COVID-19 Applicability Measure

Abstract

The COVID-19 Applicability Measure (Wirtz et al., 2023) is an empirical psychometric instrument designed to evaluate the self-reported perceived feasibility and real-world implementation success of public health non-pharmaceutical interventions (NPIs) during the COVID-19 pandemic. While conventional epidemiological instruments predominantly assess knowledge, attitudes, or retrospective compliance frequencies, this instrument specifically operationalizes applicability—the perceived ease, practicability, and operational success with which individuals integrate infection prevention behaviors into the complex demands of their daily routines. Developed against the backdrop of national prevention guidelines and qualitative field interviews in Germany, the scale targets vulnerable populations navigating heightened interpersonal, caregiving, and physiological demands, specifically pregnant women and mothers of infants.

The instrument comprises 20 items capturing five correlated latent dimensions: (1) Apply hygiene measures, (2) Avoid contact with other people, (3) Avoid public transportation, (4) Stay at home, and (5) Check infection status. Each item is evaluated along a six-point reversed Likert-type response scale ranging from 1 (“very good” / Sehr gut) to 6 (“very poor” / Sehr schlecht). Psychometric evaluation via exploratory ordinal factor analysis (EFA) with oblique GEOMIN rotation established robust construct validity, yielding excellent global fit indices ($ ext{CFI} = .977$,$ ext{TLI} = .957$,$ ext{SRMR} = .036$) and accounting for 62% of the total variance across five extractable factors satisfying the Kaiser-Guttman criterion (eigenvalues ranging from 6.96 to 1.09). Subscale internal consistencies spanned from modest to satisfactory ($lpha = .562$ to $.777$), reflecting the heterogeneous, context-dependent behavioral indicators characteristic of real-world epidemiological compliance. The instrument provides clinical researchers, health psychologists, and public health officials with a validated diagnostic metric to differentiate motivational resistance from structural barriers in pandemic behavioral adherence.

Keywords

COVID-19 Prevention Measures, Applicability, Implementation Feasibility, Apply Hygiene Measures, Avoid Contact with Other People, Avoid Public Transportation, Stay at Home, Check Infection Status, Maternal Health, Psychometrics

Authors

The scale was developed and psychometrically validated by a specialized research team in health sciences and methodology at the University of Education Freiburg (Pädagogische Hochschule Freiburg), Germany:

• Markus Antonius Wirtz, Ph.D. — Faculty of Mathematics, Natural Sciences and Technology, Research Methods in Health Sciences, University of Education Freiburg, Germany.
  ORCID: 0000-0002-9392-0484.
  Email: [email protected].
  Correspondence Address: Kunzenweg 21, 79117 Freiburg im Breisgau, Germany.
• Andrea Heiberger, M.Sc. — Faculty of Mathematics, Natural Sciences and Technology, Research Methods in Health Sciences, University of Education Freiburg, Germany.
  ORCID: 0000-0002-2446-9626.
• Carolin Dresch, M.Sc. — Faculty of Mathematics, Natural Sciences and Technology, Research Methods in Health Sciences, University of Education Freiburg, Germany.
  ORCID: 0000-0001-5593-2954.
• Anja Alexandra Schulz, Ph.D. — Faculty of Mathematics, Natural Sciences and Technology, Research Methods in Health Sciences, University of Education Freiburg, Germany.

Purpose

During global viral pandemics such as SARS-CoV-2, public health mitigation strategies rely profoundly on widespread adherence to non-pharmaceutical interventions (NPIs), including physical distancing, mask-wearing, frequent hand hygiene, surface disinfection, and mobility restrictions. However, public health discourse often conflates an individual’s willingness or intention to adhere with their actual structural, environmental, and domestic ability to execute these protective mandates. The primary objective of the COVID-19 Applicability Measure is to quantify the ecological feasibility and real-world applicability of officially mandated protective behaviors within the routine living contexts of individuals.

