Mapping Fear of Crime as a Dynamic Event in the Whole Journey Environment

Authors: Reka Solymosi, Taku Fujiyama, Kate Bowers

ABSTRACT

Perception of security largely influence people’s experiences with all forms of

transport, be it public, private, walking or cycling, and has knock on effects on

transport mode choice, as well as the accessibility and walkability of an area. It

is therefore important to gather accurate spatial and temporal information of

when and where people feel unsafe within the whole journey environment, to

identify problem areas. Currently, fear of crime information is collected using

retrospective surveys such as Transport for London’s attitudes to safety &

security survey, or the Crime Survey for England and Wales, which result in

inaccurate measures segmented over various stages of the entire journey

experience, unable to cover people’s entire activity space. To correctly identify

‘fear hotspots’ in place and time, experience sampling method applied to a

mobile phone application can be used to collect information on people’s

perceptions along with GPS and time-stamp data. By applying this novel

approach to measuring dynamic events on the move as they happen, problem

areas can be identified based on empirical evidence. This is used to target

preventative interventions effectively to improve perceptions of safety and

security in the whole journey environment, and to identify environmental

features associated with these hotspots to inform town planning and local

initiatives on situational prevention.

INTRODUCTION

What is fear of crime

Fear of crime is the negative emotional reaction elicited by crime and its

associated symbols (Warr, 2000; Hale, 1996). Not only does this fear appear in

those who have been victimised, but also those who have not, making it more

prevalent than actual rates of victimisation (Warr 2000). In the 2010-11 year

43% of Londoners reported their quality of life affected by fear of crime (Gray et

al. 2013). Nationwide, according to the British Crime Survey, 17 % of

respondents were very worried about violent crime, 13 % about burglary and

14 % about vehicle crime (Jansson 2007).Regarding feeling safe on public

transport, TfL’s survey reveals that a fifth of Londoners said that they had felt

worried about their personal security in the last three months, with 53% of

these incidents on a bus followed by 24% on the Tube and 20% on the train

(Transport for London 2013).

Why is fear of crime important

Fear of crime, evidently a present issue, is important to investigate as it has its

own, very real, consequences that are distinct from those of crime. People’s

perceptions of the environment in terms of safety impacts on their experiences

in the areas where they live, work, and travel through. Fear of crime as

experienced while traveling is particularly important, as travel mode choice is

influenced by perception of safety in the journey environment; higher fear of

crime is associated with less cycling and walking, and increased use of private

transport modes (Mitra et al. 2010; McDonald et al. 2010). To encourage use of

public transportation, perception of safety during the entire door-to-door

journey needs to be addressed (Brons & Rietveld 2001). “Perceptions of the

likelihood of being a victim of crime or antisocial behaviour affects travel

choices and can act as a barrier to travel” (Transport for London 2011). As such,

fear of crime in the journey environment poses an issue of accessibility. Groups

such as elderly and vulnerable people, women, and ethnic minorities are

disproportionately affected by fear of crime in transport (Loukaitou-Sideris &

Fink 2009) and experience further disadvantage by barring accessibility of

transport. Liska et al. (1988) report that fear of crime defined in terms of low

levels of perceived safety constrains social behaviour, creating a feedback loop

by which more fear leads to more social isolation and subsequently even higher

fear (see also Garofalo 1981). People’s sense of security is also an important

factor in assessing the walkability of an area (Kelly et al. 2011). Walkability of

the environment affects public transport ridership (Ryan & Frank 2009). Indeed,

according to Transport for London survey results, the transport modes most

affected by fear of crime are walking, travelling by bus, and cycling, with people

feeling safest traveling after dark by door-to-door modes such as car and black

cab (Transport for London 2011). Evidently, fear of crime can prevent people

from using environmentally friendly transport modes, and encourages the use

of private vehicles, increasing carbon emissions, and reducing exercise. Overall,

fear of crime in an area can affect accessibility, environmentally favourable

travel choices, and relative disadvantage as well as other social and economic

elements of a community.

This importance of fear of crime is reflected in governmental attitudes as well;

“how safe people feel and how much crime they perceive to be occurring has

become a priority for policing” (London Transport Committee 2008), and

perception of crime and antisocial behaviour is addressed under the Mayoral

goal of improving safety and security on the transport networks in London

(Transport for London 2011). In order to be able to mitigate this effect, it is

important to establish where, when, and how fear of crime is experienced by all

users of an environment, be they residents, workers, people travelling through,

or other users of public spaces, and cover their entire activity space in the

whole journey environment.

