deliverable 3.2.3 preliminary data analysis summary report · 2014. 12. 9. · 3 introduction...

Deliverable 3.2.3 Preliminary data analysis summary report

Tewkesbury, UK

Prepared by: University of Northumbria

Date: 15th December 2009

2

Table of Contents Pg No

Introduction 3

Background 3

Survey Objectives 8

Methodology 9

Data Description 11

Data Analysis 15

Discussion 25

3

Introduction

MICRODIS is an Integrated Project funded under the EU Sixth Framework Programme – Thematic Priority 6.3 Global Change and Ecosystems (Contract number GOCE-CT-2007-036877).

Disaster losses are increasing with great consequence to the survival, dignity and livelihoods of individuals and communities, particularly of the poor in developed and less developed countries. Disaster risk arises when hazards interact with physical, social, economic and environmental vulnerabilities. In the past two decades, more than 200 million people have been affected, on average, every year by these extreme events. Environmentally unsound practices, global environmental changes, population growth, urbanisation, social injustice, poverty, conflicts, and short-term economic visions are producing these vulnerable societies. This takes on particular urgency in the face of long-term risks brought about by climate change, and goes beyond environmental degradation or the mismanagement of natural resources. There is now international acknowledgment that efforts to reduce disaster risks must be systematically integrated into policies, plans and programmes for sustainable development and poverty reduction. The MICRODIS project locates itself within this above framework. The two regions which form the focus of the MICRODIS project are:

1. European Union, associated countries and new accession states: Belgium, France, Finland, Germany, the Netherlands, Norway, the United Kingdom.

2. South and Southeast Asia regions: India, Indonesia, the Philippines and Vietnam. These regions have been selected based on their high frequency of extreme events and the impact on affected communities. There are twelve broad and twenty-three sub-groups of distinct extreme events, ranging from chronic slow onset phenomena to acute rapid onset ones. The health and socio-economic impact implications differ vastly between these twenty three types and addressing all of these would compromise the quality and applicability of the project results, risking over-generalisation. In both Asia and the European Union, three types of extreme events, namely floods, earthquakes, and windstorms, account for nearly seventy-five percent of the occurrence of all extreme events. The MICRODIS project will concentrate on these three phenomena.

1) Background

1.1) Disaster Profile of UK

UK’s main vulnerability in terms of disasters is to floods and much of the UK’s emergency work is around responses to floods. As such, the Environment Agency, a government nodal agency, is mandated with flood control measures. Whilst

4

the UK has been affected by floods from time to time, the 2007 floods were considered to be the most serious inland flood since 1947. It took the lives of 13 people, affected approximately 48,000 households and 7300 businesses, causing a damage of billions of pounds1.

The government initiated an independent review of the floods of 2007, also referred to as the Pitt report, suggesting that the UK had received unprecedented rainfall during June and July 2007, the wettest ever since national records began in 1766 in the UK. The total cumulative rainfall in May, June, July 2007 was averaged at 395.1mm across England and Wales – higher by more than double the usual levels2.

Amongst the several counties affected was Gloucestershire where 350,000 people were left without mains water supply. This was referred to as the “most significant loss of essential services since the second world war”3. Other critical infrastructure and essential services disrupted were power supplies, transport links and telecommunications. Tewkesbury, the site chosen for the UK survey is one of the towns located in the county of Gloucestershire.

1.2) Tewkesbury Town: Disaster and Topographic Profile

Tewkesbury town is part of the Severn River Basin District, the third largest in England and Wales with an area of 21,590 km2 (Environment Agency, 2008). The town itself occupies an area of 1.8 km2, with much of the area comprising ‘green space’ (1.3 km2) and domestic gardens (0.1 km2). The area of water within the town is 0.07 km2 (73,770 m2).

The Severn is the longest river in Britain with a length of 280 km, while the river Avon, one of the main tributaries of the Severn, is 185 km long. Tewkesbury is situated at the junction of these two rivers and part of a broad flat floodplain valley (Figure 1) with very little woodland.

