Application of construal level theory in identifying factors affecting individual decision-making in implementing flood protection measures in rural areas of Iran

Abstract

Floods are among the most frequent and destructive natural disasters globally, inflicting extensive damage on infrastructure, causing significant economic losses, displacing communities, and triggering public health crises. Developing countries are particularly vulnerable due to inadequate infrastructure, substandard housing, widespread poverty, and low adaptive capacity—challenges that are especially pronounced in rural areas. In the absence of effective mitigation systems, the impact of floods on rural communities can intensify, potentially resulting in the complete collapse of local livelihoods. Therefore, enhancing the resilience of these communities through targeted capacity-building and resilience strategies is essential. This study aims to identify the key factors influencing the adoption of flood preparedness measures in rural regions of Iran. The primary research instrument was a questionnaire, and the data were analyzed using Structural Equation Modeling (SEM). This study employed Construal Level Theory (CLT) as its theoretical framework. The findings revealed that CLT components—including geographical distance, temporal distance, social distance, hypothetical distance, belief in climate change, and flood risk perception—significantly influence the adoption of protective measures prior to flood events, collectively explaining 67.1% of the variance. Notably, this research marks the first application of CLT in the context of flood preparedness, both in Iran and globally. As such, it offers novel insights for the development of effective flood management strategies in rural communities and lays a foundation for strengthening the resilience of their livelihoods.

Introduction

Global warming has intensified the global water cycle^1,2,3,4,5 resulting in a heightened frequency of extreme weather events^6,7,8. This intensification presents significant threats to socio-economic stability, environmental integrity, and water resource security^9,10,11. As global temperatures rise and glaciers continue to melt, sea levels increase, and precipitation patterns become more unpredictable, escalating the risk of flooding¹⁰. Flooding typically occurs during periods of heavy rainfall when rivers and streams exceed their capacity to manage excess water¹².

Floods rank among the most frequent and devastating natural disasters worldwide, causing widespread damage to infrastructure, substantial economic losses, community displacement, and public health crises^13,14,15. Globally, they remain one of the most costly disasters, impacting economies, societies, and populations^16,17,18,19. Flooding has become an increasingly frequent global event with substantial socio-economic consequences²⁰. Over the past four decades, floods have caused more than one trillion dollars in global damages, with annual losses estimated at 40 billion dollars²¹. Between 1998 and 2017, floods affected more than 2 billion people worldwide²². In 2023, the Centre for Research on the Epidemiology of Disasters (CRED) recorded approximately 165 flood events globally, resulting in over 7,700 deaths and impacting more than 32 million individuals²³. The World Water Resources Institute recently reported that by 2030, the number of people affected by floods is expected to double²⁴. Over the past five decades, floods have accounted for more than half of all disaster-related losses, impacting millions of people worldwide annually²⁵. Estimates suggest that the number of individuals affected by floods is comparable to those impacted by other natural disasters²⁶.

Asia is highly prone to natural disasters and has endured numerous devastating events. In 2015, more than half (62.7%) of all global natural disasters occurred in the region²⁷. Many Asian cities, particularly those in low-lying coastal areas, face significant vulnerability to flooding²⁸. Moreover, due to development pressures and rising population growth, these cities face an increased risk of catastrophic losses and disasters such as floods²⁹. In Iran, financial losses from floods have escalated in recent decades, with rural residents being the most affected²². Iran ranks as the sixth most disaster-prone country globally. Natural disasters in the country predominantly affect rural areas, which are more vulnerable due to inadequate infrastructure and facilities³⁰. Flood statistics in Iran indicate a rising trend in recent decades, leading to significant economic and social consequences for residents³¹. According to the latest Iranian Crisis Management Report, in 2019, floods and river overflows affected more than 25 provinces, including 200 cities and 4,304 villages, resulting in damage to over 60,000 urban and rural residential units. In rural areas, flood damage led to the complete evacuation of 350 villages across Iran³¹. Flood events pose significant risks to rural communities, the majority of which depend heavily on rainfed agriculture³². Extensive physical and socio-economic damages profoundly affect individuals’ psychological well-being. Floods not only cause psychological distress but also expose the social vulnerability of victims by resulting in financial and human losses or displacement³³. Floods significantly impact the agricultural sector, leading to reduced yields, increased production costs, and high investment demands. Escalating operational expenses, combined with inadequate infrastructure, diminish rural households’ ability to manage these risks, making them more vulnerable to flood-induced poverty³⁰. Consequently, floods pose a serious threat to rural areas in developing countries, including Iran^15,34.

In response to these challenges, local communities can develop strategies and infrastructure for protection, recovery, and adaptation to mitigate future hazards³⁵. Global disaster risk reduction policies emphasize implementing protective measures before floods occur to enhance resilience and preparedness³⁶. Preparedness encompasses both pre-flood actions aimed at minimizing impacts and enhancing coping mechanisms, as well as post-flood resilience-building efforts³⁷. Research indicates that individuals with higher levels of preparedness respond more effectively to floods, thereby strengthening community resilience and mitigating their consequences³⁸. Given the high variability and unpredictability of floods, enhancing residents’ capacity to withstand and adapt to disasters is essential³⁹. Private flood mitigation measures encompass a range of protective strategies that individuals can implement within and around their properties to minimize or prevent flood damage⁴⁰. Household-level flood mitigation seeks to reduce the severity and impact of floods through measures that prevent floodwater from entering homes (e.g., watertight doors) and strategies that minimize damage once floodwaters infiltrate (e.g., waterproof drywall)⁴¹. Flood protection behaviors or measures are commonly categorized into three main types: structural, nonstructural, and information-based adaptations. Structural measures refer to physical interventions aimed at reducing flood impact, such as building insulation, levees, dams, and elevation of structures. Nonstructural measures include policy-based and financial tools like flood insurance, land-use regulations, and relocation incentives. Information-based adaptations involve behavioral and technological strategies, such as participation in early warning systems, access to flood risk maps, and community preparedness programs⁴². Collectively, these approaches form the basis of pre-flood protection measures, which encompass a range of proactive actions and programs designed to minimize loss of life and property before flood events occur. These measures are essential components of integrated flood risk management and contribute significantly to enhancing community resilience⁴¹.

