Cross-sectional studies.

A cross-sectional study is a research design that collects, analyzes, and examines data at a single point in time for a specific population. In this study, a cross-sectional study was chosen to investigate community resilience in the aftermath of two hurricanes during the COVID-19 pandemic.

This study selected the Lempira Department of the Republic of Honduras for this study. This Department was identified as a target area for the Strengthening Systems Based on the National Health Model project for 2013-2018 in the Republic of Honduras that promoted primary health care. This study collected data from a total of 139 households in the community in Lempira from August to November 2021.

Cluster sampling was used in this study. All households under the jurisdiction of one of the ten health centers in Lepaera city, Lempira department, were included in the study. Based on the list prepared by the health center, local collaborators face-to-face interviews were done while visiting their houses. The eligibility criteria for the survey were participants at least 18 years old and females who could respond to the questionnaires.

The purpose and significance of the study, preservation of anonymity, freedom to participate, no disadvantages associated with non-participation, and freedom to withdraw from this study were explained verbally to the participants, and informed consent was obtained before the study was conducted.

Ethical approval was obtained from the Scientific Research Unit Biomedical Research Ethics Committee (CEIB) of the Faculty of Medicine of the National Autonomous University of Honduras (IRB 00003070) and the Research Ethics Committee of Himeji University (IRB 2020-N09).

In this study, community resilience was the outcome variable. The CCRAM-10 was selected for this study as a measure of community resilience "recovery from crisis" as it relates to disaster recovery and reconstruction. The CCRAM-10 was assessed using the scale was back-translated to confirm retaining the scale’s original meaning. The Cronbach's alpha was found within the acceptable level, i.e., 0.823-0.950 in this study. Responses were rated on a 5-point Likert scale.

We chose the CCRAM-10 methodology because it has identified factors that indicate community resilience, including expert opinion, and because CCRAM-10 can provide information that will help residents and decision makers anticipate emergency challenges and plan based on the results of this study. It was recognized as an indicator that could be used to identify community and resident characteristics for community building.

Another scale in this study was social capital as an explanatory variable. As a measure of social capital, this study used Szreter and Woolcock's (2004) bonding, bridging, and linking social capital framework. Responses were rated on a four-point Likert scale. For the social capital scale's indicators, Cronbach's alpha ranged from 0.75 to 0.815.

Other measurements were socio-demographic characteristics, including age, gender, education level, marital status, religion, employment, and years of residence. Similarly, other variables included disaster preparedness and whether the damages by the two hurricanes in 2021.

This study calculated the descriptive statistics such as frequencies, percentages, and means at first. For the analytic study, this study applied multiple regression analysis. The statistically significant threshold level was set to p less than 0.05. Data analysis was performed using Jamovi version 2.3.0.0 (The jamovi project, 2022).

Table 1 shows that the mean age of the respondents (housewives) was 38.4 years (SD 16.5). The average length of residence was 37.9 years (SD 17.5). Occupation is employee or housewife 130 people (93.5%). People had elementary school-level education 106 people (76.3%), higher education 24 people (17.2%), and none 9 people (6.5%). 127 people (91%) have a partner, and 9 (6.5%) are single. 113 people (81%) were Catholic, and 22 (16%) were none. The household income is low at 97 people (70%) and, on average, 42 people (30 %). Regarding their health status, 106 people (77 %) have no chronic diseases, and 29 people (21 %) have chronic diseases and no problems.

The target population of this study is characterized by healthy housewives who have completed elementary school.

Table 1 Table 1 Demographic data

		n	mean(sd)		(	%	)
Gender	Female	139
Age (group)		138	38.4(16.5)
Year of residence		138	37.9(17.5)
Occupation	Professional or boss	4			(	2.9	)
	Industry or manufacturing Merchant.	0			(	0	)
	Student	4			(	2.9	)
	Employee or housewife	130			(	93.5	)
	Retired or unemployed	1			(	0.7	)
Education	Elementary school	106			(	76.3	)
	Higher education	24			(	17.2	)
	None	9			(	6.5	)
Marital status	Single	9			(	6.5	)
	Married /Free union/Living with partner	127			(	91.4	)
	Single parent	0			(	0	)
	Separated/Divorced/Widowed	3			(	2.2	)
Religion	Catholic	113			(	81.3	)
	Protestant	0			(	0	)
	None	22			(	15.8	)
	Evangelica	4			(	2.9	)
Household income	Much lower than average	0			(	0	)
	Low	97			(	69.8	)
	Average	42			(	30.2	)
	Higher	0			(	0	)
	Much higher than average	0			(	0	)
Health Status	Chronic disease no problem	29			(	21.0	)
	Chronic disease problem	3			(	2.2	)
	No Chronic disease	106			(	76.8	)

Table 2 shows that the types of disasters that were expected to happen in their place of residence were landslides (97.1%), floods (20.1%), hurricanes (19.4%), and earthquakes (0.7%), with almost all respondents expected landslides. The response asked about their preparedness; nearly all had done nothing (99.3%).