Background in Fear of Crime research

Fear of crime has been studied extensively, and this background aims to give

only a brief introduction into its methodologies, as relevant to measuring

perceptions of security in the whole journey environment. Traditionally,

measurement of fear of crime is done using household surveys. A variant of the

question “How afraid are you walking alone at night?” is asked from

respondents, as well as a series of follow up questions, asking about the

perceived likelihood of falling victim to a particular offence. In the Crime Survey

for England and Wales (CSEW, formerly British Crime Survey (BCS)),

respondents are asked whether they are ‘very’, ‘fairly’, ‘not very’ or ‘not at all’

worried about becoming a victim of crime. Past research has claimed that this

involves an assessment of individuals’ perceived likelihood of becoming a victim

of crime, and the level of worry they attribute to this event, based on the

assumed probability rather than actual experience (Gray et al. 2008). Perceived

risk here is measured as solely subjective probability (Jackson & Gouseti 2013),

omitting many of the complexities involved with individuals experiencing fear of

crime in their lives.

Consequentially, such methodologies can result in the overrepresentation of

actual rates of fear of crime (Farrall & Gadd 2004; Fattah & Sacco 1989; Yin

1980; Bernard et al. 1984) and neglecting the possible impact of the perceived

seriousness of crime and victimization (Warr & Stafford 1983; Jackson 2011) as

well as the sense of control over the event (Jackson & Gouseti 2013: 4). Further,

“these intensity measures may often collate not just everyday worries or fears,

but also some emotionally tinged ‘attitude’ towards risk (Jackson 2006) or

future-oriented anxiety (Sacco 2005; Gray et al. 2010) which they do not

account for in the interpretation of these results. Instead the findings are

presented as accurate representations of people’s everyday experiences, when

it is widely considered that they are not.

To attain a more accurate picture of fear of crime as experienced, questions

about the frequency and intensity of the experience were intrudiced in the

2003-4 sweep of the BCS (Farrall & Gadd 2004; Farrall et al. 2007; Gray et al.

2008; Gray et al. 2010). These new measures focus on instances of ‘worry’,

referring to concrete mental events of concern (Farrall & Gadd 2004) rather

than ‘anxiety’, which refers to a more diffuse mental state (Gray et al. 2010).

This movement towards measuring fear of crime as an event experienced in

everyday life rather than an merely an underlying attitude or anxiety shows a

shift in focus towards analysing fear of crime as an experiential event. However,

by asking whether in the last 12 months the respondent has felt worried about

being a victim of crime, and asking them to remember how many times this

happened and how worried were they in each of these occasions (Experience

and Expression in the Fear of Crime, 2003-2004), a number of original issues

with the CSEW/BCS remain unaddressed. Firstly, these new measurements still

do not get around “problems associated with self-reports, including errors in

recall, demand effect, or reluctance to disclose emotions” (Warr 2000). Issues

with recall have always been a part of quantitative survey research in the social

sciences (Loftus et al. 1992; Bernard et al. 1984). Specific to fear of crime, it is

important to note that memories of emotional experience from longer than

about two-weeks prior draw on semantic knowledge and general beliefs related

to the particular event, rather than the specifics of the event itself (Robinson &

Clore 2002b; Robinson & Clore 2002a; Gray et al. 2011). Therefore even though

such experience-based questions move closer to capturing the more expressive

dimensions of public insecurities about crime (Gray et al. 2011), it still does not

reflect fully the dynamic way in which this is experienced in every-day life.

Further, these surveys use respondents’ place of residence when assigning a

geography to fear of crime. Therefore we gain a static image of how safe people

perceive their own neighbourhoods, and lose information from those who may

not live there, but do travel to, or through, that area. However it has been

shown in past research that fear of crime varies based on many variables,

besides just who feels it (age, gender, etc.). Factors such as time of day,

familiarity with an area , and psychological distance (Jackson & Gouseti 2013) all

affect perceptions of safety. It has even been shown to vary with the purpose of

the journey, that is whether the person was carrying out a voluntary or

compulsory routine activity (Rengifo & Bolton 2012). Therefore these new

measures while shifting emphasis towards addressing fear of crime as

experienced in every-day life, still fall short of being able to capture its dynamic

nature. Even surveys that concentrate on travel, such as the TfL Attitudes to

Safety & Security Survey only cover segmented elements of the whole journey

environment, and do not cover people’s entire activity space door-to-door.