Flood mitigation behavior is an emerging research area aimed at understanding why some flood-prone residents take proactive measures to reduce risks while others do not⁴³. Motivational factors vary depending on the context and the specific mitigation measures being considered⁴⁴. Additionally, personal experiences with past hazards, financial losses, and physical or psychological harm play a significant role in shaping conservation motivation and decision-making⁴⁵. Several theoretical frameworks have been employed to model flood risk reduction behavior and identify the motivational factors influencing it⁴⁶. However, these frameworks have yet to adequately explain why human societies fail to implement protective measures before a flood occurs, leaving this critical issue largely unaddressed⁴⁷. Accordingly, this study aims to examine the factors influencing the implementation of pre-flood protective measures among local communities in Iran.

Literature review

Flood management studies can generally be categorized into two principal domains. The first domain focuses on environmental and morphological determinants, examining the physical and natural variables that influence flood dynamics. These include vegetation cover, riverbed slope⁴⁸, flood probability and return intervals⁴⁹, spatiotemporal patterns of flood risk distribution⁵⁰ and a range of hydraulic parameters such as Manning’s coefficient, surface roughness, modeling time step, turbulence coefficient, and downstream boundary conditions⁵¹. Complementary studies also incorporate geological and geomorphological indicators, such as elevation, slope, curvature, topographic wetness index (TWI), normalized difference vegetation index (NDVI), land use and land cover (LULC), rainfall data, lithological composition, and proximity to infrastructure like roads and rivers to develop flood susceptibility maps⁵². Although these approaches offer essential insights for predictive modeling and technical planning, they often exhibit a critical shortcoming: the underrepresentation of human and social dimensions in flood risk management. These dimensions—comprising behavioral responses, psychological perceptions, community awareness, and adaptive decision-making—play a pivotal role in determining actual outcomes during flood events. Ignoring these human-centric factors can undermine the effectiveness of even the most sophisticated physical flood models and limit their utility in building resilient communities.

The second group has examined social and human factors and has used various analytical approaches to understand and assess the behavioral, institutional, and managerial dimensions associated with floods. However, these approaches also often have various limitations. Various theoretical frameworks have been employed to investigate the factors influencing the adoption of protective measures prior to flooding events. Among the most prominent are the Protection Motivation Theory (PMT)⁴⁷ and the Theory of Planned Behavior (TPB)⁵³. PMT emphasizes two key components: threat appraisal—which includes perceived severity and vulnerability—and coping appraisal, encompassing response efficacy and self-efficacy. Research indicates that fear, perceived vulnerability, and protection motivation are among the strongest predictors of preventive behavior. This theory is particularly effective when individuals perceive the threat as serious and believe they are capable of self-protection^54,55. TPB, on the other hand, focuses on three constructs: attitude toward the behavior, subjective norms, and perceived behavioral control. Within this framework, both behavioral intention and prior experience with disasters serve as strong predictors of preparedness actions. The TPB framework proves most effective when individuals are aware of the risks and possess the necessary resources to take appropriate action⁵³. In essence, both PMT and TPB are grounded in rational evaluation and behavioral intention, offering valuable insights into the psychological mechanisms that drive flood preparedness^53,56.

However, research suggests that psychological distance representations can promote more confident future-oriented judgments, as well as clearer evaluations and behavioral decisions⁵⁷—factors not explicitly addressed in either PMT or TPB. For instance, a study conducted in China demonstrated that psychological distance significantly influences preparedness behavior, even among residents of flood-prone regions. These findings indicate that past experience alone may be insufficient for fostering preparedness; rather, how individuals perceive risk is a critical determinant⁵⁸. Moreover, repeated exposure to flooding can sometimes lead to a form of indifference or normalization. In such cases, individuals may continue to perceive temporal or spatial distance as substantial, dampening their motivation to act. This underscores the potential value of theoretical approaches that explicitly incorporate psychological distance into their framework, which may prove more applicable to flood preparedness contexts than models like PMT and TPB⁵⁹. Research indicates that when climate phenomena such as floods are perceived as psychologically distant, individuals are less likely to engage in adaptive or coping behaviors^59,60. In contrast, when these events are perceived as psychologically close, individuals demonstrate greater willingness and motivation to take preventive action^61,62. Psychological distance significantly influences people’s attitudes, motivations, and behavioral responses⁶¹. Accordingly, when people perceive phenomena such as floods as proximate and personally relevant, they are more likely to acknowledge the threat and take proactive measures⁶³. Thus, reducing the psychological distance associated with climate change can serve as a strategic approach to enhancing motivation for action⁶¹. In this context, Spence et al.⁶⁴ proposed that Construal Level Theory (CLT) has the potential to clarify how psychological distance interacts with individuals’ perceptions of climate phenomena—insights that can be leveraged to encourage adaptive and coping responses. In rural areas of Iran, for example, the perception that floods only occur in distant regions creates a psychological barrier to preparedness. CLT offers a valuable framework for addressing this by promoting mental proximity to natural hazards⁶⁵. CLT is a social psychological framework that examines the relationship between psychological distance and the level of abstractness or objectivity in individuals’ perceptions of events and objects⁶⁶. According to CLT, psychological distance influences whether an object or concept is perceived as familiar and tangible or, conversely, abstract and distant⁶⁷. In essence, CLT explains how psychological distance shapes thought processes and behavior⁶⁶. CLT posits that distant psychological events are represented by high-level abstract concepts that encompass decontextualized general features, whereas close psychological events are characterized by specific low-level concepts that incorporate detailed contextual elements^65,66. According to CLT, when individuals perceive climate change as psychologically distant, its threats and risks appear less real, tangible, or relevant⁶². This perception serves as a significant barrier to effectively coping with climate phenomena^68,69,70.

To evaluate the relative effectiveness of CLT compared to the TPB and PMT in the context of flood prevention measures, it is essential to highlight the foundational differences in how each theory conceptualizes individual decision-making. These distinctions, which underscore each model’s assumptions and predictive mechanisms, are systematically presented in Table 1.

Table 1 Comparison of different theories in the field of flood protection measures.

