Measuring Fear of Crime Challenges
A case study on the difficulty and the progress of measuring fear of crime
Fear of crime has been the centre of attention of scientific inquiry for nearly half a century. Scholars in this field have argued that fear of crime is directly correlated to experiences of deviant activity (Blakin,1979; Skogan & Maxfiend, 1981). Additionally, fear of crime was also reasoned in terms of the risk an individual is exposed to (Jackson, 2005). Others have focused on explaining fear of crime in terms of the influence of environmental factors (Wyant,2008).
Although there has been extensive research in terms of explaining fear of crime, contemporary scholars have recognised methodological limitations when analysing it (Farrall, 2005; Gray, Jackson and Farrall, 2008). For instance, fear of crime was measured as a dichotomous response (yes or no) to fear of crime. One example includes the questions that were asked in the British Crime Survey (BCS, now called the Crime Survey for England and Wales, CSEW): ‘How safe do you feel walking alone in this area after dark?’. Therefore, this paper will argue what limitations these ‘old’ measurements have when measuring fear of crime, as well as discuss how the ‘new’ methods of analysis contribute to the better understanding of fear of crime. Firstly, this essay will draw upon the limitations of measuring fear of crime. This includes discussing the problem with defining fear of crime. Secondly, this essay will explain what progress researchers in this field made in terms of measuring fear of crime. This includes how and why these advances improved researchers’ understanding of fear of crime, as well as how these ‘new’ methods improved the validity and reliability of measuring fear of crime. Lastly, this essay will identify some examples from the literature, in terms of how these ‘new’ measures have enabled researchers to conduct better studies of measuring fear of crime. This includes the female and male paradoxes, among others.
DEFINITION
Although many attempts have been made by scholars to define fear of crime, there remains a confusion as to what fear of crime means, due to the reason that there is not a universal definition in place. Therefore, because there is no general agreed definition of fear of crime upon researchers, questions differ from survey to survey, serving as a limitation to the way it is measured. For instance, Ditton et al., (1999) defined fear as “a term … encompassing a confusing variety of feelings, perspectives and risk-estimations and thus meaning different things to different people”. Ferraro and LaGrange (1987) defined fear of crime as “the negative emotional reaction generated by crime or symbols associated with crime”. Ferraro (1995 cited in Gray et al., 2011) also defined it as “emotional response of dread or anxiety to crime”. From these two definitions, aside from the fact that they are both quite outdated definitions of fear of crime, they are also very simplified definitions, as they only focus on the term fear of crime, being generated only from criminal activity. However, more contemporary definitions of fear of crime were provided by Gray et al., (2010). They defined fear of crime by making two significant distinctions in terms of fear of crime as everyday ‘worries’ and fear of crime as ‘anxiety’. The way they explained fear of crime, not only widens the spectrums in terms of what it is categorised as fear, but it also enables researchers to evaluate and conduct studies using more specific definitions, further allowing them to better understand this topic.
OLDER MEASUREMENTS OF FEAR OF CRIME
To begin with It is important to mention how the two way of measuring fear of crime had emerged from the theoretical point of view. Firstly, the old standard measurement of fear of crime, which started to emerge in the 1960s, developed from positivist ideas, by using ‘data-driven models of research’ (Lorenc et al., 2014, p12). Additionally, it has also used observational research in order to identify factors which causes fear (Lorenc et al., 2014). Secondly, the new measurements, were developed on a more critical basis, using the foundation of sociology and criminology. Furthermore, the new development of measurement of fear of crime was also determined by the feminist ideas, as it explores a variety of attitudes and anxieties, that people experience in their daily lives fear (Lorenc et al., 2014). Although these two measurement of fear of crime, were developed from two different theories, the way they measured crime also different. For instance, the old positivist measurement, focused primarily on quantitative surveys, whereas the new critical measurements focused primarily on qualitative and ethnographic studies (Lorenc et al., 2014). With regards to, more recent research on the measurement fear of crime, both qualitative and quantitative research were used, in order to obtain more accurate and reliable data (Gray et al.,).