Aim of this paper

The project presented in this paper will attempt to address these shortcomings

and omissions in past fear of crime literature and by doing so complement and

build upon past research, to help provide a more holistic picture of fear of

crime, as experienced by people in their everyday lives. It aims to investigate

the feasibility of applying a novel methodology to fear of crime research in

order to gather data about people’s perceptions of safety in their entire activity

space, and locate them in place and time dynamically.

Purpose of this paper

The study presented here approaches fear of crime from the framework of “the

criminology of everyday life” (Garland 2001). This approach views the criminal

event as the endpoint of a decision process (which can be conscious or

subconscious), influenced by personal (e.g. readiness to commit crime), and

environmental (e.g. suitable target and lack of capable guardian) factors

(Brantingham & Brantingham 1993). Similarly, the fear of crime event can be

considered as something experienced by someone with a certain readiness to

experience fear of crime, which can depend on a number of factors which the

individual brings with them (e.g. age, gender, psychological distance, familiarity

with an area) as well as elements present in the environment (e.g. signal crimes,

graffiti). By collecting micro level spatial and temporal data on fear of crime as

experienced in people’s entire activity space, it would be possible to discern

whether fear of crime shows concentration at hotspots in place and time, like

crime. This knowledge can be used to design targeted, efficient and effective

situational interventions to enhance perceptions of safety in the built

environment. When such micro-level variation for crime data was identified, it

resulted in more effective targeted interventions, and exposed how previous,

inaccurate, measurement lead to inefficient and misguided interventions (e.g.

see Ratcliffe & McCullagh, 2001). Therefore, by discerning the feasibility of

measuring fear of crime as a dynamic event in the whole journey environment

with specific data on spatiotemporal variation, this paper can take the first step

towards identifying what are the circumstances that promote fear of crime, and

how situational interventions can be implemented to enhance perceptions of

security.

METHOD

Experience sampling

Experience sampling method (ESM) “captures the representation of experience

as it occurs, or close to its occurrence, within the context of a person’s everyday

life” (Christensen, Barrett, Bliss-moreau, Lebo, & Kaschub, 2003: 54). It offers a

method for researchers to seek information about the daily events and

experiences that make up people’s lives (Csikszentmihalyi & Larson 1987), and

provides a powerful way to understand psychological phenomena (Christensen

& Barrett 2003). The use of experience sampling emerged in the domain of

psychology, as a reaction to “a large body of research demonstrating the

inability of people to provide accurate retrospective information on their daily

behaviour and experience” (Csikszentmihalyi & Larson, 1987: 526). As in fear of

crime research, criticisms of questionnaires claim they tend to measure

people’s generalized knowledge or theories about their experiences, rather

than the episodic or experiential representations (Christensen et al. 2003).

Answers to such retrospective questionnaires, of which the CSEW/BCS is one,

tend to therefore reflect experiences that have been filtered and reconstructed.

These problems associated with using CSEW/BCS and other questionnaire

surveys measuring fear of crime can be addressed by the approach of using

ESM. As retrospective reports distort toward an ‘average’ experience, a more

effective way to “capture emotions, motivations, and cognitive processes is by

asking people to describe them at the moment they occur” (Hektner et al. 2007:

24). By asking people to report their subjective experiences with fear of crime

as they go about everyday tasks, we can build on their reports to identify trends

and patterns that may emerge from their experiences.

However, one downside to using ESM is that it has been seen as more costly

than questionnaires, and more taxing on respondents, since they have to

remember to carry the survey with them to complete it at the appropriate time,

and to eventually meet with the researcher in person or post their responses

once they have completed the survey. A solution to this limitation can be found

by applying this methodology to use with mobile phone applications.

