COVID-19 Message Fatigue, Information Overload, Message Avoidance, and Behavioral Intention–Model Inventory

1. Abstract

The COVID-19 Message Fatigue, Information Overload, Message Avoidance, and Behavioral Intention–Model Inventory (Sun & Lee, 2023) is a multidimensional psychometric measurement system developed to evaluate the psychological mechanisms linking digital infodemics to public health non-compliance. Grounded in the Stressor-Strain-Outcome (SSO) theoretical framework, this instrument assesses how environmental cognitive stressors in social media environments translate into affective strain and downstream behavioral avoidance. Specifically, the model conceptualizes Information Overload (IO) as an environmental stressor, COVID-19 Message Fatigue (comprising two distinct psychological dimensions: emotional exhaustion and subjective tedium) as psychological strain, and subsequent Message Avoidance alongside reduced Behavioral Intention (encompassing personal hygiene, social restriction adherence, and vaccination willingness) as behavioral outcomes.

The finalized inventory comprises 26 self-report items evaluated on 5-point and 7-point Likert-type scales. Following exploratory factor analysis (EFA), which eliminated four psychometrically redundant or low-loading items (BI4, BI9, BI12, and BI14), confirmatory factor analysis (CFA) substantiated a robust measurement model with acceptable fit indices: χ²(288) = 791.12, p < .001, χ²/df = 2.75, Comparative Fit Index (CFI) = 0.95, Root Mean Square Error of Approximation (RMSEA) = 0.06, and Standardized Root Mean Square Residual (SRMR) = 0.05. A second-order confirmatory model for the behavioral intention construct similarly established exceptional psychometric validity (χ²(40) = 153.42, p < .001, CFI = 0.97, RMSEA = 0.06, SRMR = 0.04). Composite reliability (CR) and Cronbach’s alpha coefficients for all subscales ranged from 0.73 to 0.94, while Average Variance Extracted (AVE) ranged between 0.51 and 0.80, confirming strong convergent validity. Heterotrait-Monotrait (HTMT) ratios below the conservative 0.90 cutoff confirmed rigorous discriminant validity. The tool serves as an essential empirical apparatus for health communication researchers, epidemiologists, and behavioral scientists investigating digital infodemic management and health crisis compliance.

2. Keywords

Behavioral Intention, COVID-19 Message Fatigue, Information Overload, Message Avoidance, Social Media, Stressor-Strain-Outcome Framework, Digital Fatigue, Health Communication Assessment, Health Information Avoidance, Infodemic

3. Authors

The COVID-19 Message Fatigue, Information Overload, Message Avoidance, and Behavioral Intention–Model Inventory was developed and validated by:

• Juhyung Sun, Ph.D. — Department of Communication, University of Oklahoma, Norman, Oklahoma, United States. ORCID: 0000-0002-0302-5196. Email: [email protected].
• Sun Kyong Lee, Ph.D. — School of Media & Communication, Korea University, Seoul, Republic of Korea. ORCID: 0000-0002-8645-8922. Email: [email protected].

Correspondence regarding the instrument, theoretical derivations, or normative validation data should be addressed directly to Dr. Juhyung Sun at the Department of Communication, University of Oklahoma, 610 Elm Ave, Norman, Oklahoma, 73019, United States.

4. Purpose

The primary purpose of the inventory is to model, quantify, and explain the paradoxical phenomenon wherein saturated public health messaging during an ongoing crisis results in psychological disengagement, message tuning-out, and the eventual deterioration of protective health behaviors. In the wake of the global COVID-19 pandemic, digital platforms and social media networks subjected populations to an unprecedented inundation of algorithmic health advisories, dynamic risk updates, and regulatory guidelines—a condition formally designated by the World Health Organization as an “infodemic.” While public health officials deployed saturated communication strategies to elevate risk perception and maintain preventive compliance, sustained exposure often triggered cognitive depletion and affective backlash.

This measurement model addresses critical clinical, communicative, and public health needs by establishing an empirical bridge between communication exposure and actual behavioral adoption. Specifically, the instrument seeks to accomplish three primary objectives:

• Isolate Environmental Overload: Quantify the degree to which individuals perceive public health information volume, complexity, and velocity as exceeding their subjective cognitive processing capacity on digital social networks.
• Diagnose Psychological Strain: Measure the affective and cognitive manifest dimensions of message fatigue—distinguishing between feelings of systemic cognitive exhaustion and routine communicative boredom or tedium.
• Predict Downstream Behavioral Attrition: Capture the direct and indirect impacts of communication strain on active avoidance tactics (e.g., muting content, bypassing news feeds) and on three core domains of public health protective adherence: non-pharmaceutical hygiene behaviors (e.g., hand sanitation, mask usage), social distancing restrictions (e.g., avoiding congregate spaces, canceling travel), and pharmaceutical interventions (e.g., initial and booster vaccination uptake).