Tabla 2 Tabla 2 Expectations anad preparedness for disaster

Preparedness		n	%
Type of disaster that people imagine may occur in my residence (choose more than one answer)	Flood	28	20.1
	Hurricane	27	19.4
	Earthquake	1	0.7
	Land slide	135	97.1
	Others	0	0
	None	0	0
Preparedness for disasters (choose more than one answer)	I have not done anything.	138	99.3
	A shelter (bunker) or safe room.	1	0.7
	Food and water for emergency use.	0	0.0
	Waterproofing equipment and protective accessories.	0	0.0
	Medication and first aid kit.	0	0.0
	Hygiene items such as masks and soap.	0	0.0
	Identify shelters and evacuation routes near your home.	0	0.0
	Are ready to take their valuables with them at any time.	0	0.0
	Have a good line of communication with my neighbors.	0	0.0
	Others	0	0.0

Table 3 Table 3 Damage from hurricane Eta and Iota

Disaster Experience		n	%
Damage from hurricane Eta and Iota		139	100.0
	No	114	82.0
	Yes	25	18.0
Type of damage	House roof damage	1	4.0
	Landslide	22	88.0
	Flooding	0	0.0
	others	2	8.0

Table 3 shows that in the two hurricane disasters experienced in 2020 (Eta and Iota), 25 people (18.0 %) experienced damage. Most of those affected were affected by landslides (88.0%), followed by damage to house roofs (4.0%).

To determine the factors that influence the status of community resilience in the aftermath of two hurricanes during the COVID-19 pandemic in this section.

To examine the impact of social capital and background information on general community resilience, this study used general community resilience and its five components (Leadership, Collective efficacy, Preparedness, Place Attachment, and Social trust) as objective variables and the explanatory variables Multiple regression analysis was conducted using "Bonding Social Cohesion," "Bridging Social Cohesion," and "Political Trust" for social capital, and "Year of residence" and "Education Level" for background information.

No influence of other explanatory variables such as "Household income" and "Marital status" was found in the Background information.

The possibility of multicollinearity was examined by VIF and was determined to be less than 3 for all explanatory variables, indicating that multicollinearity was unlikely to occur.

The coefficient of determination R² of the model was .423, and the adjusted coefficient of determination for degrees of freedom was .39. A total of five variables (three variables for social capital and two sociodemographic factors) explained approximately 40 % of the variance in general community resilience.

Table 4 shows that the number of subjects in the model, and the partial regression coefficients with a 95 % confidence interval.

The most striking result of this study was the gap in actual preparedness and awareness of disasters in Lempira. To the question "What have you done so far to prepare for disasters?" most respondents answered, "I have done nothing.” However, most of the respondents answered "Agree" to the following two questions “My place of residence is prepared for emergencies” and “Residents know what their roles are in an emergency” on disaster preparedness in CCRAM-10.

One of the reasons for this is that instead of each household preparing for a disaster, people trust the ability of local leaders trained in disaster response. It is also likely that a part of the Honduran culture believes that if you have walls and a roof over your head, you will be able to survive a disaster. Many people do not want to leave their homes during a disaster. This study believe this is because even when there are storms or floods, they would rather die inside than leave their homes because of their culture.

In other regions, several previous studies have reported disaster awareness in local communities of developing countries. The Moken, an ethnic minority affected by the Indian Ocean tsunami, had no concept of tsunami or disaster. They saw the coming tsunami phenomenon as part of the flood myth Laboon (reminiscent of the flood myth recorded in the Book of Genesis in the Old Testament) and by the oral tradition that "when the tide suddenly recedes abnormally, Laboon comes," they fled to higher ground. Their lives were saved (Suzuki, 2011). A survey in Bangladesh found that few people evacuate during disasters, indicating a widespread fatalistic view of natural disasters among the local population. Other reasons for not evacuating included fear for the safety of property and livestock, overcrowded shelters, and unsafe places for women (Jagnoor et al., 2019).

Thus, in considering preparedness as one of the components of local resilience, people's awareness of disasters varies, and understanding the culture of each region is essential in enhancing preparedness, and a culturally sensitive response is required.