The scale was deliberately designed to unpack the ubiquitous “intention-behavior gap” in health psychology. In early 2021, prior to the widespread availability of vaccines and amid surging transmission rates driven by novel variants, public directives placed unprecedented burdens on domestic life. For vulnerable cohorts—such as pregnant women and mothers caring for infants—adhering to sweeping lockdown policies presented distinct operational dilemmas. For example, maintaining complete social isolation while attending mandatory prenatal obstetric checkups, securing infant provisions, managing home confinement without external childcare support, or avoiding crowded public spaces created acute friction between theoretical compliance and empirical feasibility.

The COVID-19 Applicability Measure was engineered to serve several clinical, epidemiological, and health policy applications:

• Diagnostic Differentiation of Non-Adherence: By measuring subjective applicability rather than raw moral or behavioral compliance, the tool isolates whether non-compliance stems from ideological refusal vs. environmental and socio-economic incapacity (e.g., lack of private transit, inability to telecommute, or caregiving obligations).
• Targeted Health Communication: It enables policymakers and public health agencies to identify specific domains where public guidance fails to accommodate structural realities, thereby informing tailored behavioral nudges and community-level support structures.
• Perinatal and Maternal Vulnerability Assessment: It offers maternal-child healthcare professionals an empirical instrument to assess behavioral stress and domestic burden in expectant mothers, directly informing psychosocial support interventions during epidemic outbreaks.
• Psychometric Modeling of Behavioral Subtypes: When combined with latent class analysis (LCA), the tool allows researchers to segment populations into behavioral compliance phenotypes based on patterns of everyday execution across diverse behavioral domains.

Psychological Construct

The underlying construct assessed by the COVID-19 Applicability Measure is perceived behavioral applicability—defined as the subjectively appraised feasibility, practicability, and execution success of prescribed health-protective routines under everyday environmental, social, and structural constraints. Unlike self-efficacy, which captures an individual’s generalized prospective confidence in their capability to execute courses of action, perceived applicability assesses retrospective or contextualized operationalization: how well a behavior actually meshed with life demands.

The construct is multidimensional, operationalized across five empirically and theoretically grounded sub-domains:

1. Apply Hygiene Measures

This subscale assesses the execution of physical and barrier-based infection prevention actions. It encompasses both habitual personal hygiene routines and context-specific spatial protections. Behaviors include maintaining the recommended 1.5-meter physical distance, ventilating indoor spaces at regular intervals (e.g., every 20 minutes when sharing closed rooms), wearing certified mouth-and-nose face coverings (masks) in general settings, dynamically donning masks during spontaneous brief outdoor conversations with acquaintances, executing thorough routine handwashing, and regularly applying chemical hand disinfectants. High scores on these items indicate that personal and domestic barriers (e.g., skin sensitivity, forgetfulness, environmental unavailability of sanitizer, or awkward social dynamics) hindered consistent hygiene execution.

2. Avoid Contact with Other People

This dimension operationalizes the social mitigation of infectious transmission. Rather than measuring complete hermitism, it measures the practical management of interpersonal encounters. Items measure how successfully individuals managed to minimize the overall number of social contacts, restrict interpersonal meetings exclusively to individuals who were demonstrably compliant with pandemic regulations, and abstain from leveraging informal family networks—specifically avoiding having grandparents or external acquaintances assist with infant care and babysitting. This domain captures the psychological tension between interpersonal attachment needs, childcare requirements, and epidemic transmission risks.

3. Avoid Public Transportation

This specialized behavioral domain evaluates mobility-related infection avoidance, focusing specifically on the avoidance of municipal transit systems (such as buses, subways, and commuter trains). Mass transit hubs represent high-density vector environments where physical distancing is frequently compromised. Successfully avoiding public transport depends heavily on structural resources, such as access to private motor vehicles, bicycle infrastructure, pedestrian proximity to essential amenities, or the complete restriction of travel.

4. Stay at Home

The “Stay at Home” dimension measures domestic confinement and the systematic circumvention of shared community environments. It evaluates how successfully individuals avoided patronizing public and cultural facilities (e.g., restaurants, municipal libraries, cafes), executed occupational responsibilities remotely via telecommuting/home office setups, strategically organized grocery procurement outside of high-density peak shopping hours, and avoided interregional or non-essential geographic travel. Items also capture the technological integration of daily life, assessing how effectively participants substituted in-person activities with digital solutions, such as telecommunication platforms (Zoom, Skype) and online ordering, grocery delivery, or “click-and-collect” services.