In general, this study focuses on examining protective measures undertaken prior to the occurrence of flooding, with particular emphasis on the influential role of psychological distance in shaping individual behavior. Unlike PMT, which is predominantly applied in the context of immediate and severe threats, CLT offers a valuable framework for analyzing human behavior under non-crisis conditions and during the early stages of a potential hazard. As such, CLT was selected as the theoretical foundation for this research. Given the study’s aim of identifying the factors that influence flood-preventive actions in high-risk areas, the selection of a suitable theoretical framework is critical. CLT, rooted in social psychological theory, provides robust conceptual tools for examining individual decision-making processes when threats are perceived as distant—temporally, spatially, socially, or hypothetically. Psychological distance, a core construct of CLT, is instrumental in shaping risk perception and the likelihood of engaging in protective behaviors. CLT enables researchers to classify and interpret individuals’ responses to future threats based on the level of abstraction through which these threats are construed. Compared to TPB and PMT, CLT possesses distinct advantages. Specifically, CLT can analyze preventive actions in the absence of acute danger, whereas PMT is more suited to imminent threats, and TPB centers on behavioral intentions influenced by normative beliefs. Furthermore, CLT supports the design of targeted behavioral interventions by aligning message framing and policy communication with individuals’ construal levels. Accordingly, this study adopts CLT as its theoretical framework to investigate and interpret preventive behavior in the context of flood risk.

Research design

Theoretical framework

CLT has been effective in predicting individual perceptions and behaviors across various domains, including pro-environmental actions⁷¹, green product purchase intentions (Ibrahim et al., 2018), climate change mitigation and adaptation strategies^65,72 consumer behavior (Ibrahim et al., 2018), and climate change awareness⁶². However, its application in the context of flooding remains unexplored.

CLT illustrates the relationship between psychological distance and human behavior. Psychological distance is commonly associated with four dimensions: geographical distance (GD), temporal distance (TD), social distance (SD), and hypothetical distance (HD); however, these distinctions are categorized systematically⁶⁶. Individuals who perceive climate change as psychologically distant may believe it primarily affects remote locations and populations, reinforcing geographical distance (GD). Similarly, they may assume that while climate change is occurring, its consequences will be more severe for future generations, reflecting temporal distance (TD)⁶⁴. The uncertainty surrounding climate change can contribute to inaction, making the issue appear more distant and less threatening. Additionally, the global nature of climate change may reinforce perceptions of its remoteness and uncertainty, reflecting hypothetical distance (HD). Conversely, even if individuals acknowledge the reality, imminence, and proximity of climate change impacts, they may still experience personal detachment, perceiving these effects as socially distant (SD)²⁹. Researchers often assume that individuals generally perceive climate phenomena, such as flooding, as psychologically distant across all four dimensions⁷¹. Correlational studies have consistently demonstrated a negative relationship between psychological distance and protective behavior, suggesting that greater perceived distance reduces the likelihood of adopting preventive measures^64,65,72,73.

CLT has demonstrated that psychological distance serves as a framework for understanding beliefs related to risk perception, willingness to act, and risk responses^60,65. Risk perception involves individuals’ subjective assessment of a phenomenon’s likelihood and severity. Awareness plays a crucial role in shaping perception and motivating preventive actions⁷⁴. Risk perception is not simply a cognitive process; it also involves emotional responses that influence decision-making⁷⁵. Community perception of floods refers to how communities perceive, behave, and act based on their experience and the socio-demographic conditions prevailing in a particular situation⁷⁶. Flood risk perception is a subjective assessment of both the likelihood and severity of impending floods, as well as individuals’ perceived ability to withstand their impacts⁷⁷. Understanding the factors that drive people to respond to flood hazards is critical for effective risk management, as risk perception strongly influences reactions to warnings⁷⁸. For example, during floods, individuals may not only evaluate evacuation routes but also experience fear or anxiety regarding rising water levels⁷⁹.

Belief in climate change is another key variable within CLT⁶⁵. Belief is defined as the conviction held at a given moment that something is true or correct⁸⁰. Climate change beliefs refer to the degree to which an individual recognizes a phenomenon as being influenced by or connected to climate change^80,81. Le Dang et al.⁸¹ found that farmers who believed in the reality of climate change exhibited higher levels of perceived risk⁶⁵. Li et al.⁸² suggested that beliefs influence both risk perception and adaptive behaviors. Consequently, belief in climate change is positively associated with adaptation and promotes adaptive responses⁸³. Beliefs play a crucial role in shaping attitudes toward established issues, influencing behavioral actions and decision-making⁸⁴. Moreover, evidence suggests that individuals’ beliefs about climate change affect how they interpret related events⁸⁵. Accordingly, based on the presented material, the theoretical framework for this research can be structured using CLT (Fig. 1), with the corresponding research hypotheses formulated within this framework.

• Geographical distance (H1), Temporal distance (H2), Social distance (H3) and Hypothetical distance (H4) have negative direct effects on Belief in climate change.

• Geographical distance (H5), Temporal distance (H6), Social distance (H7) and Hypothetical distance (H8) have negative direct effects on risk perception.

• Geographical distance (H9), Temporal distance (H10), Social distance (H11) and Hypothetical distance (H12) have negative direct effects on protective measures before floods occur.

• Beliefs that climate change (H13) have positive a direct effect on protective measures before floods occur.

• Risk perception (H14) have positive a direct effect on protective measures before floods occur.

Control variables

Although the primary focus of this study was on the core components of CLT, socio-economic characteristics were also incorporated as contextual variables, given their influence on the implementation of pre-flood measures and the preparedness of local communities³⁶. While climatic and environmental factors largely determine the frequency and severity of floods, their actual impact on residents often depends on anthropogenic elements³². Accordingly, demographic and socio-economic variables were included as control variables in the research model. Prior studies have demonstrated that factors such as age, flood experience, income, household size, and education significantly influence the adoption of pre-flood measures^44,86,87,88. Thus, these variables were treated as mediators in this study, with the potential to shape the relationships among key constructs.

Fig. 1

Theoretical framework of research based on CLT.

Survey instrument

To design the questionnaire items, a comprehensive review of theoretical research literature was conducted across various fields, including perception of risk, belief in climate change, flood management, conservation methods, and social capital. Additionally, existing questionnaires from previous studies were examined and incorporated to ensure the accuracy and relevance of the measurement items in the present study. Based on this review, relevant items were designed and adapted to align with the research objectives. Specifically, 7 items were used to assess conservation behaviors, 3 items for HD, 4 items for SD, 4 items for TD, 5 items for GD, 4 items for belief in climate change, and 4 items for flood risk perception (Table 2).