The 2007 floods were a result of the surface water but also flooding of the river Severn. Flooding is the only major disaster risk in Tewkesbury, and the Severn catchment has a long and well-documented history of flooding (EA, 2008). Tidal effects of the river Severn are usually confined to the stretch of river up to Gloucester and only reach Tewkesbury in unusual circumstances (EA, 2008).

Tewkesbury is situated at the Avon and Severn confluence (Figure 1) and has thus always been prone to fluvial flooding due to its position on the main fluvial floodplain and the potential contribution of two separate river catchments to flooding events. There are 1,800 people in 800 households situated directly on the floodplain in the Tewkesbury district (Gloucestershire County Council, 2007).

Prior to the severe 2007 flood there have been five other recent major flooding events within the catchment (February 2004, February 2002, Autumn 2000, October 1998 and Easter 1998). However, Tewkesbury town itself was only affected by the flooding in autumn 2000, which had been the largest flooding event on record in the lower Severn catchment before 2007 (EA, 2008).

The first recorded flood in Tewkesbury was in 1484 and since then the town has experienced flooding events approximately every 30-50 years. The largest flood level of historic flooding events (records from 1862-1990) was recorded in 1947, triggered by warm rain falling on snow (EA, 2008).

1 The Pitt Review: Lessons learned from the 2007 floods, independent report commissioned by the UK Government. 2 ibid 3 ibid

5

Figure 1: Location of Tewkesbury in relation to the rivers Avon and Severn

The Social Flood Vulnerability Index (SFVI) for Tewkesbury is very low (EA, 2008) and people in this town are assumed to be better able to respond to and recover from flooding events than in more deprived and/or urban areas of the catchment.

1.3) Tewkesbury: The context of the floods of 2007 The district of Tewkesbury received 80-90 mm of rain on the 20 July 2007, which amounts to almost two months rainfall in just one day and caused severe flooding of the town (EA, 2008). A water level of 12.93 metres above sea level was recorded at the river Severn on 23 July 2007 (previously the highest recorded level was 12.80 in March 1947.

An estimated 810 properties were affected by flooding in Tewkesbury, with the centre of town being completely cut off due to flood waters (EA 2008). 1,500 buildings were flooded in Tewkesbury by both flash and fluvial flooding (Stuart-Menteth, 2007).

6

1.4) Tewkesbury Town: Socio-economic and demographic profile4 Tewkesbury has a well-preserved medieval character and is of considerable historical and conservation interest. For this reason, tourism is the major source of income and employment in Tewkesbury.

The district of Tewkesbury has an average annual income of 26,953 GBP (GMB, 2005), well in excess of the English average of 22,400 GBP (South West Observatory, undated) and is one of the wealthier regions of the UK.

In this report, we outline the general socio-economic and demographic profile of Tewkesbury Town only.

1.4.1 Demographics

The district of Tewkesbury has a population of 76,405 while Tewkesbury town itself has a population of 10,087. Of these, 4,997 are males (49%) and 5,117 females (51%) living in 4,523 households, with the majority (69%) living in 1-person (33%) and 2-person (36%) households. As per the census 2001, the total population aged 16-74 was 7341. Household composition is summarised in Table 1, with 26% of all the households being pensioners or retired households. Of the total population, 11% is reported to be retired or on pension.

Table 1 : Household Composition of Tewkesbury

Type of Household Number of Households Percentage of all Households

1- and 2-person all pensioner 1169 26

Couples with no children 905 20

Couples with one dependent child 321 7

Couples with two or more dependent children 528 12

Couples with all non-dependent children 290 6

Lone parent households 392 9

Other households 200 4

Source: 2001 census

1.4.2 House ownerships

1,361 of dwellings in Tewkesbury are owned outright and a further 1,849 are owned with a mortgage. 872 properties are rented from the Local Authority or other social housing providers, 291 properties are privately rented. Thus, 71% of properties in Tewkesbury are owner occupied.

1.4.3 Ethnic and Religious profile

The ethnic profile of the Tewkesbury shows that 98.7% of the population is white, of which 96.8% identifying themselves as White British.

The following table 2 shows the religious profile of Tewkesbury with 76.3% being Christians and another 15.5% not identifying with any religion.