5. Check Infection Status

This domain captures active surveillance, informational monitoring, and diagnostic gatekeeping. It addresses the individual’s proactive behaviors in tracking epidemic conditions and biological infection risks. Indicators include actively keeping informed regarding ever-shifting legal mandates and municipal incidence metrics, utilizing digital contact-tracing tools (such as the national Corona-Warn-App), and securing formal or rapid diagnostic antigen/PCR testing prior to visiting vulnerable, immunocompromised, or high-risk relatives and acquaintances. This dimension reflects an analytical, surveillance-oriented approach to personal epidemiological risk management.

Theoretical Framework

The COVID-19 Applicability Measure is conceptualized at the intersection of several prominent paradigms in health psychology, behavioral medicine, and social epidemiology:

Theory of Planned Behavior and Perceived Behavioral Control

Within Icek Ajzen’s Theory of Planned Behavior (TPB), behavioral performance is governed jointly by behavioral intentions and perceived behavioral control (PBC). PBC reflects an individual’s appraisal of the subjective ease or difficulty of performing the target behavior, accounting for anticipated barriers, resource access, and past experience. The COVID-19 Applicability Measure specifically translates the concept of perceived control into an operationalized metric of realized behavioral feasibility. When external structural impediments (e.g., retail store overcrowding, absence of teleworking options, or urgent childcare requirements) outstrip individual capability, even deeply held protective attitudes fail to translate into action. The scale maps this control boundary across 20 distinct behavioral scenarios.

The Health Action Process Approach (HAPA)

Ralf Schwarzer’s Health Action Process Approach (HAPA) bisects health behavior change into a pre-intentional motivational phase and a post-intentional volitional phase. The volitional phase is further segmented into planning, maintenance self-efficacy, and recovery self-efficacy, mediated by situational coping mechanisms and contextual barriers. The Applicability Measure targets the action execution and maintenance components of HAPA. In the context of early-2021 pandemic restrictions, individuals had already formulated intentions to protect themselves and their fetuses or infants; however, the volitional realization of these goals required navigating complex environmental demands. The scale captures where the volitional execution breakdown occurs across distinct behavioral domains.

The Social-Ecological Model

Urie Bronfenbrenner’s Social-Ecological Model posits that health behaviors cannot be understood solely as individual cognitive choices; they are embedded within microsystems (immediate family, infants, partners), mesosystems (workplaces, extended family such as grandparents), exosystems (transit systems, municipal public health ordinances), and macrosystems (national infection rates, cultural health norms). The Applicability Measure reflects these nesting dynamics: items tapping grandparent childcare avoidance assess microsystemic/mesosystemic negotiations, while items tapping public transit avoidance or remote work measure exosystemic constraints.

Perinatal Health Psychology and Maternal Protective Motivation

In perinatal health psychology, maternal behavioral adaptation is guided by unique biological, neuroendocrine, and psychological pressures aimed at safeguarding offspring survival. During a respiratory pathogen outbreak characterized by unknown vertical transmission risks, expectant and postpartum mothers experience heightened affective anxiety and maternal protective motivation. However, physiological limitations (e.g., fatigue, mobility constraints in late pregnancy) and maternal care burdens (e.g., frequent pediatric appointments, nutritional demands) clash with strict isolation directives. The theoretical framework of this instrument explicitly recognizes that maternal protective behaviors operate under heightened internal conflict, making “applicability” a far more ecologically valid metric than pure ideological compliance.

Validity

The construct, structural, and clinical validity of the COVID-19 Applicability Measure was established in an empirical investigation by Wirtz et al. (2023), published in Diagnostica. The initial scale items were developed using a mixed-methods qualitative-to-quantitative pipeline: national pandemic prevention frameworks and federal informational guidelines were synthesized alongside structured interviews conducted with expectant mothers and mothers of infants to ensure high content and face validity.