Table 2 Component measurement items.

Data analysis

Data analysis in this study was conducted using SPSS and SmartPLS software, addressing both descriptive and inferential statistics. To examine relationships between latent and manifest variables, Structural Equation Modeling (SEM) was employed to validate the proposed conceptual framework. Although the sample size technically warranted the use of covariance-based SEM, the study opted for variance-based SEM using SmartPLS due to three critical considerations: (i) Model complexity: The research framework included both reflective and formative constructs, as well as intricate indirect relationships. SmartPLS is particularly well-suited for such configurations. (ii) Data non-normality: Covariance-based techniques are sensitive to data normality assumptions. Given the dispersion and non-normal distribution of variables in this study, variance-based SEM provided a more robust analytical approach. (iii) Predictive orientation: The primary aim of this research was to predict behavioral outcomes related to flood protection decisions, rather than merely explain correlations. SmartPLS offers enhanced capabilities for predictive modeling and handling mediating effects. Therefore, SmartPLS was selected not only for its flexibility and tolerance of data irregularities but also for its superior performance in exploring both direct and indirect influences among key variables.

The SEM approach employed in this study comprises two principal stages: (1) the measurement model, which assesses the relationships between latent constructs and their observed indicators, ensuring psychometric robustness through unidimensionality, construct validity, reliability, and discriminant validity⁹³ and (2) the structural model, which analyzes the causal paths among latent variables to validate theoretical relationships⁹⁴. In this context, observed variables refer to directly measurable indicators, while manifest variables are questionnaire items explicitly linked to these indicators, as detailed in Table 1. By contrast, latent variables represent abstract, unobservable constructs—such as perceptions or beliefs—which are inferred statistically from observable measures⁹³. The latent variables evaluated in this study include Protective Measures, Geographical Distance, Temporal Distance, Hypothetical Distance, Social Distance, Risk Perception, and Belief in Climate Change.

To control for confounding influences, the model incorporated several control variables, held constant to isolate and clarify the relationships between independent and dependent constructs⁹⁵. These variables—namely age, flood experience, household income, education level, and household size—were selected based on their relevance to behavioral decision-making in flood-prone settings.

Furthermore, to assess potential multicollinearity, Variance Inflation Factor (VIF) values were calculated for each indicator. A conservative threshold of < 5 was applied, in alignment with statistical best practices, ensuring minimal redundancy and independence among constructs⁹⁶.

Materials and methods

Study area

Shushtar County, located in Khuzestan Province, has been analyzed using Geographic Information Systems (GIS) software (ArcGIS 10.5) (https://soft98.ir/software/engineering/3526-arcgis.html) (Fig. 2). As a key agricultural hub, it ranks second in the province for agricultural production. However, recurring floods pose a significant threat to this vital sector. In flood years, over 1,000 hectares of farmland become unusable, severely impacting local output. In recent years, Shushtar has experienced several major flood events, during which numerous villages have been inundated or encircled by floodwaters. Due to its geographical location, Shushtar has been more affected by flooding than other cities in Khuzestan Province³⁶. For instance, flood damage reports from 2019 indicate that in the Shoaibieh district—one of Shushtar’s districts—over 36 villages were submerged, suffering extensive devastation⁹⁷.

Fig. 2

Case study.

Statistical population and sampling method

The statistical population of this study comprised all rural households exposed to flooding in Shushtar County. To estimate the sample size, Kerjcie & Morgan’s⁹⁸ table was utilized, resulting in the selection of 353 rural household heads through a multi-stage stratified sampling method. This advanced sampling approach is typically employed when studying an entire population is impractical due to its size or complexity. It involves dividing the population into clusters (groups) and subsequently selecting samples across multiple stages, with each stage narrowing down the subset of clusters or individuals⁹⁹. This method integrates stratified sampling with multi-stage sampling, enabling researchers to obtain a more accurate and representative sample of the target population. An additional advantage is that, rather than examining the entire population, only a limited number of clusters and individuals are selected, improving efficiency¹⁰⁰. To ensure precise sample selection, the stratified sampling process was implemented following methodologies from previous studies¹⁰⁰ as outlined below.

(i) Division of the population into classes: The population is first segmented into distinct groups (clusters) based on criteria such as geographical location, organizational attributes, or other relevant factors. In this study, Shushtar County was classified into three districts according to the Statistical Center of Iran: Central, Shoaibieh, and Mian-Ab. (ii) Selection of clusters: One or more clusters are randomly chosen for analysis. In this study, two rural districts were randomly selected from each of the three districts for inclusion in the research sample. (iii) Cluster subset selection: A subset of individuals within the selected clusters is then chosen. In this study, two villages from each district were selected for analysis. (iv) Data collection selection: Finally, data is gathered from the selected sample. Villages were randomly chosen for the study, resulting in a total of 12 villages included in the research.

Validity

Questionnaire validity refers to the degree to which a questionnaire accurately measures the intended concept and effectively fulfills the research objective¹⁰¹. In this study, two methods were employed to assess validity. (1) Content Validity: Content validity assesses the extent to which a questionnaire encompasses all relevant aspects and dimensions of the concepts under study. This validity is crucial as it ensures that the questionnaire aligns with the research objectives and accurately measures the intended constructs¹⁰². Content validity is typically evaluated by subject matter experts to ensure that questionnaire items are comprehensive, relevant to the topic, and that no critical aspects of the concept under study are omitted¹⁰¹. Following the development of the questionnaire items—formulated based on a comprehensive review of prior studies—the initial draft was submitted to a panel of experts representing diverse specializations, including water resources, rural extension and development, social psychology, and environmental science. Based on their feedback and recommendations, a series of revisions were implemented to refine the instrument. This iterative process continued until the questionnaire received final approval.

(2) Construct Validity: Construct validity assesses the extent to which a measurement instrument, such as a questionnaire, accurately captures the theoretical concept or construct it is designed to measure. This type of validity ensures that the questionnaire items are appropriately aligned with the underlying theory or concept being examined¹⁰³. One method for assessing construct validity is the average variance extracted (AVE), where a value exceeding 0.5 is considered statistically acceptable¹⁰⁴. Based on the results presented in Table 1, the questionnaire demonstrates adequate validity.