4 The demographic and socio-economic data is derived from the 2001 census.

7

Table 2 : Religions of Tewkesbury Population

Christian Buddhist Hindu Jewish Muslim Sikh Other No religion Not stated

Persons 7699 10 5 10 35 3 21 1560 743

Percentage 76.3 0.1 0.05 0.1 0.4 0.03 0.2 15.5 7.4

Note: Percentage figures may not add up to 100 due to rounding.

Source: 2001 census

1.4.4 Employment:

A glance at the spread of employment undertaken by its residents (Table 3), suggests that the manufacturing sector along with wholesale and retail trade accounts for almost 40% of all employment, followed by real estate work with 12%.

Table 3: Industry of employment

Number of persons % of employed population

Agriculture, hunting and forestry 56 1.1

Manufacturing 1305 25.4

Electricity, gas and water supply 47 0.9

Construction 278 5.4

Wholesale and retail trade, repairs 835 16.3

Hotels and restaurants 290 5.6

Transport, storage and communications 271 5.3

Financial intermediation 165 3.2

Real estate, renting and business activities 620 12.1

Public administration, defence, social security

327 6.4

Education 307 6

Health and social work 405 7.9

Other community, social and personal service activities

222 4.3

Note: Percentage figures may not add up to 100 due to rounding.

Source: 2001 census

8

1.4.5 Education

There are six schools, both primary and secondary, located directly in the town of Tewkesbury, three of them local authority schools including one school for children with learning difficulties.

1.4.6 Health

There are three doctors’ surgeries and a small 49-bed NHS-run community hospital in Tewkesbury. Specialist doctors and larger hospitals are accessible in nearby Cheltenham and Gloucester.

Of the inhabitants of Tewkesbury in 2001, 7,012 (70%) reported to be in good and 2,324 (23%) to be in fairly good health; 751 (7%) people described their health as not good. 1,624 (16%) people reported that they suffered from a long-term limiting illness, 948 of these were pensioners. It was found that long-term limiting illnesses were more prevalent in people with no central heating and no access to a car.

2) Survey Objectives

MICRODIS is a project with the overall goal to strengthen preparedness, mitigation and prevention strategies in order to reduce the health, social and economic impacts of extreme events on communities.

2.1) Broad Objectives

To strengthen the scientific and empirical foundation on the relationship between extreme events and their health, social and economic impacts

To develop and integrate concepts, method, tools and databases towards a common global approach To improve human resources and coping capacity in Asia and Europe though training and knowledge sharing

For example, the MICRODIS project will, among others, specifically aim at:

developing an integrated impact methodology establishing an evidence-base of primary field research through surveys increasing the coverage accuracy and resolution of global disaster data

2.2) Objectives for Tewkesbury Survey

The survey was conducted within the overall MICRODIS framework and thus the survey objectives are firmly located within the overall project objectives and aims.

The UK field surveys have the following specific objectives:

to test, evaluate and further the development of the integrated impact tools for standardised data collection in disasters;

to further develop the scientific understanding of the health, social and economic impacts of flooding in the UK context;

to contribute to establishing a sound evidence base of field data on disaster impacts;

to design and validate a tool- and site-specific data entry system;

to identify and describe differential social impacts of flooding in the UK.

9

3) Methodology 3.1 Sampling frame for Tewkesbury

The basic information for determination of the sampling frame for Tewkesbury is the flood outline map supplied by the Environment Agency (Fig. 2), as addresses of flooded properties could not be obtained due to data protection reasons.

Flood Outline Map : Fig 2

A total of 810 residential properties were flooded in Tewkesbury in July 2007, which represents the population of flooded households for the survey site.

3.2 Sample size

Determination of required sample sizes for both populations of interest (flooded and non-flooded households, i.e. control group) was achieved by defining the variables of interest. The following primary variables have been identified for the Tewkesbury survey:

1. mental health functioning,

2. physical health functioning,

3. coping / recovery support,

4. social cohesion / social capital,

5. Willingness-to-pay.

Using Cochran’s formula, the overall required minimum sample size of flooded households for the Tewkesbury survey was 428 households.

