A Preliminary Study of the Development of a Theory-Based Scale for Human Norovirus Prevention Behaviors: Distinguishing Handwashing and Social Distancing

Abstract

Human norovirus is a leading cause of acute gastroenteritis, with prevention largely dependent on individual hygiene behaviors. Understanding the cognitive determinants of these behaviors is essential for designing effective prevention strategies; however, validated hygiene-specific measures grounded in behavioral theory remain limited. This study presents an initial, exploratory examination of Protection Motivation Theory constructs, some of which overlap with core principles of Social Cognitive Theory, as they relate to hygiene behaviors for human norovirus prevention. Survey data were collected from a sample of college students, and principal component analysis was used to examine the dimensionality and internal structure of theory-informed scale items. Sampling adequacy was supported by Kaiser–Meyer–Olkin indices indicating acceptable to meritorious suitability for exploratory analysis. Results revealed coherent component structures for threat and coping appraisal constructs, while a distinct response cost component did not emerge. Item retention decisions were guided by both statistical criteria and theoretical interpretability. Findings are discussed within the context of early-stage scale development, with attention to implications for theory-informed measurement refinement and future validation efforts. This preliminary work provides a foundation for advancing the assessment of hygiene behaviors relevant to human norovirus prevention.

1. Introduction

Human norovirus (hNoV) remains a leading cause of acute gastroenteritis worldwide, with prevention relying largely on adherence to basic hygiene and behavioral practices in communal settings. In the United States, it accounts for an estimated 19–21 million illnesses annually and contributes to significant disruptions in institutional environments such as schools, healthcare facilities, and college campuses [1].

College campuses are highly susceptible to hNoV outbreaks due to dense living conditions, shared dining facilities, and frequent social interactions. Over the past decade, multiple campus outbreaks have resulted in hundreds of reported cases per incident, forcing temporary closures of facilities, transitions to remote instruction, and extensive sanitation efforts [2,3,4]. These outbreaks not only pose public health concerns but also disrupt academic continuity and campus operations.

Symptoms of hNoV, including vomiting, diarrhea, and dehydration, typically persist for 24 to 72 h, but the rapid transmissibility of the virus often amplifies its impact across campus populations [1]. As a result, outbreaks frequently lead to widespread absenteeism among students, faculty, and staff, creating both educational and economic consequences. Despite ongoing advancements in vaccine development, there is currently no widely available, approved vaccine for hNoV, making prevention strategies, particularly hygiene behaviors such as proper handwashing, surface disinfection, and isolation while symptomatic, the primary means of mitigating transmission [4,5,6].

Given the reliance on individual and collective prevention behaviors, understanding the factors that influence compliance with recommended hygiene practices is critical. However, accurately assessing these behaviors requires the use of valid measurement instruments [7,8,9,10]. The development of psychometrically sound scales is essential to ensure that constructs such as hygiene perceptions, risk awareness, and behavioral intentions are measured consistently and accurately across populations [9,11,12,13]. Without valid scales, findings may be inconsistent, biased, or difficult to generalize, ultimately limiting the effectiveness of interventions and policy recommendations. Furthermore, validated instruments enable researchers and practitioners to track changes over time, compare results across studies, and design evidence-based strategies tailored to specific populations, such as college students.

Accordingly, the present study emphasizes the importance of developing and validating measurement scales that capture hygiene-related perceptions and behaviors in the context of hNoV prevention. By strengthening measurement rigor, this research contributes to a more precise understanding of how to promote effective disease prevention behaviors and reduce the impact of outbreaks in campus environments.

Effective prevention of hNoV transmission depends heavily on adherence to hygiene-related behaviors, particularly handwashing and situational avoidance of illness exposure in shared environments. While public health guidance consistently emphasizes these behaviors, less attention has been paid to the systematic measurement of the cognitive and motivational factors that influence individuals’ willingness to adopt and sustain them. Advancing prevention efforts therefore requires tools that can validly capture not only hygiene behaviors themselves but also the perceived risks, efficacy beliefs, and intentions that underlie compliance.

Through preliminary scale development and providing an initial dimensionality assessment of a theory-based instrument that distinguishes between handwashing and social distancing intentions, this study addresses a critical measurement gap in infection prevention research. The resulting scale offers a structured approach for assessing hygiene-related perceptions and behavioral intentions in communal settings where hNoV transmission is common and pharmaceutical interventions remain unavailable. Strengthening such measurement capacity supports more precise evaluation of hygiene promotion strategies and contributes to efforts aimed at reducing outbreak risk through behaviorally informed prevention practices.