ESM and mobile phones

Mobile applications offer a convenient platform to survey people about their

every-day activities as they are not an extra burden to carry around, but

something that people already have. Further, the use of mobile phones offers

added benefits; sensors such as GPS and an internal clock allow for the

collection of data such as geographic location and time of response without

having to explicitly ask the participant. This feature helps to reduce burden on

participants, as well as to ensure greater accuracy by eliminating the

opportunity for human error. An example of such a tool is a mobile phone

software called EpiCollect, which collects data “submitted by phone, together

with GPS data, to a common web database and can be displayed and analysed,

along with previously collected data” (Aanensen et al. 2009). EpiCollect has

been used for various studies from epidemiology (e.g. Olson et al., 2013) to

ecology (e.g. Madder et al., 2012).

Another example of a mobile application using experience sampling to collect

data on participants’ subjective perceptions is the Mappiness project.

Mappiness extends experience sampling “ to incorporate satellite (GPS) location

data” by using an app to collect a panel data set from volunteers (MacKerron

2012). In this way, MacKerron (2012) was able to collect information on

people’s happiness from a large sample of accurately geocoded responses, and

calculate particularly good indicators of environmental quality, and make

conclusions about momentary happiness and its environmental correlates.

By building a mobile application to make possible experience sampling of fear of

crime, and also the collection of spatial and temporal information about the

fear of crime events, many of the issues mentioned above with fear of crime

research methodologies can be addressed, and benefits similar to those that

came out of Mappiness’ methodology may be achieved.

FOCA

The fear of crime application (FOCA) was developed in Java programming

language, for use on Android mobile devices. It was written and tested by the

author and is not based on code from any other mobile application. It was

created using the Android Software Development Kit in the Integrated

development environment for Java, Eclipse. It was designed to be simple,

efficient, usable, and to be able to call built-in applications, such as the GPS

sensor, using intents, in order to fully make use of the capabilities of the mobile

phones. Learning how to do this eliminates the need for using separate GPS

trackers and paper surveys or diaries that users may forget at home or find

taxing to wear or use.

The basic functions of the application are to allow individuals to submit reports

of fear of crime events they experience and collect spatial & temporal data as

well as demographic information. The following will outline and illustrate how

this works:

1. Participants download the application from an app store onto their own

mobile devices.

2. The first time the app is launched, users fill out a quick demographic

survey to collect information such as age, gender, etc. This only needs to

be completed once, and is linked to all submitted reports.

3. Participants are sent reminders or “ping”s to complete the questionnaire.

This can be set to remind them at certain times (e.g. during morning peak

travel times and evening peak travel times) or in certain locations (e.g.

when they come within 50 meters of Camden Town Station), and

corresponds to a signal-contingent protocol of the experience sampling

method. As early as 1985, ‘beeper technology’ was being used to signal

to participants a request for them to record their experiences while going

about their days (Pervin 1985). This method is most suited for projects

that measure on-going behaviours, susceptible to retrospective memory

bias and to cognitive or emotional regulation, such as fear of crime. To

carry out this type of experience sampling, there are traditionally two

required instruments; one is a signalling device that emits the ping

according to a pre-determined schedule, and the other is an experience-

sampling form (ESF) (e.g. a short answer questionnaire) where the

participant records information on the momentary situation and

psychological state (Csikszentmihalyi & Larson 1987). This method is

useful to ensure that enough responses are submitted during the course

of this study. The innovation of using an app is to combine these two

tools.

4. When participants receive this ping they open the application and

complete the survey questionnaire (ESF) about their current state.

5. Participants can also report something un-prompted; if they experience a

fear of crime event at a non-ping time, they can still send a report about

it. This option incorporates an event-contingent protocol of experience

sampling. This method involves reporting an experience immediately or

closely following the event of interest (Christensen & Barrett 2003), in

this case, an experience of fear of crime. Since we are trying to measure

instances of fear of crime as they are experienced in every-day life, the

opportunity for people to report something as it happens is necessary.

6. An issue with asking people to report fear of crime events using

smartphones is the possible danger into which participants may put

themselves or their valuables. When experiencing a fear of crime event, a

participant may not be inclined to make their valuable mobile device

visible and vulnerable. To account for this, retrospective annotation was

chosen as a third option to send a report. In this version participants are

offered the option of finding the location of the event on the map, and

telling how many hours ago the incident happened, allowing for an

adjusted GPS and time-stamp value that is still accurate to the report.

7. Finally the data collected will tell us who sent the report (demographic

information), where it was sent from (GPS), when it was sent (date and

time), what the person felt (fear of crime), which option they used to

send the report (responding to a ping, voluntarily, or retrospectively), and

any additional questions put to participants.