Structural and Factorial Validity

Structural validity was evaluated through ordinal factor analytic procedures designed specifically to address the polytomous, non-normal characteristics of behavioral ratings. Global goodness-of-fit indices demonstrated an exceptional representation of the empirical data by the hypothesized five-factor solution:

• Comparative Fit Index (CFI): $.977$ (exceeding the standard $.95$ benchmark for exemplary model fit).
• Tucker-Lewis Index (TLI): $.957$ (exceeding the standard $.95$ threshold).
• Standardized Root Mean Square Residual (SRMR): $.036$ (well below the conservative $.08$ cutoff, demonstrating negligible residual covariance).

The extraction of five latent dimensions was further corroborated by the Kaiser-Guttman criterion, where five empirical eigenvalues exceeded unity (Factor 1: 6.96, Factor 2: 1.61, Factor 3: 1.42, Factor 4: 1.29, Factor 5: 1.09), followed by a sharp drop-off (Factor 6: 0.89). Together, these five rotated factors accounted for 62% of the cumulative variance in the 20-item battery.

Criterion and Typological Validity (Latent Class Analysis)

Wirtz et al. (2023) supplemented factor modeling with latent class analysis (LCA), demonstrating that individual response profiles across the applicability dimensions systematically differentiated distinct behavioral phenotypes within the perinatal cohort. These latent classes differentiated mothers who experienced comprehensive applicability across all domains from those who demonstrated selective domain-specific execution deficits (such as severe impediments in avoiding family-based childcare or navigating remote work). This confirmed the instrument’s discriminant sensitivity in capturing real-world heterogeneity in health-protective execution.

Reliability

The internal consistency of the COVID-19 Applicability Measure was evaluated across its five latent subscales using Cronbach’s alpha ($lpha$). Across the subscales, reliability coefficients ranged from insufficient/modest to moderate/satisfactory ($lpha = .562$ to $.777$):

• The more homogeneous behavioral clusters (such as core physical hygiene and direct personal distancing practices) exhibited satisfactory internal consistencies approaching or exceeding $.70$ to $.78$.
• Specific sub-dimensions exhibited lower coefficients (e.g., $lpha pprox .56$ to $.65$).

In psychometrics, behavioral inventories measuring multidimensional, ecologically constrained activities frequently yield lower internal consistencies than homogeneous cognitive, personality, or affective scales. Individual compliance with one specific measure (e.g., using a digital tracing app) does not necessarily require or correlate perfectly with another measure within the same cluster (e.g., antigen testing before visits). Rather than indicating poor measurement quality, these modest alpha values reflect the breadth of the behavioral domain and the causal-formative nature of environmental barriers. Each item represents an autonomous behavioral hurdle shaped by distinct external constraints.

Factor Analysis

The dimensional architecture of the 20 items was thoroughly investigated via an Exploratory Factor Analysis (EFA) utilizing an ordinal estimator to accommodate the ordered categorical nature of the 6-point response scale. An oblique GEOMIN rotation was applied, which allows latent dimensions to correlate realistically while minimizing the epistemic cross-loadings of individual items.

Eigenvalue Distribution and Dimensionality Criteria

According to the classical Kaiser-Guttman criterion ($lambda > 1.0$), five distinct latent factors were retained from the correlation matrix. The scree plot demonstrated a severe inflection point between the first factor and subsequent factors, followed by a leveling off after the fifth factor:

• Factor 1: $lambda = 6.96$ (capturing substantial primary variance across baseline protective routines)
• Factor 2: $lambda = 1.61$
• Factor 3: $lambda = 1.42$
• Factor 4: $lambda = 1.29$
• Factor 5: $lambda = 1.09$
• Factor 6: $lambda = 0.89$ (discarded, below the Kaiser-Guttman threshold)

Collectively, the five extracted GEOMIN-rotated factors accounted for 62% of the total empirical variance, demonstrating robust explanatory parsimony.

