Criminology Research Council grant ; (5/78)
This study found that the identification of a valid measure of emotional response to frightening stimuli was proving to be fairly elusive. Because of ethical considerations, clinical experimentation tends to be confined to the lower animals. In non-laboratory situations it is difficult to arrange for immediate physiological examinations, and since behaviour change provides at best only part answer, there is growing support for a greater use of introspection reports by subjects.
The Australian Bureau of Statistics 1975 National Crime Victims Survey probed the level of the 'fear of crime' in this country by asking respondents how safe they felt if they were out alone at night walking in their neighbourhood. Publication of the result did not occur until 1979. This release generated a paper by Australian researchers which discussed the relationship of the answers to this question with those to other items in the survey, such as age, sex, locality, unemployment. Presentation of the paper at an international conference later that year was accompanied by a submission from an American scholar which contained well-argued criticisms of this type of survey question.
Apart from the hypothetical form of the question, one major problem was to establish a precise definition of the concept 'fear of crime' as distinct from 'fear of strangers' and from general anxiety. This proved to be difficult. For these reasons a more comprehensive program was adjourned in favour of concentrating on one specific aspect, that was, to seek initially to measure the emotion which is aroused by a subject's perception of a threat by a criminal of imminent physical injury. For this purpose information was gathered on the frequency, intensity and duration of such experiences.
Attention was drawn to four aspects of the subjects' daily programs: while they were at home, while they were in some other place, while travelling between the two, or while they were thinking of the safety of someone emotionally close. They were asked the nature of the threat they actually experienced, what they had done about it, and whom had they told.
To minimise errors due to interviewer bias, the same person conducted all interviews. For this reason the number of subjects had to be restricted to 25. These were selected from those categories who were considered most likely to be able to test the responsiveness of the method of recording high scores, i.e., the elderly, especially females living alone, and others who had already been victims of violent crime.
To reduce errors caused by memory lapses subjects were visited every fortnight. It was found that in the case of elderly subjects their recollection of events could be anchored around pension payments which occur every two weeks. Earlier interviews had shown that subjects were familiar with the principle of thermometer scales to indicate graduations whereas not all seemed to have sufficient command of words to discriminate degrees as well as verbal descriptions. Subjects were therefore asked to indicate the intensity of the feeling of 'fear' they had experienced by pointing to an appropriate spot on a thermometer scale.
Acceptance of the interviewer was readily forthcoming since he was already well known as a person active in the interests of victims of crime. However regular contact could not be maintained with four subjects for reasons associated with their victimizations.
Results at the end of a three month period showed that of the 21 remaining subjects only seven registered as having actually experienced 'fear of crime'. The other 14 consisted of elderly persons mostly living alone, a recent female rape victim, a middle-aged engineer who had been shot and his wife, a young sales girl who had been present at an armed robbery, and the parents of a young girl who had been raped and murdered.
Scoring of results presented difficulties because the three components, duration, intensity, and frequency, ought to be appropriately weighted if a composite total of any meaning is to be achieved. In the absence of any known way of doing this, a crude calculation of time in seconds, x degrees of intensity, x frequency, was used. Using this method, the results were dominated by those subjects who had been the victims of a violent crime and who feared repetition. In the case of those victims whose attacker had been imprisoned it seemed their 'fears' increased as a release date became nearer.
The report represents a statement of progress so far in the endeavour to produce a quantitative measure of the fear of crime. Further research is necessary on almost every stage. For example, little seems to be known about the capacity of individuals to accurately estimate the duration of an incident especially when they are under heavy stress, or to make subsequent judgments of it when asked at a later date. The shape of the 'fear' curve which seemed to be usually a sharp rise followed by a plateau, and then a long decline, presented problems of measurement. More work is required on the validity of the top ranges of the fear thermometer for it may not be able to provide sufficient discrimination to register high levels of emotion.
Even though this project did not fulfil all of its aims, it has had valuable consequences as it led the researcher to establish the South Australian Victims of Crime Service, an organisation which now has counterparts in a number of other jurisdictions.