10

3.3 Sampling method: Process followed

• Sample households were selected geographically on the basis of the flood outline map, as addresses of affected and non-affected households are not available due to data protection act. Thus systematic sampling method was used.

• The above flood outline map – fig2 was used to identify and outline target 1/3 non-flooded and 2/3 flooded households in the Old town and Newtown areas of Tewkesbury.

• This EA map used to outline target areas with Validation of households done through walk-about method by team of interviewers.

• Households counted (over 1200, with a target of 600), pathways created in uniform manner, every other house was chosen at random start of each pathway.

An example of how pathway was created and houses chosen is shown through this Newtown map:

However the team faced several challenges in meeting the required number of sample size. There was a high refusal rate – almost 30%. Several reasons can be contributed to this the chief being survey fatigue – several had participated in previous surveys, cold calling, trauma felt in recalling the flood event, interview technique and profile of the interviewer.

Systematic “pathway” chosen

Pathway with no households

11

The survey fatigues was evident, as many said that they had already been interviewed by other agencies and were not keen to participate in a new survey now. Further, the residents were not given adequate prior notice about the team’s presence and the reasons for undertaking the survey, which resulted in cold calling. Although the team tried to rectify this by dropping leaflets in the sampled household, this could have affected the response rate. Some other respondents were not very keen to recall the flood event again resulting in the refusal. Lastly, the interviewer team was a mixed team in terms of gender as well as ethnic backgrounds and experience, and this could have been one of the reasons for the refusal for the interviews. Female respondents – particularly if alone in the households seemed reluctant to allow male interviewers in the house. The ethnic composition of the interviewers could have affected the response rate, although the experience as well as personality of the interviewers to engage with the community could have counted even more.

In the end the following sample size was achieved:

Completed: 212 (136 flooded and 76 non flooded) (42.1% of total)

Refused: 158 (31.4% of total)

Completed plus refused = 370 (~74% of total)

HH not contactable after many visits: 133 (26% of total)

Final sample number: 503

3.4) Questionnaire development The Tewkesbury questionnaire generally draws from the generic questionnaire developed by the Microdis thematic teams. To maintain the comparability between UK and other sites in Asia, the Tewkesbury questionnaire kept most of the Microdis core, social, health and economic questions, but adapted the content of some of them to UK. The Microdis questionnaire was piloted in Morpeth, and based on them, these adaptations were made.

3.5) Analytical Methods used: The team is using SPSS software for analysis and the preliminary findings in this report are based on basic descriptive analysis as well as chi-square tests.

4) Data Description This subsection outlines the characteristics of the sample of the flooded population interviewed in the survey.

All 136 households which were flooded were interviewed. The data analysis will also treat most of the

presented variables in this subsection as the independent socio-economic variables for further analysis.

4.1) Age Group

The respondents belonged to diverse age groups, with the majority of them (53%) belonging to the age

group between 40-64 years. Around 35% were belonging to the age group above 65 years.

12

Age group

Frequency Percent Valid Percent Cumulative Percent

Valid 18-24 yrs 2 1.5 1.5 1.5

25-39 yrs 15 11.0 11.0 12.5

40-64 yrs 72 52.9 52.9 65.4

65+ 47 34.6 34.6 100.0

Total 136 100.0 100.0

4.2) Gender of Respondents5

About 56% of the respondents were female, whilst 44% were male as seen from the table below.

Gender


Valid Male 60 44.1 44.1 44.1

Female 76 55.9 55.9 100.0

Total 136 100.0 100.0

4.3) Religion of Respondents

The respondents were predominantly Protestant. About 17% said that they did not belong to any religion.

Religion


Valid Muslim 1 .7 .7 .7

Catholic 8 5.9 5.9 6.6

Protestant 100 73.5 73.5 80.1

None 23 16.9 16.9 97.1

Christian 3 2.2 2.2 99.3

Nonconformist 1 .7 .7 100.0

Total 136 100.0 100.0

5 The deviation of the representation of the sample from the census population which outlines the female population at 51% and male at 49% can be attributed to the fact that the sample drawn is from the flooded area only and as such cannot be said to be representing the entire Tewkesbury town.