Factor Fit and Structural Alignment

The five resulting latent dimensions showed clear alignment with public health behavioral domains:

1. Factor I: Apply Hygiene Measures: High factor loadings from items assessing distance maintenance (1.5m), regular indoor ventilation, general mask wearing, mask wearing during spontaneous conversations, hand hygiene, and surface/hand disinfection.
2. Factor II: Avoid Contact with Other People: Marked by substantial loadings from items concerning social contact minimization, restriction of meetings to rule-compliant peers, and the avoidance of informal infant care by grandparents or acquaintances.
3. Factor III: Avoid Public Transportation: Defined predominantly by items addressing mobility management and transit abstention.
4. Factor IV: Stay at Home: Substantial loadings from avoidance of public venues (libraries, eateries), execution of telecommuting, off-peak grocery shopping, avoiding interregional transit, and utilization of digital communication and delivery services.
5. Factor V: Check Infection Status: Strong loadings from information-seeking regarding current public ordinances, engagement with the national digital exposure tracing app, and pre-exposure antigen testing before encounters with vulnerable individuals.

Instrument / Measurement Tool

• Instrument Name: COVID-19 Applicability Measure
• Authors: Markus Antonius Wirtz, Andrea Heiberger, Carolin Dresch, and Anja Alexandra Schulz
• Test Type: Original self-report survey questionnaire / behavioral applicability inventory
• Item Count: 20 items
• Administration Format: Standardized self-administered questionnaire (paper-and-pencil or online digital survey)
• Target Population: Adults (18 years and older); evaluated and validated specifically in expectant mothers and mothers of infants
• Original Language: German (Maßnahmen zur Eindämmung der COVID-19-Pandemie: Anwendbarkeit im Alltag)
• Response Scale: 6-point reversed Likert-type rating scale:
  • 1 = Very good (Sehr gut)
  • 2 = Good (Gut)
  • 3 = Rather good (Eher gut)
  • 4 = Rather poor (Eher schlecht)
  • 5 = Poor (Schlecht)
  • 6 = Very poor (Sehr schlecht)
• Scoring and Directionality:
  • Items are scored numerically from 1 to 6. Lower numerical scores (e.g., 1, 2) reflect superior perceived applicability and ease of implementation in daily life. Higher numerical scores (e.g., 5, 6) reflect severe applicability deficits and behavioral implementation friction.
  • Researchers may calculate mean subscale scores across each of the five dimensions, or reverse-code items (1 = 6, 2 = 5, 3 = 4, 4 = 3, 5 = 2, 6 = 1) if positive indexing of feasibility is preferred for clinical or structural equation modeling.
• Subscale Composition:
  1. Apply hygiene measures (Items tapping distancing, room ventilation, masks, handwashing, disinfection)
  2. Avoid contact with other people (Items tapping contact restriction, peer compliance, avoiding grandparent care)
  3. Avoid public transportation (Item tapping avoidance of public buses, trams, and trains)
  4. Stay at home (Items tapping public venues, remote work, off-peak retail, travel avoidance, digital tools)
  5. Check infection status (Items tapping legal updates, digital tracking apps, pre-contact antigen testing)

Permissions & Fee and Test Year

• Publication Year: 2023 (evaluating behavioral retrospection to the early-2021 pandemic wave)
• Copyright & Licensing: The instrument and its primary validation study are distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0).
• Fee: Free of charge ($0.00). No licensing or administration fees are required for academic, clinical, or non-commercial research use.
• Permissions: Researchers, psychologists, and epidemiologists may freely administer, adapt, translate, and reproduce the measure, provided that appropriate citation and attribution are credited to the original authors (Wirtz et al., 2023) and the publication source.

References

• Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T
• Bronfenbrenner, U. (1979). The Ecology of Human Development: Experiments by Nature and Design. Harvard University Press.
• Schwarzer, R. (2008). Modeling health behavior change: How to predict and modify the adoption and maintenance of health behaviors. Applied Psychology, 57(1), 1–29. https://doi.org/10.1111/j.1464-0597.2007.00325.x
• Wirtz, M. A., Heiberger, A., Dresch, C., & Schulz, A. A. (2023). Erfassung der Bedeutung von Maßnahmen zur Eindämmung der COVID-19-Pandemie und deren Anwendbarkeit im Alltag von Schwangeren und Müttern von Säuglingen: Ordinale Faktorenanalyse und Latente Klassenanalyse [Assessment of the importance of COVID-19 prevention measures and their applicability in the daily life of pregnant women and mothers of infants: Ordinal factor analysis and latent class analysis]. Diagnostica, 69(1), 14–24. https://doi.org/10.1026/0012-1924/a000293

Items of the Scale