The present study represents an initial, exploratory phase of scale development focused on understanding the cognitive determinants of hygiene behaviors relevant to human norovirus (hNoV) prevention. Drawing on Protection Motivation Theory (PMT), this work emphasizes early identification of meaningful dimensions and item performance rather than definitive construct confirmation. Given the limited availability of validated, hygiene-specific PMT instruments and the evolving behavioral landscape following the COVID-19 pandemic, an exploratory approach was warranted to assess the salience and structure of theoretically informed constructs prior to more advanced psychometric testing. Findings from this stage are intended to inform subsequent validation efforts, including reliability assessment, confirmatory analyses, and tests of predictive validity.

2. Review of Literature

Understanding and promoting hygiene-related behaviors that reduce infectious disease transmission requires attention not only to recommended practices but also to the cognitive and motivational processes that shape individuals’ compliance. In the context of hNoV, prevention relies almost entirely on non-pharmaceutical interventions, making behavior a central determinant of outbreak risk and severity. Consequently, theoretical frameworks that explain why individuals adopt, or fail to adopt, protective hygiene behaviors provide an essential foundation for both measurement and intervention. The present study draws primarily on Protection Motivation Theory (PMT), supplemented by key elements of Social Cognitive Theory (SCT), to inform the development of a theory-based measurement instrument for hNoV prevention behaviors.

2.1. Protection Motivation Theory and Preventive Health Behaviors

PMT is a widely used framework for explaining how individuals cognitively evaluate health threats and decide whether to engage in protective behaviors. Originally developed by Rogers and colleagues, PMT posits that behavior change is driven by two parallel appraisal processes, threat appraisal and coping appraisal [14,15]. Threat appraisal reflects perceptions of the severity of a health threat and one’s susceptibility to it, whereas coping appraisal encompasses beliefs about response efficacy, self-efficacy, and response costs associated with recommended behaviors [13,15].

Extensive empirical research has demonstrated the utility of PMT in predicting a wide range of preventive health behaviors, including infection prevention, vaccination uptake, and hygiene practices [10,16,17]. In communicable disease contexts, PMT constructs have been shown to predict behaviors such as hand hygiene, illness avoidance, and adherence to public health guidance, highlighting the relevance of the framework for studying hygienic practices that mitigate disease spread [11,18].

In the case of hNoV, where no widely available vaccine exists, PMT offers a particularly appropriate lens for understanding prevention behavior. Individuals’ perceptions of hNoV severity and susceptibility may influence whether they recognize infection as a credible threat, while coping appraisals determine whether they believe actions such as handwashing or avoiding sick individuals are both effective and feasible. Accordingly, PMT provides a conceptual structure for examining how hygiene-related beliefs translate into behavioral intentions.

2.2. Social Cognitive Theory and the Role of Self-Efficacy

SCT, developed by Bandura, complements PMT by emphasizing the role of self-efficacy in behavior change [19]. SCT posits that behavior is shaped through reciprocal interactions among personal factors, environmental influences, and behavioral experiences. Central to this framework is self-efficacy, defined as an individual’s confidence in their ability to perform a specific behavior under varying circumstances.

Research consistently identifies self-efficacy as a robust predictor of preventive health behaviors, including hygiene practices and disease avoidance [20,21]. Individuals who believe they are capable of washing their hands consistently or avoiding exposure to illness are more likely to enact these behaviors, even when faced with situational barriers or social pressures.

The emphasis on self-efficacy within SCT aligns closely with PMT’s coping appraisal processes. Both frameworks suggest that perceived threat alone is insufficient to motivate action; individuals must also believe that recommended behaviors are effective and that they possess the capability to carry them out successfully. Integrating PMT and SCT therefore allows for a more comprehensive understanding of hygiene-related behavior, particularly in communal environments where behavioral decisions are influenced by both personal beliefs and contextual constraints.

2.3. Norovirus Transmission and the Importance of Individual Behaviors

Historically, hNoV research has emphasized food handling and foodservice environments as primary sources of transmission [22,23]. However, more recent evidence highlights the substantial role of person-to-person transmission in dense, communal settings such as healthcare facilities, cruise ships, and college campuses [5,24,25]. Outbreaks in these environments are often introduced by symptomatic or pre-symptomatic individuals and amplified through close contact, contaminated surfaces, and inadequate hygiene practices [3,4,26].