The traditional single questions used to measure fear of crime is identified by researchers to be a formless measure, as it does not make any reference to real or specific offences. Therefore, it is still not certain if those question measured fear of crime, or measured any other approach (Ferraro & LaGrange,1987). For instance, on the British Crime Survey in 1995 questions such as ‘How fearful are you when walking alone at night? ‘were asked. Fearful is ambiguous in this question. The fear expressed by respondents may be about their physical or psychological circumstances, such as having a bad vision at night. In addition, the fear may also be expressed due to other factors, such as low street lightning. Therefore, it could be argued that those type of questions are mainly about fear of personal safety, rather than fear about crime (Radar,2004).
Questions about the validity of the instruments used to collect the data were also raised by various researchers (Gray et al., 1997 and Skogan & Maxfiend, 1981). Additionally, the concept of fear of crime is being amplified from the research conducted via surveys (Farrall et al., 1997: 665-666).
Furthermore, although some may argue that the traditional questions do measure fear of crime, those questions only measure general fear of crime, and not fear of certain type of offences (Gray et a.,). Therefore, participants might respond to questions based on offences they have experienced or offences they think about, rather than give specific, narrow answers. For instance, someone might think street robbery, while others might think of rape. It is important to specify the type of crime in questions about fear of crime, because the level of fear of crime that respondents experience, depend on different offences. One study shows that respondents expressed a higher level of fear of being a victim of burglary, when compared to the fear of being murdered (Warr and Stafford, 1983). It was also found that women’s main reason of worry was being a victim of sexual attacks (Ferraro and Jackson, 1995).
Nevertheless, the dichotomous single questions provide only two answers for respondents to choose from: ‘yes’ or ‘no’. Therefore, it is difficult, if not impossible to precisely pinpoint the exact fears of crime experiences by the respondents. Additionally, because these questions measure broad areas, it is difficult to generalise different studies (Lane et al., 2014). For example, there is extensive research to show that females express a higher level of worry when are alone in the night, when compared to men (Ferraro, 1995). However, it can be argued that, the only reason women may express higher levels of fear is only because they are alone in the night, and not because females generally express a higher level of fear of crime.
There is also a problem with perceived risk and fear of crime, in the formless measurements. For instance, LaGrange and Ferraro (1989) made a clear district ion between fear of crime and perceived risk. However, before this distinction, the surveys made no differentiation between these two concepts. Researchers who wanted to measure perceived risk of crime, used the same form of measurement used for fear of crime (La Grange and Ferraro, 1989). Also others have tried to measure fear of crime, as well as perceived risk of victimisation but they failed to distinguished between the two (Ferraro and Jackson, 1995). This could cause problems with the validity of the answers, as respondents might respond to experience of perceived risk, as fear of crime (Ferraro and Jackson, 1995).
Various researcher also identified the problem with time. The earlier questions to measure fear of crime, does not include any time period; frequency (Farrall and Ditton, 1999). Due to this, some respondents might refer to their fear of crime, from an incident that happened years ago, while other might refer to an incident they witness recently (Nair, 1993). Problems with differentiating between frequency that fear of crime happens and the intensity of fear of crime also arises. These considerations are important because it is important to be able to make the distinguish between people who fear crime intensively and the frequency to which they experience those fears (Gray et al., 2008). Hough (2004) however opposes the view that these type of questions, require the need to know more about the frequency. He argues that researching the intensity of the fear is more important, than researching the frequency of the fear. He stated that: “When we talk about mental states such as anxiety or worry, we are concerned with intensity, not frequency. Leaving aside acute anxiety attacks, anxiety is not comprised of a series of events that can be located in space and time” (Hough, 2004, pp. 174).
Although some researcher opposes the fact that frequency should be acknowledged when measuring fear of crime, Farrall (1997 cited in Jackson,2008) found that when giving the respondents the opportunity to express their perceptions and feelings in detail (i.e. qualitative interview), showed that anxieties were happening rare. On the contrary, the old standard measurements revealed a more exaggerates image of the anxieties that respondents experience in their day to day life (Farrall, 1997 cited in Jackson, 2008).