8. See Figure 1 for a diagram

Figure 1: Methodology of the experience sampling application

FEASIBILITY STUDY

Sample characteristics

In order to determine the feasibility of this methodology when applied to the

study of fear of crime, a pilot study was conducted. We recruited people who

live or work in one particular part of London, to ensure that despite the small

sample size (only a pilot study) responses will not be scattered all over the city.

Of the 34 people who registered for the study, 7 dropped out, leaving a final

sample of 27 people. 52% were female and 48% male. Mean age was 27.32, the

youngest participant being 20, and the oldest 47. 22% were in full time

employment at the time, 11% in part time employment, and 67% were full time

students. Regarding ethnicity, the sample contained 1 Asian person, 1 Chinese

person, 3 Latin Americans, 2 people who did not answer this question, and the

majority were white British (41%) or white Other (30%).

Since this was a pilot study to ascertain the feasibility of using the novel

methodology proposed by this paper for the study of fear of crime, the

representativeness of the sample was not a concern. However it is important

that this data can be collected by the application, because for the real study, it

will be vital to ensure that the sample we use is representative of the

population at large.

Response characteristics

Besides demographic information, the pre-experiment survey also asked the

same retrospective question that is asked by the CSEW /BCS, for comparability.

Responses revealed that the pilot sample’s distribution of fear of crime

measured with this retrospective survey matched that of findings for the whole

country with the BCS; our sample’s fear of crime was not significantly different

from that of the overall population.

Regarding the experience sampling study, participants submitted a total of 467

data points over the course of two weeks. Number of reports per person varied

from a minimum of 2 and a maximum of 44. On average, people submitted, 16

reports (median was 13.5) with a standard deviation of 11.2 (Figure 2).

Figure 2: Distribution of reports sent by participants

Of all these responses, 121 were sent using retrospective annotation (about a

quarter of all responses). However, the majority of these were sent within the

hour (n = 94), where participants indicated that there was 0 hours delay

between the event and them reporting it. Out of the reports where a delay in

reporting was indicated, the shortest time between event and report was one

hour, and the longest 23 hours, with a mean delay of 5.6 hours (median = 3).

However it is interesting to note that the retrospective analysis option was not

chosen to report instances when participants were fairly or very worried about

crime. Considering that many of these reports were sent also with 0 hours

delay, it is possible that the use of option 3 with the map was utilised by

participants as a way to get around the problems or delays associated with the

phone’s GPS finding signal. Instead, reports of being fairly and very worried

about crime were submitted in almost equal parts using option 2, un-prompted

self-reports (53%,) and using option 1, response to signal-contingent protocol

pings (47%s).

Regarding sampling times of the day, Figure 3 shows the number of responses

sent per hour.

Figure 3: Fear of crime reports by hour of day

Adjusting for time in each segment, we received in total 17.3 reports for every

hour in the morning commute section, 20.3 for each hour during the day, 37.3

for each hour within the evening commute, and 14.9 reports per hour on

average for the night period. While evening commute is over-sampled with this

method, it is important to keep in mind that the night time frame contains

several hours where most participants are inactive (sleeping). Therefore this

method is considered effective for sampling different times of the day. Further,

sampling people at various times means that we get information about people

when they are out and about making journeys through London, and are not

restricted to collecting information about their place of residence.

The phone’s inbuilt GPS sensor also proved available to be utilised for collecting

the location of where people were feeling safe or unsafe during those two

weeks. Data about time and location of reports revealed multiple reviews of the

same locations, which will allow for the creation of aggregate ‘fear maps’.

Figure 4 shows the pilot data for Camden and Islington boroughs aggregated to

street level. This map reveals variation in fear of crime in very small areas, which

highlights the need for precise measurement over aggregate categories that

homogenise areas.

Figure 4: Fear of crime map for pilot study

DISCUSSION

Applicability of ESM via mobile app to FOC

The results from the pilot indicate that this method is indeed applicable to the

study of fear of crime. It gives an insight into people’s everyday experiences,

and is not restricted spatially to the places where they live or work. Instead it is

available to people as they move about their entire activity space, allowing

them to send reports about any location that they encounter at any time. The

flexibility in time also moves us away from the day-night dichotomy present in

current research using retrospective surveys. Narrowing down exactly at what

times certain areas are perceived to be unsafe can help strategically target

limited resources in order to enhance feelings of security in problem areas.