College campuses are particularly vulnerable due to shared living spaces, communal dining facilities, and frequent social interaction [2,4]. Studies have shown that behaviors such as inadequate hand hygiene, failure to isolate while symptomatic, and low perceived risk significantly contribute to the spread and duration of hNoV outbreaks in these settings [10,11,27]. Despite widespread awareness of recommended prevention strategies, adherence remains inconsistent, underscoring the need to better understand the psychological drivers of compliance.

2.4. Measurement Gaps in Hygiene-Related Prevention Research

Although PMT and SCT have been widely applied to health behavior research, relatively few studies have focused on developing validated measurement instruments specifically designed to assess hygiene-related perceptions and behavioral intentions in the context of hNoV prevention. Existing research often relies on single-item measures or general hygiene scales that do not distinguish between different types of preventive behaviors or their underlying cognitive drivers.

Accurate measurement is critical for advancing hygiene research and practice. Psychometrically sound instruments allow researchers to compare findings across populations, evaluate intervention effectiveness, and identify which beliefs should be targeted to promote behavior change [7,8,12]. Without validated scales, efforts to improve hygiene compliance may be poorly targeted or based on incomplete understanding of behavior determinants.

Accordingly, the present study aims to address this gap through preliminary scale development and initial dimensionality assessment of a multidimensional instrument grounded in PMT and informed by SCT. By distinguishing between handwashing and social distancing intentions and assessing their associated threat and coping appraisals, this research provides a more nuanced approach to measuring hygiene-related prevention behaviors in high-risk communal environments.

3. Materials and Methods

Building on the theoretical foundations of PMT and SCT and following IRB which included the informed consent statement approval, a quantitative research design was used in preliminary scale development and initial dimensionality assessment, comprising a scale measuring factors that influence intentions to engage in protective health behaviors related to hNoV. In scale item wording, the term norovirus was used without specifying human norovirus (hNoV), as the study population consisted exclusively of human participants. This decision was intended to enhance readability and reduce unnecessary repetition while preserving construct validity and interpretive clarity within a human behavioral context. Additionally, participants were instructed to read each survey item under the context of “While on a college campus” (e.g., While on a college campus my chances of contracting Norovirus are quite small). This methodology outlined the procedures used to operationalize key constructs, collect data from an opportunity sample of college student population, and assess the validity of the proposed measurement instrument with data collected via an online survey system (Qualtrics). Participants initiated the survey by responding to an informed consent inquiry, indicating agreement to proceed.

3.1. Description of the Sample

The sample for this study comprised 120 undergraduate students enrolled in specific classes at a large public university in the Mid-south United States. Participants ranged in age from 18 to 25 years, with a mean age of 20.36 years (SD = 2.00). The sample was relatively balanced by sex, with 48.3% identifying as male (n = 58) and 51.7% identifying as female (n = 62). With respect to academic classification, a proxy for age, the sample was primarily composed of sophomores (45.8%, n = 55) and juniors (29.2%, n = 35), followed by freshmen (15.8%, n = 19) and seniors (8.3%, n = 10), and there were no graduate students. The racial composition of the sample was predominantly White (84.2%, n = 101), with smaller representations of Asian (5.8%, n = 7), Hispanic (5.0%, n = 6), Black or African American (4.2%, n = 5), and American Indian (0.8%, n = 1) participants.

In addition to demographic characteristics, two screening questions were included to assess participants’ prior exposure to hNoV-related information and experiences (e.g., prior illness or awareness of outbreaks). Responses indicated that most participants had at least a basic awareness of hNoV, although direct personal experience with the illness varied across the sample [28]. These preliminary measures provided important context for interpreting participants’ perceptions of severity, susceptibility, and behavioral intentions related to hNoV prevention.

3.2. Measures

3.2.1. Protection Motivation

Consistent with PMT, protection motivation is most appropriately operationalized through behavioral intentions, which reflect an individual’s readiness to adopt protective health behaviors [29,30]. In the context of hNoV, protection motivation was conceptualized as two distinct, yet related, behavioral domains: Handwashing intentions and social distancing intentions. Separating these constructs allows for a more precise assessment of behavioral responses, as each represents a unique prevention strategy with different practical and psychological considerations.

Handwashing Intentions

Handwashing was defined as washing hands thoroughly with soap and water for at least 20 s, consistent with guidelines from the CDC [22]. This construct was measured using five items adapted from prior PMT-based health behavior research [6,12,29] Items were modified to reflect the context of hNoV prevention (e.g., “I would wash my hands to protect myself from a Norovirus infection,” “I would wash my hands after using the restroom”).