Reliability

Reliability refers to the capacity of a measurement instrument or questionnaire to consistently produce stable and accurate results over time or across varying conditions. In other words, it reflects the degree to which the instrument yields similar outcomes when applied multiple times under identical circumstances¹⁰⁵. In this study, two methods were employed to assess reliability. (1) Cronbach’s alpha: Cronbach’s alpha is a statistical index used to evaluate the internal consistency of questionnaires or measurement instruments. It indicates how well the items within an instrument function together and the extent to which all items contribute to measuring a specific construct. This method provides an estimate of the correlation between the variables that comprise the construct or scale under investigation⁹⁵. A Cronbach’s alpha value above 0.7 is considered statistically acceptable. After confirming the validity of the questionnaire, the next step involved calculating the reliability of the research items. To ensure validity, the initial draft of the questionnaire was distributed to 20 individuals from the statistical population prior to the main data collection. Subsequently, Cronbach’s alpha was calculated to assess reliability. Items with low reliability were carefully reviewed and revised as necessary. This process continued until the questionnaire achieved final approval.

(2) Composite Reliability (CR): CR is a statistical measure used to evaluate the internal consistency of measurement instruments, such as questionnaires. It assesses the degree to which items within an instrument function cohesively and accurately measure the intended construct. Unlike Cronbach’s Alpha, which assumes equal importance for all items, composite reliability assigns greater weight to items with higher factor loadings, making it a more precise measure of reliability¹⁰⁶. A CR value exceeding 0.6 is considered statistically acceptable⁹⁶. Based on the results presented in Table 1, the research instrument demonstrates sufficient reliability.

Ethical statement

Informed consent was obtained from all individuals who participated in the study. All materials and methods were carried out according to instructions and regulations, and this research has been approved by a committee at Agricultural Sciences and Natural Resources University of Khuzestan, Mollasani, Iran. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Results

Personal and professional characteristics of respondents

The mean age of respondents was 45.92 years, with a standard deviation of 11.44 years, placing them within an age range of 23 to 73 years. The gender distribution revealed that 90% of respondents were male. Regarding education, the majority (36%) had completed secondary education, while 5.6% were illiterate. The average household size was 5.58 individuals, and the annual agricultural income averaged 484.07 million Tomans. Notably, 94% of respondents had experienced at least two floods. Additionally, the results indicated that 65.1% of respondents had never attended flood prevention training classes.

Measurement model

To verify the measurement model, researchers must follow key steps to ensure that the manifest variables accurately reflect the latent variables. Once these steps are confirmed, the structural model of the research can be examined and validated.

Unidimensionality: This criterion determines whether the observed variables effectively measure the same latent variable^93,95. At this stage, factor loading values and their significance levels are assessed. If the factor loading exceeds 0.5 and the t-value is greater than 1.96, the variable can be considered valid in terms of unidimensionality⁹⁵. The results confirm that the research variables exhibit high accuracy and have been appropriately selected (Table 3).

Validity and reliability: Reliability and Validity: At this stage, the values for AVE, Cronbach’s alpha, and CR should be assessed⁹⁶. Based on the results, the research tool—specifically, the questionnaire—demonstrates appropriate validity and reliability, as the reported AVE, Cronbach’s alpha, and CR values exceed the established standards.

Furthermore, to ensure the independence of the constructs within the model, multicollinearity was assessed through the calculation of VIF values for all indicators. The results revealed that all VIF values remained below the recommended threshold of 5, indicating the absence of significant multicollinearity among the constructs. This outcome affirms that the indicators are free from redundancy and that each construct uniquely and reliably represents distinct dimensions of the phenomena under investigation.

Table 3 Assessment of measurement model.

Discriminant validity: Discriminant validity assesses the uniqueness of a measurement tool. A test exhibits discriminant validity when its correlation with instruments measuring different constructs is low or nonexistent¹⁰⁴. The evaluation criterion at this stage is based on the square root of the extracted variance and the correlation between constructs^104,107. According to the results presented in Table 2, the square root of the extracted variance (0.730 < AVE < 0.849) exceeds the correlation values (-0.384 < r< -0.569), confirming the discriminant validity of the research instrument (Table 4).

Table 4 Discriminant validity of research constructs.

Heterotrait-Monotrait ratio (HTMT): HTMT is a technique used to assess the divergent validity of measurement models by determining whether there is sufficient differentiation between constructs. It has been proposed as a more accurate alternative to the classic Fornell-Larcker method⁹³. This approach calculates the ratio of the average correlations between different constructs (Heterotrait) to the average correlations within a single construct (Monotrait). If the HTMT value falls below the accepted threshold (typically < 0.85), the constructs can be considered to exhibit adequate divergent validity. The findings of this study confirm that the research instruments meet this criterion (Table 5).

Table 5 HTMT value for research constructs.

Evaluation of the research structural model

Following the validation of the measurement model, the structural model must be reviewed and assessed. To ensure accuracy, indicators proposed in previous studies Hair et al.⁹³. were applied. The results confirmed a strong model fit, as presented in Table 6.

Table 6 Structural model fit index.

Following the confirmation of the structural model, the next step involves testing the research hypotheses. The research path model is presented in two formats: factor loadings and t-values (Figs. 3 and 4). The evaluation results indicate that the research variables collectively explain 67.1% of the variance in the implementation of protective measures prior to flood occurrence.

The findings of the study reveal that the various dimensions of psychological distance—GD, TD, SD, and HD—not only exert a direct influence on individuals’ engagement in flood-related protective measures, but also function through indirect pathways mediated by two key constructs: belief in the reality of climate change and risk perception.

Path analysis results demonstrate that psychological distance dimensions account for 32.1% of the variance in individuals’ belief in climate change, and 30.5% of the variance in risk perception. These values underscore the central mediating role of these cognitive factors in the causal framework. In practical terms, the results suggest that efforts to reduce psychological distance—for example, by localizing risk communication or emphasizing present and personal implications of climate hazards—can substantially enhance individuals’ belief in climate threats and elevate their perceived vulnerability. Such shifts in cognition are instrumental in fostering adaptive behaviors, including preemptive protective measures. Therefore, integrating psychological and perceptual strategies into flood risk communication is essential for designing more effective and community-sensitive resilience policies.

Fig. 3

Model in standard mode (factor loading).