In applied research, health communication strategists utilize this inventory to optimize message frequency, identify the tipping point at which campaign volume produces diminishing returns, and design targeted intervention messages that do not trigger avoidance mechanisms among vulnerable demographics, particularly young adults who consume high volumes of social media.

5. Psychological Construct

The inventory operationalizes four interconnected psychological constructs situated within an integrated structural model:

1. Information Overload (IO)

Information overload represents a subjective state wherein the perceived volume, rate of presentation, and complexity of environmental stimuli exceed an individual’s finite cognitive capacity for information processing. Drawing upon foundational conceptualizations by information science and communication scholars, this construct evaluates users’ feelings of being overwhelmed by the sheer abundance of COVID-19 information on social networking platforms. It captures the perceptual bottleneck experienced when individuals attempt to filter, verify, and comprehend competing risk messages.

2. COVID-19 Message Fatigue

Message fatigue is defined as an aversive, depleted psychological state resulting from prolonged, repetitive, and pervasive exposure to messages concerning a specific health issue. Rather than viewing fatigue as a unidimensional construct, the inventory models it as a dual-faceted psychological strain consisting of:

• Cognitive and Emotional Exhaustion: A state of psychological depletion where the mental effort required to attend to, evaluate, and integrate recurrent crisis communications leaves the user feeling drained, overwhelmed, and emotionally spent.
• Tedium and Boredom: A motivational state characterized by subjective monotony, weariness, and disinterest triggered by uniform, unvarying, and repetitive message content across digital channels.

3. Message Avoidance

Message avoidance reflects active, deliberate cognitive and behavioral coping maneuvers enacted to evade exposure to relevant health messaging. Unlike passive inattention, avoidance represents purposeful behavior to defend cognitive equilibrium against further depletion. Tactics captured by this construct include scrolling past health-related postings, unfollowing or muting institutional accounts, closing applications when pandemic topics arise, and deliberately navigating away from conversational threads addressing the disease.

4. Behavioral Intention

Behavioral intention represents a respondent’s subjective probability that they will perform recommended preventive health actions. Grounded in the Theory of Planned Behavior, behavioral intention is operationalized in this inventory as a second-order factor comprised of three distinct first-order subdimensions:

• Hygiene Practices: Intentions to engage in non-pharmaceutical personal sanitation, such as washing hands frequently, utilizing sanitizer, wearing protective face coverings, and sanitizing personal items.
• Social Restriction Adherence: Intentions to limit interpersonal physical exposure, including social distancing, avoiding crowded indoor venues, foregoing social gatherings, and postponing non-essential domestic or international travel.
• Vaccination Willingness: Intentions to obtain recommended immunizations, schedule booster doses, and adhere to established pharmaceutical preventive schedules.

6. Theoretical Framework

The architecture of the inventory is theoretically anchored in the Stressor-Strain-Outcome (SSO) framework (Koeske & Koeske, 1993; Cooper, Dewe, & O’Driscoll, 2001), a transactional stress paradigm adapted from organizational psychology to health communication and human-computer interaction contexts.

The Stressor-Strain-Outcome (SSO) Continuum

The SSO framework posits that environmental stimuli acting as external demands (Stressors) trigger disruptive internal psychological, emotional, or cognitive reactions within the person (Strains), which subsequently drive observable behavioral adaptations, coping strategies, or performance deficits (Outcomes). Within this inventory:

• The Stressor: Information Overload acts as the ambient environmental stressor. The modern high-choice media environment continuously broadcasts crisis messaging, creating sensory and cognitive saturation.
• The Strain: Message Fatigue (exhaustion and tedium) represents the mediating internal strain. When cognitive resources are repeatedly drained without adequate recovery intervals, the individual suffers psychological burnout specifically targeted toward the communicative stimulus.
• The Outcome: Message Avoidance and diminished Preventive Behavioral Intentions serve as the primary behavioral outcomes. To mitigate emotional strain, users employ behavioral flight mechanisms (avoidance), which compromises preventive compliance by diminishing institutional trust and perceived risk salience.

Integration with Cognitive and Motivational Models

The inventory additionally incorporates principles from the Cognitive Load Theory (Sweller, 1988) and the Limited Capacity Model of Motivated Mediated Message Processing (LC4MP) (Lang, 2000). LC4MP posits that human cognitive resources for encoding, storage, and retrieval are strictly limited. When incoming message streams exceed cognitive bandwidth, cognitive overload ensues, prompting subconscious defensive motivational systems that down-regulate attention and prompt selective exposure or active avoidance. Furthermore, the model aligns with Psychological Reactance Theory (Brehm, 1966), explaining how pervasive, coercive health mandates generate affective resistance, viewing message avoidance and non-adherence as psychological assertions of autonomy.