Old standard measures used to also ask respondents for a summary of the intensity of their worries. Researchers argued that, it is difficult to put emotions into writing, especially when the traditional questions were not clear, but where rather general (Jackson, 2005). Additionally, there is a problem with over-estimating the fear of crime. Researchers found that traditional measurements of fear of crime tend to exaggerate the results of the respondents. The reasons why this fear was amplified in the surveys, was because respondents summarised their most vivid and threatening experience of fear of crime as well as overestimated the frequency of those threatening experiences (Jackson, 2005). One example is from the 1995 British Crime Survey. it was found that there was a strong link between fear of crime, the experience of crime (which is either direct or indirect) and offences such a vandalism and drug use (Hough, 1995). Additionally, it was also found that fear of vehicle crime and mugging has increased since 1982 (Hough, 1995). It is difficult to say if those questions were accurately measured, due to the type of questions asked.
NEWER MEASUREMENTS OF FEAR OF CRIME
Various researchers have since modified the old standard questions used in the surveys such as British Crime Survey (Gray et el., 2008), as well as researchers who wanted to investigate fear of crime in their studies (Warr and Stafford, 1983, Fisher and May, 2008). For instance, Fisher and May (2009). They conducted a study, and asked questions about fear of four particular offences, rather than asking about fear of crime in general. By this, Fisher and May (2009) minimized the changes of obtaining measurement error, as respondents only had to measure fear of crime in relation to those type of crime. Furthermore, this type of questions not only diminishes the chances of measurement error, but also improves validity and reliability of data obtained from those responses (Jackson, 2005). Lastly, this also makes the questions clear and unambiguous, unlike the standard old measures of fear of crime.
Other researchers such as Smith and Hill (1991) developed questions which have more specific context; such as: inside, outside, day, night, etc. These type of questions also adds to the validity and reliability of the questions, when compared to the non-specific questions. The scale of which the questions are answered was also modified in order to obtain more accurate and more unambiguous responses on fear of crime. For instance, the traditional questions used a two way answer (yes;no), whereas the new questions used the Likert-scale questions (Lane et al., 2014). For examples, researchers such as Fisher and Sloan (2003) used for their studies a 10-point scale (such as 1 being not afraid and 10 being very afraid). Gray et al., (2008) also used Liked scale (i.e. 1 being not worried and 5 being very worried).
The new measurements of fear of crime also include questions about frequency and intensity of the fear/worry/anxiety felt by the respondents. These new questions were not taken into account when the old standard questions were used. However, when the new questions started to be integrated into the fear of crime measurement questions about the frequency and intensity of fear of crime, it could be seen that fear of crime was not as frequent as one might think, as often respondents’ answers using the traditional survey was exaggerated due to the measurement error caused (Farrall et al., 2004). Nevertheless, the standards question, never asked the frequency of the fear, until the new measurements of fear of crime started to take control (Gray et al.; Jackson, 2005).
The newer critical measurements of fear of crime used in its questions terms such as worry, rather than fear of anxiety (Jackson, 2005). It is suggested that although it is difficult to predict how respondents understand these terms, terms such as fear or anxiety could have a negative impact, as well as ambiguous responses to questions. One reason for this could be because fear is understood, as a present respond to a stimulus, as well as anxiety. It is believed that these two words have a strong significance, which in turn could have a negative impact on people’s responses (i.e. exaggeration) (Jackson, 2005). However, the term worry could be a better word due to the fact that “it includes a mental state, a concern about potential danger … a chain of thoughts and images about an unpleasant and uncertain outcome” (Jackson, 2005, p. 301).
Although it is believed by researchers (Farrell, 1997; Gray et al, 2008; Jackson, 2005) that the newer forms of measuring fear of crime are more accurate, some researchers still argue that even these measures have its flaws. Due to these flaws, these measures can lack validity and reliability. Some people might be more influenced, than other by certain factors, which in turn can make them report a higher level of worry (Jackson, 2005). For instance, those who reported the highest level of worry, were most likely to live in the low crime areas, experience fewer victimisation levels, as well as know fewer people who experience crime (Jackson et al, 2008 cited in Farrall and Lee, 2008). Media is also believed to be an important factor, which influences people’s perception of fear of crime, as well as perceived risk. Also, knowing people who have been exposed to victimisation can also increase respondent’s fear of crime. (Jackson et al., 2008 cited in Farrall and Lee, 2008).
