CI [.00, .12], SRMR = .04). A chi-square difference test confirmed that constraining factor loadings to equality (tau-equivalence) did not significantly degrade model fit compared to an unconstrained congeneric model (Δχ²(2) = 2.95, df<\/em> = 2, p<\/em> = .229), empirically validating the interchangeable weighting of items in composite scoring<\/span>.<\/p>\n Convergent construct validity was examined in Study III by correlating the FIRE-G composite with validated subscales from the comprehensive FIRE core questionnaire (Schulte & Thielsch, 2019). The FIRE-G exhibited a strong, statistically significant positive correlation with the FIRE Group<\/em> scale (r<\/em> = .53, p<\/em> < .001, n<\/em> = 373). Because the FIRE Group<\/em> subscale assesses general social cohesion and mutual support across the entire course cohort, this substantial correlation verifies that the FIRE-G captures relevant peer interaction dynamics while focusing specifically on collaborative task performance<\/span>.<\/p>\n Discriminant validity was established through several planned comparisons:<\/p>\n The reliability of the FIRE-G was thoroughly evaluated across independent samples using classical test theory and structural equation modeling frameworks.<\/p>\n Because the FIRE-G contains three items, evaluating its internal consistency requires precise alignment with the underlying measurement model:<\/p>\n An internal consistency coefficient of .83 for a three-item metric is exceptionally high, demonstrating that the items tap into a unified, coherent psychological dimension without introducing excessive item redundancy.<\/p>\n The factorial validity of the FIRE-G was examined across two sequential empirical phases: initial item reduction via exploratory factor analysis (EFA) and subsequent model confirmation via confirmatory factor analysis (CFA).<\/p>\n The original item battery comprised seven candidate items: four assessing peer learning interactions (G_1, G_3, G_5, G_6) and three assessing organizational\/instructor facilitation (G_2: instructor support; G_4: debriefing quality; G_7: clarity of task instructions). Candidate item G_5 (“Participants cared about a good result”) exhibited marked ceiling effects and restricted variance (M<\/em> = 6.14, SD<\/em> = 0.82, kurtosis = 4.67) and was excluded prior to exploratory factor extraction<\/span>.<\/p>\n An exploratory principal component analysis with oblimin rotation was conducted on the remaining six items (EFA 1; N<\/em> = 240) using the R package<\/span> In Study III (N<\/em> = 375), the structural integrity of the final three-item configuration was formally tested via CFA using the R package<\/span> The resulting CFA demonstrated outstanding fit to the empirical data:<\/p>\n Following the standard fit criteria established by Hu and Bentler (1999), these results confirm an exemplary structural representation of the construct.<\/p>\n Descriptive statistics and psychometric properties for the manifest items in Study III (N<\/em> = 375) are presented below<\/span>:<\/p>\n All items demonstrate strong discriminatory power (all rit<\/sub><\/em> ≥ .67) and acceptable distribution parameters typical of positive training evaluations<\/span>.<\/p>\nKeywords<\/h2>\n
Authors<\/h2>\n
\n
Purpose<\/h2>\n
\n
Psychological Construct<\/h2>\n
1. Reciprocal Peer Learning and Knowledge Construction<\/h3>\n
2. Equitable Competence-Based Participation and Social loafing Mitigation<\/h3>\n
3. Facilitated Multidirectional Discourse and Peer Exchange<\/h3>\n
Theoretical Framework<\/h2>\n
Sociocognitive Interaction and Knowledge Elaboration<\/h3>\n
Phases of Group Work in Emergency Vocational Training<\/h3>\n
\n
Mitigating Dysfunctional Group Phenomena<\/h3>\n
Validity<\/h2>\n
Content Validity<\/h3>\n
Structural Construct Validity<\/h3>\n
Convergent Validity<\/h3>\n
Discriminant Validity<\/h3>\n
\n
Reliability<\/h2>\n
Internal Consistency Across Validation Cohorts<\/h3>\n
\n
Factor Analysis<\/h2>\n
Item Development and Exploratory Factor Analysis (Study II)<\/h3>\n
psych<\/code> (Revelle, 2021). All six items displayed substantial factor loadings ranging from .66 to .80 on a single general dimension, with item communalities (h<\/em>²) ranging between .41 and .61. However, to preserve test economy and construct cleanliness, the authors removed items G_2, G_4, and G_7 because they indexed instructor-dependent organizational aspects already captured by other specialized tools (the FIRE core scale and the FIRE-DO instructor evaluation module; Kleinstück et al., 2021). A subsequent oblimin-rotated EFA on the three peer-focused items (EFA 2) revealed clean, exceptionally high factor loadings:<\/p>\n\n
Confirmatory Factor Analysis (Study III)<\/h3>\n
lavaan<\/code> (Rosseel, 2012) with robust maximum likelihood (MLR) estimation. Because a standard congeneric model with three manifest indicators has exactly zero degrees of freedom (df<\/em> = 0)—representing a just-identified model wherein goodness-of-fit indices cannot be mathematically calculated—the authors applied an essentially tau-equivalent specification. In an essentially tau-equivalent model, unstandardized factor loadings are constrained to equality while error variances remain freely estimated, yielding two degrees of freedom (df<\/em> = 2; Czerwiński & Atroszko, 2021)<\/span>.<\/p>\n\n
Item Characteristics in the Validation Sample<\/h3>\n
\n
Instrument \/ Measurement Tool<\/h2>\n
\n
\n
\n
\n
Item_Mean = ∑(Item_Points) \/ N_respondents<\/code>.<\/li>\nScale_Mean = (Item_Mean_1 + Item_Mean_2 + Item_Mean_3) \/ 3<\/code>.<\/li>\n<\/ol>\n<\/li>\n<\/ul>\n<\/li>\nPermissions & Fee and Test Year<\/h2>\n
\n