7. Validity

The psychometric validity of the instrument was rigorously established through multiple analytical methods in a study comprising young adults in the United States, an age demographic demonstrating exceptionally high social media integration and susceptibility to digital infodemics.

Convergent Validity

Convergent validity demonstrates that the items operationalizing each latent construct share a high proportion of common variance. In this inventory, convergent validity was established via the Average Variance Extracted (AVE) for all measurement constructs:

• All AVE values across the constructs ranged comfortably between 0.51 and 0.80.
• Because every AVE value met or exceeded the universally accepted methodological benchmark of 0.50 (Fornell & Larcker, 1981; Hair et al., 2010), each latent variable explains more than half of the variance of its corresponding indicator items, verifying robust convergent validity.

Discriminant Validity

Discriminant validity confirms that each latent construct represents a distinct psychological phenomenon not captured by other measures within the structural model. To overcome known limitations of the traditional Fornell-Larcker criterion, the authors evaluated discriminant validity using the rigorous Heterotrait-Monotrait Ratio of Correlations (HTMT):

• All estimated HTMT ratios between paired latent variables were strictly below the conservative threshold criterion of 0.90 (and adhered to the stricter 0.85 threshold across the majority of variable pairings).
• These findings confirm that Information Overload, Message Fatigue (Exhaustion and Tedium), Message Avoidance, and the Behavioral Intention dimensions are empirically distinct and non-redundant.

Common Method Bias

Because all survey data were collected via self-report instruments, Common Method Variance (CMV) posed a potential threat to validity. The authors addressed this via Harman’s single-factor test. An unrotated exploratory factor analysis demonstrated that the first, single extraction factor accounted for 41.76% of the total variance. Because this value falls well below the standard cut-off threshold of 50%, common method bias was determined not to be a confounding threat to the validity of the structural findings.

8. Reliability

The internal consistency and composite reliability of the instrument were assessed across all primary constructs and subdimensions, demonstrating exceptional measurement precision and replicability.

Internal Consistency Metrics

Both Cronbach’s alpha (α) and Composite Reliability (CR) coefficients were evaluated to account for potential tau-equivalence violations:

• Cronbach’s Alpha Range: Observed values ranged from α = 0.73 to α = 0.94 across the various model subscales.
• Composite Reliability Range: Calculated values similarly ranged from CR = 0.73 to CR = 0.94.

All values comfortably exceeded the standard psychometric cutoff of 0.70 for research instruments (Nunnally & Bernstein, 1994) and surpassed the 0.80 benchmark indicative of high-precision diagnostic and modeling applications. The observed consistency indicates that items within each respective domain (Information Overload, Message Fatigue dimensions, Avoidance, and Behavioral Intention facets) systematically tap into their targeted psychological phenomena with minimal measurement error.

9. Factor Analysis

The structural dimensionality of the inventory was systematically established through sequential exploratory and confirmatory factor analytic procedures in accordance with psychometric best practices (Hair et al., 2010).

Exploratory Factor Analysis (EFA) & Item Refinement

Initial construct screening involved EFA utilizing principal axis factoring and oblique rotation to account for expected theoretical intercorrelations among the constructs. During this phase, factor loadings were inspected against a stringent retention criterion (≥ 0.40). Four items from the initial behavioral intention pool were identified as problematic:

• Items BI4, BI9, BI12, and BI14 exhibited low primary factor loadings below 0.40 and secondary cross-loadings.
• In accordance with factor retention guidelines, these four items were systematically eliminated, yielding a refined, psychometrically sound 26-item final instrument.

Confirmatory Factor Analysis (CFA) & Global Model Fit

Confirmatory factor analysis using maximum likelihood estimation was conducted on the full measurement model containing the refined item set. The measurement model demonstrated excellent goodness-of-fit indices across absolute, parsimonious, and incremental metrics:

• Chi-Square (χ²): 791.12 with 288 degrees of freedom (p < .001)
• Normed Chi-Square (χ²/df): 2.75 (well within the acceptable range of ≤ 3.0)
• Comparative Fit Index (CFI): 0.95 (exceeding the ≥ 0.95 benchmark for superior fit)
• Root Mean Square Error of Approximation (RMSEA): 0.06 (below the ≤ 0.08 threshold, indicating minimal approximation error; 90% CI [0.05, 0.07])
• Standardized Root Mean Square Residual (SRMR): 0.05 (well below the ≤ 0.08 cutoff)

Second-Order CFA for Behavioral Intention

To confirm that the three distinct behavioral intention facets (hygiene, social restrictions, and vaccination) reflect an overarching preventive behavioral orientation, a second-order CFA was executed. The second-order hierarchical model exhibited superior fit:

• χ²(40) = 153.42, p < .001
• χ²/df = 3.84
• CFI = 0.97
• RMSEA = 0.06
• SRMR = 0.04

These empirical findings confirm that behavioral compliance during a health crisis operates as a multi-tiered latent construct, supporting the hierarchical integrity of the measurement tool.