Further the low-level variation in fear of crime reported within these

preliminary results further underlines the need for a tool that can allow people

to report where they feel unsafe on a micro-level, in the whole journey

environment.

CONCLUSIONS

Of course this study is a pilot study and will be followed by the actual

experience sampling study of fear of crime, that will cover greater areas of

London, include a large and more representative sample, and include validation

of the technique and findings.

One possible issue is that online-based data collection is based on volunteer

sampling, rather than probability sampling (Lefever et al. 2007). However it is

still possible to use theoretical sampling, in an effort to provide a cross-section

of different socio-economic, age, and gender backgrounds (Innes 2004). To be

able to stratify our sample based on these variables, the methodology may

propose a large sample from which we can select strategically based on

demographic information collected by the pre-experiment questionnaire.

Next steps will also include validation of this study by carrying out interviews

with experience sampling study participants to ascertain whether the data

collected by FOCA accurately reflects their everyday experiences with fear of

crime.

Overall, this new insight into fear of crime can inform efficient, effective and

targeted interventions on the local level, and enhance perception of safety in

the environment, also enhancing walkability, people’s willingness to cycle, and

choose public modes of transport.

REFERENCES

Aanensen, D.M. et al., 2009. EpiCollect: linking smartphones to web applications for epidemiology, ecology and community data collection. PloS one, 4(9), p.e6968.

Bernard, H. et al., 1984. The problem of informant accuracy: The validity of retrospective data. Annual review of …, 13, pp.495–517.

Brantingham, P.L. & Brantingham, P.J., 1993. Environment, Routine and Situation: Towards a Pattern Theory of Crime. In R. V Clarke & M. Felson, eds. Routine Activity and Rational Choice. London: Transaction Publishers, pp. 259–294.

Brons, M. & Rietveld, P., 2001. Improving the Quality of the Door-to-Door Rail Journey : A Customer-Oriented Approach. , 35(1).

Christensen, T. & Barrett, L., 2003. A practical guide to experience-sampling procedures. Journal of Happiness …, 4, pp.53–78.

Christensen, T.C., Wood, J. V & Barrett, L.F., 2003. Remembering everyday experience through the prism of self-esteem. Personality & social psychology bulletin, 29(1), pp.51–62.

Csikszentmihalyi, M. & Larson, R., 1987. Validity and reliability of the experience-sampling method. The Journal of nervous and mental …, 175(9), pp.526–536.

Farrall, S. et al., 2007. QUESTIONING THE MEASUREMENT OF THE ’ FEAR OF CRIME ' Findings from a Major Methodological Study. , 37(4).

Farrall, S. & Gadd, D., 2004. Research Note: The Frequency of the Fear of Crime. British Journal of Criminology, 44, pp.127–132.

Fattah, E.A. & Sacco, V.F., 1989. Crime and Victimisation of the Elderly, New York: Springer-Verlag.

Garland, D., 2001. The Culture of Control: Crime and Social Order in Contemporary Society, Chicago: University of Chicago Press.

Garofalo, J., 1981. The fear of crime: Causes and consequences. The Journal of Criminal Law and Criminology (1973-), 72(3), pp.839–857.

Gray, D. et al., 2013. London’s Quality of Life Indicators 2012 Report,

Gray, E., Jackson, J. & Farrall, S., 2010. Feelings and Functions in the Fear of Crime: Applying a New Approach to Victimisation Insecurity. British Journal of Criminology, 51(1), pp.75–94.

Gray, E., Jackson, J. & Farrall, S., 2011. In search of the fear of crime: Using interdisciplinary insights to improve the conceptualisation and measurement of everyday insecurities. … Crime: Using Interdisciplinary Insights ….

Gray, E., Jackson, J. & Farrall, S., 2008. Reassessing the Fear of Crime. European Journal of Criminology, 5(3), pp.363–380.

Hale, C., 1996. Fear of Crime: A Review of the Literature. International Review of Victimology, 4(2), pp.79–150.

Hektner, J.M., Schmidt, J.A. & Csikszentmihalyi, M., 2007. Experience sampling method: Measuring the quality of everyday life, Thousand Oaks, California: Sage.