13

Note: Those in the ‘Christian’ category identified themselves as such and not as belonging to any subcategory such as Catholics, Protestants or nonconformists.

4.4) Marital Status of Respondents

The sample had around 60% married respondents, whilst around 15% were single and another 13% widowed.

Around 9% were found to be divorced. Marital status


Valid Married 81 59.6 59.6 59.6

Single 21 15.4 15.4 75.0

Separated 1 .7 .7 75.7

Divorced 12 8.8 8.8 84.6

Widowed 18 13.2 13.2 97.8

Common law 3 2.2 2.2 100.0

Total 136 100.0 100.0

4.5) Highest level of education

The sample population was fairly well educated with about 85% having some level of school qualifications or

higher degree or professional qualification, as seen from the table below.

Highest level of education


Valid None 20 14.7 14.7 14.7

O levels/CSE/GCSE/School cert 41 30.1 30.1 44.9

A/AS levels, Higher School Cert 13 9.6 9.6 54.4

First degree 21 15.4 15.4 69.9

Higher degree 9 6.6 6.6 76.5

NVQ/GNVQ/HNC/HND/Nation

al Certificate 12 8.8 8.8 85.3

Professional 13 9.6 9.6 94.9

City & Guilds 3 2.2 2.2 97.1

Other 4 2.9 2.9 100.0

Total 136 100.0 100.0

14

4.6) Occupation of Respondents

Although the sample has people working from diverse occupations as seen from the table, 40% belonged to

the category of retired people6. Occupation currently

Frequency Percent Valid Percent

Cumulative

Percent

Valid unskilled and elementary 10 7.4 7.4 7.4

skilled and intermediate 22 16.2 16.3 23.7

professional/managerial 25 18.4 18.5 42.2

self-employed and business

owner 11 8.1 8.1 50.4

Retired 55 40.4 40.7 91.1

Other 12 8.8 8.9 100.0

Total 135 99.3 100.0

Missing System 1 .7

Total 136 100.0

5) Data Analysis

In this preliminary report, we mainly enquire into the social impacts through the lenses of two social variables: household structure and gender. In particular, we analyse households with children and household without children and the differential impact experienced by each. Similarly, we analyse the differential effects of the flood due to gender differences. We also intend to do further analysis of these preliminary findings as well as take into consideration other variables not discussed in this report such as age, religion, occupations, etc. Since the analysis is mainly comparative, the appropriate levels of statistical tests were used which took into account a) the number of groups to be compared b)level of data involved –nominal, ordinal, scale.

Preliminary observations:

6 The representation of retired persons or pensioners is higher in the sample (40%) than in Tewkesbury town which stands at 11%. There are two reasons for this difference: one that the represented sample is only from the flooded area, and not from the entire town and second; it is likely that the interviewers were able to contact pensioners more than the rest of population, as pensioners were more likely to be found at home during the survey hours than the economically active population.

15

5.1) Increase in community contact due to floods was observed.

In general, the community contact increased and the flood experience brought people together. As one respondent says, “when we were in the same situation, we were struggling together in the neighbourhood, so we got to know more people”.

5.2) Displacement patterns and Household structure

Null Hypothesis Ho: There is no relation between displacement pattern and household structure in the population

Alternative hypothesis: Displacement patterns are influenced by the household structure in the population

Displacement due to disaster * Household structure/relationships within household Cross tabulation

Household structure/relationships within household

16

household without

children

household with

children Total

Displacement due to disaster temporary Count 43 36 79

Expected Count 50.3 28.7 79.0

permanent Count 2 1 3


None Count 41 12 53


Total Count 86 49 135


Chi-Square Tests

Value Df

Asymp. Sig. (2-

sided)

Pearson Chi-Square 7.223a 2 .027

Likelihood Ratio 7.463 2 .024

Linear-by-Linear Association 7.169 1 .007

N of Valid Cases 135

a. 2 cells (33.3%) have expected count less than 5. The minimum

expected count is 1.09.

Conclusion:

Since the Chi square significance value for this is 0.027 (less than 0.05), we reject the null hypothesis and accept the alternative hypothesis that there is a relation between displacement patterns and household structure in the population. The data suggests that households without children faced less displacement than those with children.