Note: It is important to note that in order to facilitate understanding of the model results and reduce the visual complexity of the graphs, the final results related to the control variables are presented in Table (6). This table provides a concise, coherent, and understandable view of the relationships between the research variables and allows for greater focus on meaningful findings.

Fig. 4

Model in significant mode (t-value).

Hypothesis testing: To evaluate the research hypotheses, the bootstrapping method was applied in the PLS analysis, using two scenarios with 3,000 and 4,000 hypothetical samples. The bootstrap method relies on two key criteria: path coefficients (β values) and the t-value statistic, as outlined by Hair et al.¹⁰⁸. The results indicated that, in both scenarios, the significance of the coefficients remained unchanged, with only the t-values varying due to the influence of sample size. These findings confirmed that the conceptual framework designed for this study possesses high validity, and all research hypotheses and relationships were statistically significant (Table 7).

Table 7 Results of examining research hypotheses.

Discussion

This study aimed to identify the factors influencing the implementation of protective measures before a flood occurs. As one of the first research efforts in flood management within Iran, it explores an under examined area. Based on the authors’ knowledge and a review of the theoretical literature, no prior study has investigated pre-flood protective measures using the CLT framework. Thus, this research not only helps bridge a gap in the theoretical literature but also provides valuable insights for policymakers and flood-affected countries seeking to enhance disaster preparedness. The study results demonstrated the high effectiveness of the applied theory, as it successfully explained 67.1% of the variance in the use of protective measures. Additionally, all relationships between variables were statistically significant, aligning with theoretical predictions and supported by the research data. SEM was employed to analyze the research hypotheses, which were systematically examined. Next, the influence of each variable examined in this study on individuals’ flood protection behaviors is analyzed. This analysis is conducted independently for each variable to discern its unique contribution and to evaluate its impact on decision-making regarding preventive measures. By adopting this approach, key determinants can be identified, intervention priorities can be established, and a scientific foundation can be laid for the development of educational initiatives and policymaking in the field of flood risk management.

Control variables

The findings of this study demonstrate that formal and targeted education—delivered through training courses, workshops, and organized information programs—plays a crucial role in fostering protective behaviors¹⁰⁹. As a structured process, education enhances practical skills, reshapes attitudes, and strengthens individuals’ psychological readiness in confronting natural hazards. In contrast, knowledge, whether acquired through education or personal experience, encompasses conceptual understanding, risk awareness, and subjective information processing. Education serves as a catalytic force that generates meaningful knowledge, which is instrumental in promoting effective protective actions¹¹⁰. In other words, education facilitates the creation and reinforcement of knowledge, ultimately increasing individual motivation and capacity to engage in protective measures. Therefore, distinguishing between education—the transmission of information and skills—and knowledge—the cognitive outcome of that process—is essential for a more accurate assessment of behavioral models in response to flood risks.

Numerous studies have demonstrated that prior experience with flooding enhances risk perception and preparedness. Households affected by floods are more inclined to modify their behavior and support improved flood management strategies^111,112. Evidence suggests that individuals who have experienced flood events tend to recalibrate their risk perceptions¹¹³ and are more motivated to engage in preparedness actions¹¹⁴. Personal encounters with local weather extremes and climate-related events often render climate change more tangible, immediate, and personally relevant¹¹⁵. Therefore, it is essential to document past flood experiences accurately, systematically, and consistently, capturing details on the extent of damage, community-level responses, the effectiveness of policy enforcement, and encountered challenges. When such documentation is well-organized and made accessible to local populations—through public archives, local institutions, or digital platforms—it serves as a powerful tool for enhancing public awareness, facilitating social learning, and encouraging proactive community engagement in flood prevention measures. Moreover, leveraging documented experiences fosters greater trust in management strategies and contributes to strengthening social resilience, enabling communities to better withstand and recover from future flood events.

GD

The study results indicate that GD has a significant negative impact on beliefs, risk perception, and protective responses before a flood, confirming hypotheses H1, H5, and H9. These findings align with previous research conducted by^64,65,72,73. This finding suggests that individuals may acknowledge the occurrence of climate change and the likelihood of flooding but still perceive its impacts as psychologically distant, primarily affecting other locations. Additionally, many people recognize flooding as a geographically close threat yet believe its consequences will be more severe in distant regions. For instance, while warmer summers—one of the more benign aspects of climate change—may be perceived as psychologically close, the risk of severe impacts such as flooding is often viewed as relatively distant⁶⁴. Additionally, people tend to perceive future environmental conditions as more severe in geographically distant areas⁶⁸. Similarly, individuals exhibit a moderate tendency to view the impacts of climate change in developing countries as more critical than in their own local region¹¹⁶ and to consider environmental problems more serious at a global level rather than a local one¹¹⁷. This asymmetry may stem from a motivational inclination to avoid threatening information¹¹⁸. When confronted with evidence of a serious threat such as climate change, individuals may be inclined to perceive its impacts as geographically distant, even without fully understanding the scientific basis of such predictions. Thus, reducing the psychological distance of climate change could serve as a strategy to enhance individuals’ motivation to take action⁶¹. In other words, making climate change more tangible, more local, more relevant, and more immediate may help reduce feelings of alienation and encourage proactive responses⁶⁴. This suggests that people acknowledge the significant impact of climate change events, such as floods, and recognize the severe consequences these events may have on their lives and livelihoods⁶². Specifically, if rural households perceive flood events as psychologically close in terms of GD and accept that they will be affected in the near future, they are more likely to implement protective measures before the flood occurs.

TD

The study results indicate that TD has a significant negative impact on beliefs, risk perception, and protective responses before a flood, confirming hypotheses H2, H6, and H10. These findings are consistent with previous research conducted by^62,65,72. This finding suggests that, even when individuals acknowledge the likelihood of a flood and its severe consequences, they may still perceive it as psychologically distant due to its potential impacts being far in the future. In essence, people tend to attribute the probability of a flood to another time—typically the distant future—rather than considering its immediacy^68,69,70. National survey data from the United States support this perspective, indicating that while people recognize climate change impacts occurring in the present, they tend to perceive severe effects affecting large populations as events that will unfold in the relatively distant future, beyond 2050⁸⁵. Yazdanpanah et al.⁶⁵ examined the effect of psychological distance on climate change adaptation behaviors and found that individuals who believe climate change will primarily impact future generations did not alter their activities accordingly. Similarly, Rasool et al.⁹¹ in a study on flood management, concluded that individuals who perceived flood risks as imminent were more proactive in implementing protective measures. Therefore, within this context, myths can play a critical role in fostering community preparedness by reducing TD toward destructive events such as floods. In this context, the media can play a crucial role in minimizing the effects of TD on destructive phenomena such as floods while fostering community preparedness for implementing protective measures. By making climate risks feel more immediate and locally relevant, the media can enhance public awareness and encourage proactive action.