In understanding the awareness of disaster preparedness, this study have to know first how people in the community think about risk. In Australia, a checklist has been developed to link efforts to understand people's attitudes to behavior change. As a first step, they assessed their preferences and value priorities when at risk of being severely affected by disaster losses. By gaining new knowledge in the form of stories, concepts, understandings, narratives, and data about key drivers of vulnerability from a broad range of people, these stories and identified system patterns highlight the tensions and value conflicts among different parts of society and various roles within organizations, and the other they highlighted how ideas could arise (UNDRR, 2022).

The results of this study indicate that in general community resilience, which integrates the five items, the “Bonding Social Cohesion” hurt Community Resilience, while “Bridging Social Cohesion” and “Political Trust” had a positive impact on Community Resilience.

The classification of whether social capital is conjunctive, bridging, or linking is important (Berkman et al., 2014). Coupled social capital is a network whose members share similar background factors in terms of hierarchy, race/ethnicity, etc., or the resources members have access to within such a network. Bridging social capital is a resource that "bridges" across class, race/ethnicity, and other social characteristics and is accessible between networks (Berkman et al., 2014). In Lempira, a rural area in the target region of the Republic of Honduras, networks of remarkable homogeneity were a factor that lowered community resilience, and bridging networks among heterogeneity were a factor that raised community resilience.

There are several possible explanations for this result. In particular, the fact that those around them are not prepared for disaster preparedness leads to their lack of preparedness, causing homogeneity to drag them down in terms of disaster preparedness. This can be described as a homeostatic bias, where people around them are not doing it, so they do not need to prepare themselves. These behaviors during disasters are consistent with people are influenced by the evacuation behavior of those around them (Ohtake et al., 2020) in evacuation behavior.

It can also be explained with the Social Contagion Theory (Scherer & Cho, 2003), which states that people's attitudes and behaviors are formed under the influence of others with whom they are directly connected.

As a proposed solution, “informing people that their evacuation behavior is influenced by the evacuation behavior of others around them and making them aware and aware that their evacuation behavior affects others around them will promote evacuation behavior” (Ohtake et al., 2020, P.91). This study can propose a way to convey messages that lead to disaster prevention behavior by utilizing the nudges.

The UNDRR defines a disaster as

Serious disruption of the functioning of a community or a society at scale due to hazardous events interacting with conditions of exposure, vulnerability, and capacity, leading to one or more of the following: human, material, economic and environmental losses, and Impacts. A disaster is a situation that exceeds the response capacity of the area and requires assistance from outside sources.(UNDRR, 2009, P.9)

It makes sense that bridging would be an effective tool for community resilience in the context of external relationships in situations that exceed the local response capacity to accept and receive support from outside sources. “Bridging" is a connection to the outside world that transcends generations and roles and can be interpreted as a “weak tie.” Novel information is more likely to be obtained from "weak ties" with low frequency of contact and high heterogeneity than from "strong ties" with a high frequency of contact and high similarity (Granovetter, 1973). Disasters are a place and a time in which relationships in peacetime and "weak ties," loose relationships with the outside world, overlap. The weak ties with the outside world may work effectively as external support.

"Political trust" refers to the "linking social capital" in this study. Higher linking leads to higher community resilience.

The community resilience for those with elementary school education was higher than for those with no education, but not different from those with more than elementary school education.

There is no study about education and community resilience.

In Honduras, primary education is provided to children between 6 and 15. Although there is a movement to make nine years of primary education compulsory, a total of six years is currently compulsory.

Compulsory education in Honduras consists of one year of preschool, nine years of basic education, and two to three years of secondary education (two years only for the technology program), for a total of 12 to 13 years. However, students with poor grades repeatedly fail out of school, and only 50 % of children graduate from grades 1 to 6 of basic education (equivalent to elementary school in Japan) in 6 years, with some dropping out and the completion rate being 86 % (JICA, 2020).

One of the purposes of compulsory education is to nurture people who can live in a world of "I" and the world of "we" (Ministry of Education, Culture, Sports, Science and Technology, 2005). It is essential to have a view of oneself not only about oneself but also about the people around one.

This study found differences between those who had compulsory education and those who did not. The contrast between educational background and community resilience may be due to the difference between the world of "I" and "we" nurtured through compulsory education or the difference between the connection with other people and social skills.

Limitations of the current study are that the study subject was limited to women and that information on culture was not taken. Therefore it is difficult to determine how gender and cultural background affect disaster preparedness and community resilience. Future research should be conducted from these perspectives.