Social Distancing Intentions

Social distancing was defined as intentionally increasing physical distance from individuals who are ill or potentially infectious [1,25]. Like handwashing, this construct was measured using five adapted items reflecting avoidance behaviors in high-risk environments (e.g., “I would intentionally avoid people who are sick to protect myself from a Norovirus infection,” “I would not sit next to someone who is actively sick in the classroom”).

All protection motivation items were measured on a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree), with higher scores indicating stronger behavioral intentions to engage in preventive actions.

3.2.2. Threat Appraisal Constructs

Perceived Severity

Perceived severity reflects the extent to which an individual believes a health threat would have serious consequences [14,29]. Five items were adapted from prior PMT research and contextualized for hNoV (e.g., “Norovirus would make me very sick,” “Norovirus would cause me to miss class/work”) [6,12]. Responses were measured on a 7-point Likert scale (1 = strongly disagree; 7 = strongly agree), with higher scores indicating greater perceived severity.

Perceived Susceptibility

Perceived susceptibility refers to an individual’s belief regarding their likelihood of experiencing a health threat [15,29]. Five items were adapted from Lwin and colleagues and modified for hNoV (e.g., “It is possible that I will get Norovirus,” “The chance of my peers getting Norovirus is rather large”) [9,12]. Responses were measured on a 7-point Likert scale, with higher scores indicating greater perceived vulnerability.

Coping Appraisal Constructs

To align with the multidimensional nature of prevention behaviors, coping appraisal constructs were measured separately for handwashing and social distancing. The coping appraisal constructs consisted of response efficacy, self-efficacy, and response costs for handwashing and social distancing.

Response Efficacy

Response efficacy refers to the belief that a recommended behavior effectively reduces or prevents a health threat [17].

Handwashing Response Efficacy: Items were adapted from prior PMT research and modified for hNoV (e.g., “Handwashing would impact whether or not I get sick from Norovirus”) [9].

Social Distancing Response Efficacy: Parallel items assessed beliefs about the effectiveness of avoiding sick individuals (e.g., “Avoiding people who are sick would reduce my chances of a Norovirus infection”).

Self-Efficacy

Self-efficacy reflects an individual’s confidence in their ability to perform a recommended behavior [17,19].

Handwashing Self-Efficacy: Items assessed confidence in performing proper hand hygiene practices (e.g., “I would be able to wash my hands when I want to”).

Social Distancing Self-Efficacy: Items assessed confidence in avoiding exposure to illness (e.g., “I would be able to avoid people who are sick when I want to”).

All coping appraisal items were measured on a 7-point Likert-type scale (1 = strongly disagree; 7 = strongly agree), with higher scores indicating stronger efficacy beliefs.

Response Costs

The construct for response cost is defined as the perceived barriers to the practice of good protective behaviors [12,17].

Response Costs Handwashing and Social Distancing: Items assessed time and convenience constraints in handwashing (e.g., “I would wash my hand every time I should, even if it takes a lot of time”; “It is not convenient to always avoid people who are sick”). The responses for these items were measured on a 7-point Likert type scale ranging from 1 (strongly disagree) to 7 (strongly agree). Higher scores indicate increased response costs.

3.3. Data Analysis

The data (n = 120) were analyzed using IBM SPSS (Version 29). Analyses focused on examining the dimensionality and internal structure of the proposed hygiene-related PMT and informed by SCT constructs during an early stage of instrument development. Principal Component Analysis (PCA) was selected as an exploratory approach to identify dominant patterns of shared variance and to inform item retention when theoretical constructs are conceptually well defined but remain empirically understudied. Given the modest sample size and the exploratory aims of the study, PCA provided a practical initial strategy for evaluating scale structure. Consistent with recommendations for early-stage scale development, PCA is appropriate when the primary goal is data reduction and the identification of a parsimonious item set, particularly under conditions of modest sample size, where common factor models may yield unstable or uninterpretable solutions [31,32]. Given the conceptual relatedness of the constructs, an oblique rotation was used.

Although PCA does not model latent constructs in the same manner as Exploratory Factor Analysis (EFA), it provides a pragmatic and empirically defensible first step for evaluating structure and refining items prior to more rigorous latent variable modeling [32]. Accordingly, the results are interpreted as preliminary, and future research using larger samples will employ EFA and Confirmatory Factor Analysis (CFA) to more fully assess the latent structure, construct validity, and model fit.