SD

The study results indicate that SD has a significant negative impact on flood-related beliefs, risk perception, and protective responses, confirming hypotheses H3, H7, and H11. These findings align with previous research conducted by Sacchi et al.⁷³, Carmi & Kimhi⁷², Jones et al.⁶⁰ and Yazdanpanah et al.⁶⁵. Even when individuals acknowledge the reality, immediacy, and relative proximity of flood impacts, they may still perceive these impacts as socially distant. That is, they may recognize the severity of the issue but attribute its effects to others, feeling less personally responsible or directly affected⁷². Although social and spatial distance are often intertwined, making them challenging to separate, this research suggests that individuals may perceive distance differently along these dimensions. Specifically, climate change impacts tend to be viewed as more severe in developing countries and remote geographical regions (Reser et al., 2012). This psychological factor can act as a barrier to collective action in response to climate-related disasters such as floods, which require the coordinated mobilization of entire communities.

HD

The study results indicate that HD has a significant negative impact on beliefs, risk perception, and protective responses to floods, confirming hypotheses H4, H8, and H12. These findings align with previous research conducted by Trope & Liberman⁶⁶ and Yazdanpanah et al.⁶⁵. In this context, HD refers to an individual’s subjective confidence or expectation regarding the occurrence of a future event, such as a flood. This confidence is shaped by existing information, past experiences, and personal perceptions of the environment. Hypothetical distance influences decision-making, planning, and behavioral responses to future risks⁶⁶. Given the complexity of climate phenomena, HD can manifest in various ways. It may shape perceptions regarding whether climate change is occurring (or will occur) and/or influence views on the severity of its impacts, assuming its occurrence¹¹⁹. Perceptions of certainty regarding climate change may be shaped by skepticism propagated in the media. This deliberate “production of doubt” has been linked to declining belief in and concern about climate change^119,120. The uncertainty surrounding climate phenomena such as flooding can lead to inaction, as individuals may perceive the issue as distant and, therefore, non-threatening. Additionally, the global scale of climate change may further contribute to its perceived remoteness and uncertainty¹²⁰. These two effects can interact, with the perceived remoteness of the issue reinforcing uncertainty and the “production of doubt” further distancing individuals from the problem. Essentially, the availability of arguments questioning the existence of climate change contributes to psychological distancing along the HD dimension⁶⁷. A key challenge in reducing HD regarding climate impacts is that individuals often misunderstand climate change predictions, such as flood occurrence, and frequently misinterpret probabilities¹²¹. A lack of understanding of climate change predictions and misinterpretation of probabilities can diminish the urgency of addressing climate risks. This issue may stem from the complexity of scientific information, the way data is presented, or inherent cognitive biases. Therefore, when individuals accurately comprehend the probability of flooding, they are more likely to implement protective measures in advance and ensure adequate preparedness.

Belief in climate change

Moreover, belief in climate change has a significant positive impact on flood protection responses, confirming hypothesis H13. These findings are consistent with prior research by Hyland et al.⁸⁰, Le Dang et al.⁸¹ and Savari et al.⁹⁷. The analysis found that belief in climate change plays a crucial role in shaping individuals’ behaviors, particularly in the adjustments they make to their agricultural practices in response to its consequences^{122,123,124,125}. Such an understanding can assist planners in designing and implementing effective climate change mitigation policies^{36,126,127,128,129,130}. Additionally, one of the most significant and influential determinants of conservation behavior is individuals’ awareness or belief in climate change⁸¹. Vanclay et al.¹³¹ emphasized that farmers’ beliefs about climate change are critical in determining the likelihood of adopting adaptation strategies, as ignorance or disbelief can lead to inaction or outright rejection of climate threats¹³². While beliefs exist independently of risk perceptions and the negative impacts of climate change, risk perceptions are largely shaped by underlying beliefs⁸⁰. Local communities’ beliefs and perceptions of climate change play a crucial role in shaping their responses to environmental phenomena such as flooding. When individuals perceive climate change as a serious and immediate threat to their community, they are more likely to adopt preventive measures and respond effectively. Conversely, if they view climate change as insignificant or non-threatening, they may exhibit indifference toward its consequences and fail to prepare adequately.

Flood risk perception

The study results indicate that flood risk perception has a significant positive impact on local communities’ adoption of protective measures, confirming hypothesis H14. These findings align with prior research conducted by Savari et al.³⁶, Zaalberg & Midden¹³³, Calogero et al.¹³⁴ and Wang & Zhou¹³⁵. Understanding and awareness of risk are essential for enhancing flood preparedness and response behaviors. Direct experiences with flooding increase vulnerability awareness and strengthen coping strategies¹³³. Additionally, research suggests that negative emotions often drive precautionary measures, while concerns about future floods can serve as strong predictors of protective behaviors¹³⁶. Research has primarily emphasized direct and indirect hazard experiences as crucial factors in shaping risk perception¹³⁷. Individuals who have personally experienced or witnessed the severe consequences of hazards—such as life-threatening incidents or damage to housing—tend to perceive a heightened level of risk¹³⁴. Individuals who have experienced severe hazard effects often retain vivid memories of these events, which can evoke intense fear. However, over time, if they do not encounter similar hazards again, their perception of high risk may diminish³⁶. Conversely, those who have suffered little or no harm tend to perceive a lower level of risk¹³⁸. Individuals who effectively manage lower-severity risks often overestimate their ability to handle future risks, assuming their severity will remain unchanged¹³⁹. This misjudgment can hinder their ability to accurately predict the potential consequences of future hazards¹³⁵. However, risk perception varies among individuals, and experiencing similar hazardous events does not necessarily lead to uniform assessments of risk¹⁴⁰. The social and economic contexts in which disaster risk emerges play a crucial role in shaping perceptions of risk and associated factors⁹⁷. Many low-income residents in informal settlements prioritize proximity to employment opportunities over housing conditions, often choosing to live in high-risk areas that are susceptible to flooding. Halpern-Felsher et al.¹⁴⁰ found that individuals who have experienced natural hazards perceive themselves as more vulnerable to future events such as flooding compared to those who have not. This finding suggests that people shape their hazard risk perceptions based on their interpretation of past experiences¹¹⁴. Public flood risk perception is a key factor in the success of flood resilience and climate risk management strategies¹⁴¹. Research indicates that individuals with heightened flood risk awareness are better prepared to handle disasters¹⁴². Research indicates that risk perception and protective behaviors toward natural hazards—such as floods—are significantly influenced by socio-economic and educational factors across different regions of the world³⁶. For instance, notable differences emerge when comparing South Asia and Sub-Saharan Africa. In South Asia, flood risk perception is generally high, particularly in flood-prone countries like India and Bangladesh. However, the extent to which protective behaviors are adopted often depends on individuals’ income levels and access to resources; in lower-income areas, financial limitations may prevent households from implementing adequate preventive measures. Conversely, in parts of Sub-Saharan Africa, flood risk perception tends to be lower, especially in regions where flooding is infrequent or where there is limited access to information on climate change. Cultural and social norms also influence the adoption of protective strategies. In some communities, traditional solutions are often favored over scientific or technical approaches, which can affect the effectiveness of risk mitigation efforts.