5.3) Daily routine affected by floods and household structure

Ho: There is no relation between effect on daily routine and household structure in the population

H1: The effect on daily routine is influenced by household structure in the population

Daily routine affected by disaster * Household structure/relationships within household Cross tabulation

17

Household structure/relationships within household

household without

children

household with

children Total

Daily routine affected by

disaster

Yes Count 77 48 125


No Count 9 0 9




Chi-Square Tests

Value Df

Asymp. Sig. (2-

sided)

Exact Sig. (2-

sided)

Exact Sig. (1-

sided)


Continuity Correctionb 3.844 1 .050


Fisher's Exact Test .026 .016



a. 1 cells (25.0%) have expected count less than 5. The minimum expected count is 3.22.

b. Computed only for a 2x2 table

Conclusion:

18

Since Chi square significance value is 0.020 (less than 0.05), we reject the null hypothesis and accept the alternative hypothesis that is effect on daily routine is influenced by the household structure in the population. The data suggests that households with children were more prone to change in daily routine than those without children

5.4) Experience of trauma of the flooding and household structure

Ho: There is no relationship between experience of trauma and household structure in the population H1: There is a relationship between experience of trauma and household structure in the population

Mann-Whitney Test Ranks

Household

structure/relationships within

household N Mean Rank Sum of Ranks

How traumatic flooding

experience at the time

HH with children 48 82.21 3946.00

HH without children 88 61.02 5370.00

Total 136

Test Statisticsa

How traumatic

flooding

experience at the

time

Mann-Whitney U 1454.000

Wilcoxon W 5370.000

Z -3.112

Asymp. Sig. (2-tailed) .002

a. Grouping Variable: Household

structure/relationships within household Conclusion:

Since the significance level is 0.002 (less than 0.05), we reject the null hypothesis and accept the alternative hypothesis that there is a relation between trauma and household structure in the population. Also, from the mean rank 82.21 and 61.02 for HH with children and those without children respectively, it appears that HH with children have a higher trauma than those without children.

5.5) Change in personal relationships due to flood and household structure

19

Ho: There is no relationship between changes in the personal relationships due to flood and household structure in the population

H1: Changes in the personal relationships due to flood are influenced by the household structure in the population


Household



Change in personal relationships

due to flood



Total 131

Test Statisticsa

Change in personal

relationships due

to flood


Wilcoxon W 5075.500

Z -2.855



structure/relationships within household

Conclusion:

20

Since significance level is .004 (less than 0.05), we reject the null hypothesis and accept the alternative hypothesis that changes in the personal relationships are influenced by household structure in the population. From the mean rank of 77.62 for HH with children and mean rank of 59.71 for HH without children, we can say that HH with children did have more changes in the personal relationships than HH without children. Many more households without children reported no change in personal relationships due to floods than those with children. As one family with children said “[We were] affected to a high extent, more stressful, there were more family arguments, obviously as we were all staying in a caravan”.

5.6) Change in the monthly household expenditure and household structure

Ho: There is no relationship between monthly household expenditure and the household structure in the population.

H1: The monthly household expenditure and household structure are influenced by the household structure.


Household



Change in monthly household

expenditure



Total 134

Test Statisticsa

Change in monthly

household

expenditure


Wilcoxon W 2634.000

Z -2.661



structure/relationships within household

21

Conclusion:

Since significance level is 0.008 (less than 0.05), we reject the null hypothesis and accept the alternative hypothesis that experience of change in the household monthly expenditure is influenced by household structure in the population. In general, more households without children noted that there was no change in their monthly expenditure than those with children. Households with children noted an increase in their monthly expenditure.

5.7) Change in personal relationships and gender Mann-Whitney Test Ranks

Gender N Mean Rank Sum of Ranks

Change in personal relationships

due to flood

Male 59 57.57 3396.50

Female 72 72.91 5249.50

Total 131

Test Statisticsa

Change in personal

relationships due to

flood


Wilcoxon W 3396.500

Z -2.549


a. Grouping Variable: Gender

Females Males

22

Conclusion: Since significance value is 0.011 (less than 0.05), we reject the null hypothesis and accept the alternative hypothesis that gender and changes in personal relationships due to flood are related in the population. More females experienced a higher change in personal relationship than males did.