Sampling adequacy was evaluated using the Kaiser–Meyer–Olkin (KMO) measure; values above 0.60 are generally considered acceptable, values above 0.70 indicate good adequacy, and values above 0.80 reflect meritorious suitability for factor analytic techniques [33]. Component retention was guided by eigenvalues greater than one, scree plot inspection, and theoretical interpretability. Item retention criteria were guided by prior PMT-based scale development studies [9,12]. Specifically, items demonstrating cross-loadings (≥0.35) on multiple factors were removed, as were items that loaded higher on a non-targeted factor. Consistent with recommendations for exploratory instrument development, analyses prioritized parsimony and conceptual coherence over maximization of explained variance. Under oblimin rotation, negative loadings indicate inverse relationships with correlated factors and are not, in themselves, grounds for item removal. Following this, reliability analyses (McDonald’s omega) were used to estimate internal consistency without assuming equal item loadings. In line with recommendations for exploratory research, values of approximately 0.70 or higher were considered acceptable, with slightly lower values tolerated given the early stage of scale development [34,35].

4. Results

The means and standard deviations for all items are found in Supplemental Material Table S1. In addition, the initial PCA loadings for all items are found in Supplemental Materials Tables S2–S4.

4.1. Threat Appraisal Constructs

Based on the initial PCA, one item was removed from Severity and two items were removed from Susceptibility for threat appraisals because of strong cross-loadings (Supplemental Material Table S2). The subsequent PCA results provided preliminary support for a two-component threat appraisal structure, consisting of perceived severity and perceived susceptibility (

Table 1. Principal component analysis item loadings between severity and susceptibility.

Factor Loadings *
Items		SEV	SUS	h2	Corrected Item-Total r
Severity (SEV)
Norovirus would make me very sick		0.8		0.68	0.63
Norovirus would cause me to be hospitalized		0.73		0.50	0.48
Norovirus would cause me to miss class/work		0.81		0.70	0.66
Norovirus would affect my overall attitude regarding the semester		0.69		0.51	0.50
Susceptibility (SUS)
My chances of contracting Norovirus are quite small (reversed coded)			0.83	0.65	0.41
It is possible that I will get Norovirus			0.62	0.46	0.37
The chance of my peers getting Norovirus is rather large			0.77	0.7	0.53
Factor Correlation
Factor	SEV
SEV	1
SUS	0.23
KMO	0.74

* Factor loadings ≤ 0.30 omitted.

).

Factor loadings for severity items ranged from 0.69 to 0.81, while susceptibility items loaded between 0.62 and 0.83, providing preliminary evidence that retained items clustered as expected within the two proposed threat appraisal domains. Communality estimates for severity items ranged from 0.51 to 0.70, while susceptibility items ranged from 0.46 to 0.70, indicating moderate to high levels of item variance accounted for by the extracted components. These values suggest that the retained items were adequately represented by the extracted components in this exploratory analysis. The KMO measure of sampling adequacy was within the acceptable-to-good range, supporting the suitability of the data for exploratory component analysis. The corrected item-total correlations ranged from 0.50 to 0.66 for the severity items and from 0.37 to 0.53 for the susceptibility items. McDonald’s omega values of 0.66 and 0.76, respectively, indicated tolerable or acceptable internal consistency for an exploratory study.

4.2. Coping Appraisal Constructs

Based on the initial PCA, one item each was removed from Response-Efficacy Handwashing, Response-Efficacy Social Distancing and Self-Efficacy Social Distancing because of ambiguous cross-loadings (Supplement Material Table S3). In addition, the entire construct of Response Costs was removed because the initial PCA revealed that the items loaded on four rather than two expected factors and across the conceptualized domains of handwashing and social distancing (Supplemental Material Table S4). For two of the factors, the construct also did not meet the three-indicator rule for factor analytic models [36].

The coping appraisal analysis revealed a four-component solution corresponding to self-efficacy and response efficacy for both handwashing and social distancing behaviors (

Table 2. Principal component analysis item loadings among coping appraisal constructs.