10. Instrument / Measurement Tool

The inventory is structured as an operational self-report battery designed for digital, paper-and-pencil, or mobile administration. Below are the structural parameters of the instrument:

• Instrument Designation: COVID-19 Message Fatigue, Information Overload, Message Avoidance, and Behavioral Intention–Model Inventory.
• Test Classification: Multi-construct Psychometric Assessment / Structural Model Inventory.
• Target Demographics: Adults (18 years and older), specifically validated with young adults (ages 18–29) and individuals in their thirties (30–39). Adaptable across diverse gender populations (Male, Female, Nonbinary).
• Administration Modality: Self-administered electronic survey or paper-and-pencil questionnaire.
• Total Item Count: 26 finalized items (post-EFA item reduction).
• Response Formats: Multi-point Likert-type scales:
  • Information Overload & Message Avoidance: Evaluated on 5-point Likert scales ranging from 1 (“Strongly Disagree”) to 5 (“Strongly Agree”).
  • Message Fatigue & Behavioral Intention: Evaluated on 7-point Likert scales ranging from 1 (“Strongly Disagree”) to 7 (“Strongly Agree”).
• Scoring and Computational Rules:
  • Subscale scores are calculated by computing the unweighted arithmetic mean of the indicator items belonging to each construct.
  • Higher scores on Information Overload denote greater perceived cognitive saturation from digital communication.
  • Higher scores on Exhaustion and Tedium reflect greater affective depletion and communication burnout.
  • Higher scores on Message Avoidance signify active deliberate disengagement from pandemic communications.
  • Behavioral Intention scores can be interpreted at the first-order subdimension level (Hygiene, Social Restrictions, Vaccination) or aggregated into a composite second-order Preventive Health Intention Index.

11. Permissions & Fee and Test Year

• Year of Development: 2023.
• Copyright & Intellectual Property: © 2023 Sun & Lee. Published by Springer Nature in Current Psychology.
• Fee: Free for non-commercial academic, scientific, and educational research purposes.
• Commercial Exploitation: Commercial use, organizational consulting use, or integration into proprietary software platforms requires explicit written licensing permissions from the copyright holders.
• Permissions Inquiries: Researchers seeking to reproduce the scale, adapt items for alternative epidemics, or obtain specific item-level implementation materials should contact the corresponding author, Dr. Juhyung Sun, at [email protected].

12. References

The theoretical, psychometric, and empirical foundations of this instrument are documented in the following peer-reviewed literature:

• Brehm, J. W. (1966). A theory of psychological reactance. Academic Press.
• Cho, H. (2004). The role of message fatigue and counterarguing in persuasive health campaigns (Doctoral dissertation). Rutgers, The State University of New Jersey.
• Cooper, C. L., Dewe, P. J., & O’Driscoll, M. P. (2001). Organizational stress: A review and critique of theory, research, and applications. SAGE Publications. https://doi.org/10.4135/9781452231235
• Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50. https://doi.org/10.1177/002224378101800104
• Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate data analysis (7th ed.). Prentice Hall.
• Koeske, G. F., & Koeske, R. D. (1993). A preliminary test of the stressor-strain-outcome model for clinicians working with emotionally disturbed clients. Journal of Social Service Research, 17(3–4), 107–126. https://doi.org/10.1300/J079v17n03_06
• Lang, A. (2000). The limited capacity model of mediated message processing. Journal of Communication, 50(1), 46–70. https://doi.org/10.1111/j.1460-2466.2000.tb02833.x
• Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw-Hill.
• So, J., Kim, S., & Cohen, H. (2017). Message fatigue: Conceptual definition, operationalization, and initial empirical examination. Communication Monographs, 84(1), 5–29. https://doi.org/10.1080/03637751.2016.1250429
• Sun, J., & Lee, S. K. (2023). “No more COVID-19 messages via social media, please”: The mediating role of COVID-19 message fatigue between information overload, message avoidance, and behavioral intention. Current Psychology, 42(24), 20347–20361. https://doi.org/10.1007/s12144-023-04726-7
• Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. https://doi.org/10.1207/s15516709cog1202_4
• Wisniewski, P., & Lu, H. (2010). Resolving information overload on social media through social filtering: An empirical study. In Proceedings of the American Society for Information Science and Technology, 47(1), 1–10. https://doi.org/10.1002/meet.14504701235

13. Items of the Scale