Innes, M., 2004. Signal crimes and signal disorders: notes on deviance as communicative action. The British journal of sociology, 55(3), pp.335–55.

Jackson, J., 2006. Introducing fear of crime to risk research. Risk analysis : an official publication of the Society for Risk Analysis, 26(1), pp.253–64.

Jackson, J., 2011. Revisiting Risk Sensitivity in the Fear of Crime. Journal of Research in Crime and Delinquency, 48(4), pp.513–537.

Jackson, J. & Gouseti, I., 2013. Fear of Crime and the Psychology of Risk. In Encyclopedia of Criminology and Criminal Justice. pp. 1–11.

Jansson, K., 2007. British Crime Survey-measuring crime for 25 years,

Kelly, C.E. et al., 2011. A comparison of three methods for assessing the walkability of the pedestrian environment. Journal of Transport Geography, 19(6), pp.1500–1508.

Lefever, S., Dal, M. & Matthíasdóttir, Á., 2007. Online data collection in academic research: advantages and limitations. British Journal of Educational Technology, 38(4), pp.574–582.

Liska, A., Sanchirico, A. & Reed, M., 1988. Fear of crime and constrained behavior specifying and estimating a reciprocal effects model. Social Forces, 66(3), pp.827–837.

Loftus, E.F. et al., 1992. memory and mismemory for health events. In Questions about questions: Inquiries into the cognitive bases of surveys. p. 109.

London Transport Committee, 2008. London Assembly Transport Committee Crime and disorder on London ’ s buses January 2008 The Transport Committee, London, England.

Loukaitou-Sideris, A. & Fink, C., 2009. Addressing Women’s Fear of Victimization in Transportation Settings: A Survey of U.S. Transit Agencies. Urban Affairs Review, 44(4), pp.554–587.

MacKerron, G., 2012. Mappiness. LSE. Available at: http://eprints.lse.ac.uk/36319/1/Landscape_MacKerron.pdf [Accessed October 1, 2012].

Madder, M. et al., 2012. e-Surveillance in animal health: use and evaluation of mobile tools. Parasitology, 139(14), pp.1831–42.

McDonald, N.C., Deakin, E. & Aalborg, A.E., 2010. Influence of the social environment on children’s school travel. Preventive medicine, 50 Suppl 1, pp.S65–8.

Mitra, R., Buliung, R.N. & Roorda, M.J., 2010. Built Environment and School Travel Mode Choice in Toronto, Canada. Transportation Research Record: Journal of the Transportation Research Board, 2156(-1), pp.150–159.

Olson, D.H. et al., 2013. Mapping the global emergence of Batrachochytrium dendrobatidis, the amphibian chytrid fungus. PloS one, 8(2), p.e56802.

Pervin, L., 1985. Personality: Current Controversies, Issues and Directions. Annual review of psychology, 36, pp.83–114.

Ratcliffe, J. & McCullagh, M., 2001. Chasing ghosts? Police perception of high crime areas. British Journal of Criminology, 41, pp.330–341.

Rengifo, A.F. & Bolton, A., 2012. Routine activities and fear of crime: Specifying individual-level mechanisms. European Journal of Criminology, 9(2), pp.99–119.

Robinson, M.D. & Clore, G.L., 2002a. Belief and feeling: Evidence for an accessibility model of emotional self-report. Psychological Bulletin, 128(6), pp.934–960.

Robinson, M.D. & Clore, G.L., 2002b. Episodic and semantic knowledge in emotional self-report: Evidence for two judgment processes. Journal of Personality and Social Psychology, 83(1), pp.198–215.

Ryan, S. & Frank, L., 2009. Pedestrian environments and transit ridership. Journal of public transportation, pp.39–57.

Sacco, V., 2005. When crime waves, Sage.

Transport for London, 2013. Safety and security annual report,

Transport for London, 2011. Travel in London Report 4, London, England.

Warr, M., 2000. Fear of crime in the United States: Avenues for research and policy. Criminal justice, 4(5), pp.451–489.

Warr, M. & Stafford, M., 1983. Fear of victimization: A look at the proximate causes. Social forces, 61(4), pp.1033–1043.

Yin, P.P., 1980. Fear of Crime Among the Elderly: Some Issues and Suggestions. Social Problems, 27(4), pp.493–500.