5.8) Gender and seeking help

Ho: There is no relationship between gender and seeking help after the floods in the population H1: There is a relationship between gender and seeking help after the floods in the population

Help sought * Gender Cross tabulation

Gender

Male Female Total

Help sought Yes Count 33 54 87


no Count 27 21 48




23

Chi-Square Tests

Value Df

Asymp. Sig. (2-

sided)

Exact Sig. (2-

sided)

Exact Sig. (1-

sided)


Continuity Correctionb 3.495 1 .062


Fisher's Exact Test .048 .031



a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 21.33.

b. Computed only for a 2x2 table Females: Males

Conclusion: Since the chi square significance value is 0.048 (less than 0.05), we reject the null hypothesis and accept the alternative hypothesis that there is a relationship between gender and seeking of help in the population. Women have sought help much more than the men after the floods.

5.9 ) Marital Status and change in personal relationships experienced

Ho: There is no relation between marital status and change in personal relationship experienced due to flood in the population H1: There is a relationship between marital status and change in personal relationships experienced due to flood in the population

24

Kruskal-Wallis Test

Ranks

Marital status N Mean Rank

Change in personal

relationships due to flood

Married 78 71.96

Single 21 44.33

Divorced 11 60.14

Widowed 17 54.26

Total 127

Test Statisticsa,b

Change in

personal

relationships due

to flood

Chi-Square 13.572

Df 3

Asymp. Sig. .004

a. Kruskal Wallis Test

b. Grouping Variable: Marital status

Conclusion: Since significance level is .004 (less than .005), there is a relationship between marital status and change in personal relationships due to flood. Married persons have experienced higher levels of changes in personal relationships than others as indicated by the higher means.

6) Discussion

As the preliminary analysis shows, household structure as well as gender seem to have had some role to play in differential impacts felt, and suggest a need for a deeper analysis; this particularly in combination with some other variables such as age, occupations etc.

25

The UoN team would like to compare the datasets with India as well other sites such as Philippines which have an explicit interest in exploring gendered impacts of disasters. This initial analysis shows that disasters are experienced as well as responded to differently by men and women. One of our hypotheses is that these differences in the experiences as well as responses could have a bearing on mental health impacts, one of our outcome variables. We plan to explore the gendered impacts of mental health and the predictors for the same. We suspect that they may reveal certain interesting masculine and feminine models that are at work in the different country contexts. These in turn could help us understand how gender works in different contexts and their implications for mental health issues in disasters. Further, this would potentially contribute to not only gender theories but also to a gendered disaster theory and fulfill a relevant gap in the literatures concerned. We may explore similar hypotheses with respect to mental health and social groups in household structures, age, marital status as they too have revealed differential impacts in this preliminary analysis, explored further in conjunction with gender.

Sources :

1) Environment Agency (2008) Managing Flood Risk: River Severn Catchment Flood Management Plan, Consultation Draft Plan May 2008, Environment Agency, Bristol

2) Gloucestershire County Council (2007) Demography of floodplain in Gloucestershire Gloucestershire County Council, Gloucester

3)Gloucestershire County Council (2005) Health Needs and Care Provisions in Gloucestershire, Environment Research Team, Gloucestershire County Council, Gloucester

4) GMB – Britain’s General Union (2005) First time buyers and low paid priced out of home ownership in 39 southwest areas [online] http://www.gmb.org.uk/shared_asp_files/uploadedfiles/D014776E‐529B‐4459‐B4CB‐2F06EC84806F_Housingsw.pdf accessed: 27.09.2008

5) Stuart‐Menteth A. (2007) UK Summer 2007 Floods Risk Management Solutions Ltd., London

6) South West Observatory (undated) State of the South West: Income, Expenditure and Wealth [online] http://www.swo.org.uk/SoSW2007/web/section_229.html accessed: 27.09.2008

deliverable 3.2.3 preliminary data analysis summary report · 2014. 12. 9. · 3 introduction...

Documents