Factor Loadings *
Items	SE.HW	RE.HW	RE.SD	SE.SD	h2		Corrected Item-Total r
Self-Efficacy Handwashing (SE.HW)
I would know how to wash my hands effectively to reduce my risk of Norovirus	−0.75				0.54		0.55
I would be able to wash my hands when I want to	−0.61				0.51		0.56
I would be capable of successfully following proper handwashing information	−0.88				0.75		0.72
I would have no difficulty practicing proper handwashing procedures	−0.83				0.63		0.57
I would know how to properly wash my hands to reduce my risk of Norovirus infection	−0.75				0.65		0.68
Response-Efficacy Handwashing (RE. HW)
I think handwashing would be one of the best ways to prevent an illness caused by Norovirus		0.62			0.59		0.48
Following advice about proper handwashing would help me not get sick from Norovirus		0.46			0.46		0.44
Using hand soap reassures me that I am safe from Norovirus		0.89			0.79		0.5
Handwashing would impact whether or not I get sick from Norovirus		0.61			0.55		0.51
Response-Efficacy Social Distancing (RE.SD)
I think avoiding people who are sick would be one of the best ways to prevent an infection from Norovirus			0.79		0.75		0.76
Avoiding people who are sick would have an impact on whether or not I am infected by Norovirus			0.79		0.75		0.74
Avoiding people who are sick would reduce my chances of a Norovirus infection			0.89		0.69		0.66
Actively avoiding people who appear sick would help keep me free from Norovirus infection			0.56		0.71		0.68
Self-Efficacy Social Distancing (SE.SD)
I would know how to effectively avoid people who are sick				−0.63	0.62		0.62
I would be able to avoid people who are sick when I want to				−0.87	0.76		0.73
I would have no difficulty avoiding people who are sick				−0.78	0.7		0.74
I would be confident in my ability to avoid people who are sick				−0.81	0.71		0.71
Factor Correlations
Factor		SE.HW		RE.HW		RE.SD
SE.HW		1
RE.HW		−0.31		1
RE.SD		−0.45		0.33		1
SE.SD		0.29		−0.31		−0.34
KMO				0.87

* Factor loadings ≤ 0.30 omitted.

).

The overall KMO value indicated excellent sampling adequacy. Communality estimates for the coping appraisal items indicated moderate to high levels of variance accounted for by the extracted components (self-efficacy handwashing: 0.51 to 0.75; response-efficacy handwashing: 0.46 to 0.89; response-efficacy social distancing: 0.69 to 0.75; self-efficacy social distancing: 0.62 to 0.76). These values suggest that the retained coping appraisal items were adequately represented by the extracted components.

Self-efficacy for handwashing demonstrated strong loadings (−0.61 to −0.88), while self-efficacy for social distancing showed similarly strong loadings (−0.63 to −0.87). Response efficacy for social distancing exhibited consistently strong loadings (0.56 to 0.89), while response efficacy for handwashing showed more variable loadings (0.46 to 0.89). The corrected item-total correlations ranged from 0.55 to 0.72 for the self-efficacy handwashing items, from 0.44 to 0.51 for the response-efficacy handwashing items, from 0.66 to 0.76 for the response-efficacy social distancing items and from 0.62 to 0.74 for the self-efficacy social distancing items. Factor correlations among coping appraisal components were generally strong, suggesting that the retained item groupings were related but distinguishable. Because several components were oriented in a negative direction, the signs of the component loadings and correlations should be interpreted cautiously. The broader pattern suggests related, but empirically distinguishable, efficacy-related item groupings. McDonald’s omega values of 0.82, 0.72, 0.86 and 0.86, respectively, indicated acceptable internal consistency for an exploratory study.

4.3. Behavioral Intentions

Separate PCA analyses were conducted for handwashing and social distancing intentions, providing preliminary support for treating these intention items as two distinct behavioral domains. No items were initially removed from either construct because of low factor loadings. Handwashing intention items loaded strongly onto a single component (0.64–0.80) with a KMO value that indicated acceptable sampling adequacy (

Table 3. Principal component analysis: item loadings for handwashing intentions.

Item	HWI	h2	Corrected Item-Total r
I would wash my hands to protect myself from a Norovirus infection	0.69	0.47	0.47
I would wash my hands before eating	0.78	0.61	0.61
I would wash my hands after eating	0.67	0.45	0.48
I would wash my hands after using the restroom	0.80	0.63	0.60
I would wash my hands after opening doors	0.64	0.41	0.45
KMO	0.77

). The corrected item-total correlations ranged from 0.45 to 0.61. McDonald’s omega of 0.74 indicated acceptable internal consistency for an exploratory study.

Social distancing intention items similarly demonstrated strong loadings (0.68–0.79) and a KMO value considered acceptable-to-good (

Table 4. Principal component analysis: item loadings for social distancing intentions.

Item	Factor
	SDI	h2	Corrected Item-Total r
I would intentionally avoid people who are sick to protect myself from a Norovirus infection	0.71	0.51	0.54
I would not sit next to someone who is actively sick in the classroom	0.68	0.47	0.50
I would leave a public restroom if there is someone actively sick in one of the stalls	0.78	0.61	0.62
I would avoid going to a self-service dining hall because it might get me sick with Norovirus	0.79	0.63	0.67
I would order food to my room/home to avoid eating around others in the dining hall	0.78	0.61	0.62
KMO	0.71

). The corrected item-total correlations ranged from 0.50 to 0.67. McDonald’s omega of 0.79 indicated acceptable internal consistency for an exploratory study.