THE BENEFITS OF USING NEW MEASUREMENTS OF FEAR OF CRIME
However, although researchers believe these new forms of measurements can also attract ambiguous responses, these measurements, are the most advanced and accurate methods that exist today to measure fear of crime.
For instance, due to the advancements of measuring fear of crime, it allowed researchers to distinguish between experiential and expressive fear of crime (Jackson et al., 2009). The experiential being the former, whilst the expressive fear of crime being the latter. A research was conducted and it was found that on one hand lower income areas with more crimes, experienced more experiential fear of crime, whilst on the other hand it has was found the expressive far was correlated with the sympathy expressed for the victims (Jackson et. al, 2009). These conclusions were drawn using questions about experience of fear of crime, as well as the frequency of fear experience by the respondents (Jackson et al., 2009). Additionally, Jackson (2008 cited in Farrall and Lee, 2008) also identified functional fear. This is when individuals who are worried about crime, in order to protect themselves, take precautionary measures, which in turn it does not affect their daily life, but makes it better.
The new methods of measuring fear of crime were also beneficial in terms of distinguishing between the levels of fear of crime amongst men and women. For instance, it is well known about the fear paradox.
References
Balkin, S. (1979). Victimization Rates, Safety and Fear of Crime. Social Problems, 26(3), pp.343-358.
Ditton, J., Bannister, J., Gilchrist, E. And Farrall, S. (1999). Afraid or Angry? Recalibrating the ‘fear’ of Crime. International Review of Victimology, 6(2), pp.83-99.
Farrall, S. And Ditton, J. (1999). Improving the measurement of attitudinal responses: An example from a crime survey. International Journal of Social Research Methodology, 2(1), pp.55-68
Ferraro, K. And Grange, R. (1987). The Measurement of Fear of Crime. Sociological Inquiry, 57(1), pp.70-97.
Gray, E., Jackson, J. And Farrall, S. (2008). Reassessing the Fear of Crime. European Journal of Criminology, 5(3), pp.363-380.
Gray, E., Jackson, J. And 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.
Hough, M. (2004). Worry about crime: mental events or mental states?. International Journal of Social Research Methodology, 7(2), pp.173-176
Jackson, J. (2005). Validating New Measures of the Fear of Crime. International Journal of Social Research Methodology, 8(4), pp.297-315.
Jackson, P. And Ferraro, K. (1995). Fear of Crime: Interpreting Victimization Risk. Contemporary Sociology, 25(2), p.246.
Lagrange, r. And ferraro, k. (1989). Assessing age and gender differences in perceived risk and fear of crime. Criminology, 27(4), pp.697-720.
Lane, J., Rader, N., Henson, B. And May, D. (2014). Fear of crime in the United States: Causes, Consequences, and Contradictions. 1st ed. Durham, NC: Carolina Academic Press.
Nair, G., Ditton, J. And Phillips, S. (1993). Environmental Improvements and the fear of crime. British Journal of Criminology, 33(4), pp.555-561.
Rader, n. (2004). The threat of victimization: a theoretical reconceptualization of fear of crime. Sociological Spectrum, 24(6), pp.689-704.
Skogan, W. And Maxfield, M. (1981). Coping with crime. 1st ed. Beverly Hills: Sage Publications.
Warr, M. And Stafford, M. (1983). Fear of Victimization: A Look at the Proximate Causes. Social Forces, 61(4), pp.1033-1043
Wyant, B. (2008). Multilevel Impacts of Perceived Incivilities and Perceptions of Crime Risk on Fear of Crime: Isolating Endogenous Impacts. Journal of Research in Crime and Delinquency, 45(1), pp.39-64.
.