Policy implication

To support rural communities in reducing flood risks and implementing protective measures, this study offers recommendations for local authorities and policymakers, grounded in the dimensions of psychological distance.

• (1) Reducing the geographical distance.

• Producing local educational content: Enhancing public preparedness for flood hazards requires educational interventions that are culturally resonant, visually informative, and experientially grounded. This can be achieved through collaborative efforts involving agricultural jihad centers, village councils, rural schools, NGOs, and community-based organizations to develop localized educational content. This content—comprising images, videos, and narratives of historical flood events in comparable regions—should be produced in the local language and reflect indigenous cultural elements to strengthen the community’s emotional connection to the risk.

• Experimental storytelling: Experiential storytelling, through the documentation and dissemination of lived experiences from flood-affected residents, can be delivered in the form of cultural sessions held in mosques, cooperatives, NGOs, or localized digital platforms. These narratives deepen individual and collective understanding by contextualizing the threat within familiar and relatable experiences.

• Interactive risk maps: Furthermore, the introduction of interactive flood risk maps, accessible via digital platforms or installed as simplified visual displays in village councils, schools, and other public rural spaces, can effectively communicate spatial vulnerability. Accompanied by plain-language explanations, these tools empower residents with clear and actionable insights tailored to their specific geographic context.

• (2) Reducing the temporal distance.

• Climate-driven scenario campaigns: In coordination with the Meteorological Organization, agricultural centers, and village Islamic councils, regular outreach campaigns should be launched to inform the public about rainfall patterns, flood risks, and climate change trends. These initiatives should include comparative analyses with historical data to contextualize emerging threats and foster informed community discourse.

• Personalized local warning notifications: Leveraging district-level SMS platforms and simple mobile applications developed by local administrations or humanitarian organizations such as the Red Crescent, timely and practical flood warnings can be disseminated. These systems should focus on seasonal risk periods, particularly the rainy season, and provide household-specific guidance to promote immediate protective actions.

• Urgency-focused public messaging: To emphasize the imminent nature of flood threats, the design and distribution of visual and audio materials—such as posters, brochures, and announcements—should prioritize near-term risks over long-term environmental issues. These materials should be strategically placed in high-traffic community spaces, including health centers, schools, and village council venues, to maximize visibility and impact.

• (3) Reducing the social distance.

• Leveraging trusted local figures: Recognized community leaders—including village trustees, school principals, and village heads—should be engaged to communicate risk-related information, deliver safety guidance, and encourage preventive action. Utilizing these respected figures ensures that critical messages are received with credibility and cultural resonance, thereby improving public responsiveness.

• Establishing local action groups: In partnership with the district administration, Red Crescent Society, and Islamic councils, trained resident-led groups should be formed to facilitate peer-to-peer knowledge dissemination, coordinate emergency evacuations, and administer basic first aid during crises. These community teams serve as frontline agents of preparedness, bridging institutional support with on-the-ground action.

• (4) Reducing the hypothetical distance.

• Simulated emergency drills: Annual flood simulation exercises should be organized in strategically selected locations—such as village squares or riverside zones—with broad community participation, including students, residents, and local institutions. These drills enhance psychological readiness and ensure community members acquire essential procedural knowledge for evacuation and crisis response.

• Immersive multimedia education: Utilizing virtual reality (VR) or 360-degree imagery, simple and context-specific educational content can be developed to depict post-flood scenarios of affected villages. Displaying this content in schools, health centers, and community meetings allows individuals to visualize disaster impacts, strengthen risk perception, and promote proactive engagement with preparedness measures.

Conclusions and limitations

This study aimed to examine the factors influencing the adoption of protective measures prior to flood events in southwestern Iran. The research employed CLT as its theoretical framework, demonstrating high effectiveness in explaining 67.9% of the variance in flood preparedness. Specifically, geographical distance, temporal distance, social distance, hypothetical distance, belief in climate change, and flood risk perception all exert a significant positive impact in this domain. Despite its significant findings, this study presents two key limitations that warrant consideration in future research. (1) The variables in this study accounted for 67.1% of the variance in protective measures taken before a flood, leaving 32.9% unexplained. To address this gap, future research should aim to strengthen the predictive power of the CLT in this area by identifying additional influencing variables or integrating it with other psychological models. (2) This study employed a quantitative approach. To gain deeper insights, future research should consider qualitative paradigms such as phenomenology and grounded theory to capture individuals’ experiences and protective strategies in flood management. (iii) One of the primary limitations of this study is the timing of data collection, which was conducted immediately following the flood event. Under such conditions, participants’ responses may have been influenced by heightened emotions and the stressful atmosphere associated with the crisis, potentially leading to an increased tendency to report protective behaviors. This introduces the possibility of response bias, which should be taken into account when interpreting the findings. To address this concern, it is recommended that future research adopt longitudinal or mixed-method designs to assess the persistence and stability of protective behaviors over time.