Communality estimates for handwashing intentions ranged from 0.41 to 0.63, while those for social distancing intentions ranged from 0.47 to 0.63, indicating that a moderate proportion of item variance was accounted for by the extracted components. These values suggest that the retained intention items were adequately represented in the exploratory component analyses.

As indicted previously, items removed during analysis were concentrated primarily within the response cost domain, suggesting weaker conceptual coherence relative to other PMT constructs (Supplemental Material Table S5). Across analyses, removed items were disproportionately drawn from this domain, with some efficacy-related items also excluded due to ambiguous cross-loadings. This pattern may indicate that participants differentiated hygiene behaviors more strongly through perceptions of effectiveness and intention than through perceived barriers or costs. Accordingly, item removal was guided not only by statistical criteria but also by theoretical consistency and clarity of interpretation within an exploratory framework.

5. Discussion

The purpose of this exploratory study was to aid in the initial development and examination of preliminary evidence for the internal structure of a theory-based measurement instrument grounded in PMT to assess factors influencing protective health behaviors related to hNoV. Overall, the findings provided preliminary support for a multidimensional structure of protection motivation that includes threat appraisal, coping appraisal, and behavioral intentions among college students. Importantly, these findings should be interpreted as early-stage evidence of dimensionality and item performance rather than as evidence that the scale has been fully validated. From a hygiene and infection prevention perspective, these results underscore the importance of examining the cognitive determinants of specific preventive behaviors, rather than treating hygiene compliance as a single, undifferentiated construct.

Consistent with PMT, the results provided preliminary support for a two-component threat appraisal structure, with perceived severity and perceived susceptibility emerging as distinct but related constructs. The strength of factor loadings and acceptable sampling adequacy suggest that participants were able to meaningfully differentiate between the seriousness of hNoV and their perceived likelihood of infection. Prior research has indicated that both severity and susceptibility independently contribute to protection motivation, particularly in communicable disease contexts where perceived risk may fluctuate over time or across settings [10,18,29]. These findings support the continued use of threat appraisal constructs when evaluating hygiene-related prevention behaviors for enteric pathogens such as hNoV.

The coping appraisal results provided preliminary evidence of item clustering broadly consistent with the theoretical distinction between self-efficacy and response efficacy, with self-efficacy and response efficacy emerging as distinct constructs for both handwashing and social distancing behaviors. Because several components were oriented in a negative direction, the signs of the component loadings and component correlations should be interpreted cautiously. Rather than emphasizing the direction of these associations, the broader pattern suggests that efficacy-related dimensions were related but empirically distinguishable in this exploratory sample. Self-efficacy for both behaviors demonstrated consistently strong factor loadings, reinforcing its central role in determining whether individuals feel capable of engaging in recommended hygiene practices. This finding aligns with both PMT and SCT, which emphasize that confidence in one’s ability to perform a behavior is a key determinant of compliance [17,19,20,21]. In communal environments such as college campuses, where situational barriers and social pressures are common, strong self-efficacy may be particularly important for sustaining preventive hygiene behaviors.

Response efficacy for social distancing demonstrated clear and consistent factor loadings, suggesting that participants perceived avoidance of sick individuals as an effective strategy for reducing infection risk. In contrast, response efficacy for handwashing exhibited comparatively lower loadings, although values remained within acceptable limits. This pattern may reflect variability in beliefs about the effectiveness of handwashing for preventing hNoV infection, particularly given the highly contagious nature of the virus and ongoing public uncertainty regarding the relative effectiveness of handwashing versus other preventive measures [13,15]. From a hygiene promotion standpoint, this finding highlights the importance of reinforcing evidence-based messaging that clearly communicates how proper handwashing reduces hNoV transmission, rather than assuming uniform understanding or confidence in its effectiveness [1,5].

Contrary to theoretical expectations, a distinct response cost component did not emerge. This finding is theoretically important because response cost is a central component of PMT’s coping appraisal process. One possible explanation is that hygiene behaviors such as handwashing and surface cleaning are perceived as low-effort or minimally burdensome among college students. Alternatively, the response cost items included may not have adequately captured the most salient barriers experienced by this population, such as time pressure, environmental constraints, social discomfort, or limited control over shared campus environments. These findings suggest that response cost may operate differently for hygiene behaviors than for other health behaviors traditionally examined within PMT and underscore the importance of context-specific measurement. As a result, the retained instrument should be described as PMT-informed rather than as a complete operationalization of PMT, because one central coping appraisal construct was not retained in the final item structure. Future work should refine and test response cost items in larger and more diverse samples, particularly in settings where barriers to prevention may be more pronounced.

The findings support the decision to conceptualize protection motivation as two distinct behavioral intention constructs, handwashing and social distancing. Both intention measures demonstrated strong unidimensionality and acceptable sampling adequacy, indicating that students distinguish between these behaviors rather than viewing them as interchangeable forms of prevention. This distinction has meaningful implications for hygiene research and practice. Handwashing behaviors may be shaped largely by habit formation and environmental accessibility, while social distancing behaviors may be more strongly influenced by social norms, perceived inconvenience, and situational constraints [10,11]. Treating these behaviors as distinct allows for greater precision in both measurement and intervention design.

Although convergent and discriminant validity are commonly examined in scale validation studies, the present analysis focused on examining the preliminary internal structure of the retained items using PCA with oblique rotation. Evaluating convergent and discriminant validity will require larger samples and appropriate comparison measures, which were beyond the scope of this initial study and represent a critical next stage of validation. Accordingly, the current findings should not be interpreted as evidence of comprehensive construct validity. Instead, they provide preliminary empirical support for subsequent studies that can test convergent, discriminant, criterion-related, and predictive validity using independent samples and external comparison measures. This phased validation approach supports cautious interpretation while identifying priorities for future psychometric testing.

Several limitations should be noted. First, although the sample size was adequate for exploratory analysis, a higher participant-to-item ratio would be preferable to yield more stable component solutions and enhance generalizability. Given the modest sample size relative to the original item pool, the retained component structure should be considered preliminary and in need of replication. At the same time, the stability of the retained components was partially supported by evidence of factor overdetermination, as most components were defined by three or more items with moderate to strong loadings and generally acceptable communality estimates. These patterns suggest that the identified components reflect coherent underlying dimensions; however, confirmation in larger samples is needed to ensure their robustness.

Second, the use of PCA, while appropriate for exploratory data reduction, does not allow for inference about latent constructs in the same manner as common factor analysis or confirmatory methods. Additionally, as noted previously, convergent and discriminant validity were not assessed at this stage, limiting psychometric conclusions to internal structure and item performance alone. Third, the self-reported and cross-sectional nature of the data may introduce response bias and preclude causal inference. Finally, the relative homogeneity of the sample, particularly the predominance of White undergraduate students from a single institution, limits the generalizability of the findings to more diverse populations. Future research should further evaluate and validate the instrument across broader samples, including different age groups, geographic regions, and higher-risk contexts such as healthcare and foodservice settings. Collectively, these limitations reflect the study’s exploratory intent and help to inform directions for subsequent validation work rather than detracting from the utility of the present findings.

From a practical standpoint, the retained item set provides a useful starting point for public health practitioners, campus health professionals, and researchers seeking to assess hygiene-related cognitions associated with human norovirus (hNoV) prevention behaviors. Although the term human norovirus was not explicitly used in item wording, the exclusive use of human respondents ensured that interpretation remained unambiguous and contextually appropriate. Researchers interested in applying this measure are encouraged to use the refined set of retained items (Supplemental Material Table S6), rather than the full pool of originally developed items. By identifying whether gaps in prevention compliance are driven by low perceived risk, limited self-efficacy, or uncertainty regarding behavioral effectiveness, interventions can be more strategically designed to address specific cognitive barriers. Such targeted hygiene promotion efforts may be especially valuable in communal settings, where outbreak prevention depends on the timely and widespread adoption of multiple preventive behaviors [3,5,16].

6. Conclusions

Despite these limitations, this study contributes an empirically grounded, theoretically informed starting point for measuring hygiene-related cognitive determinants using PMT and SCT. By identifying which constructs demonstrated coherence and which required reconsideration, the findings provide a practical roadmap for future scale refinement and validation. Improved measurement precision in this domain has the potential to strengthen intervention design by enabling more targeted alignment between cognitive determinants and behavioral strategies for hNoV prevention.

The findings revealed that preventive behaviors commonly grouped under the umbrella of hygiene are psychologically distinct and influenced by different motivational factors. Recognizing these distinctions is essential for designing effective, behavior-specific hygiene promotion strategies, particularly in communal environments such as college campuses where hNoV transmission risk is high and pharmaceutical prevention options remain limited.