24/7 Customer support here

A critical examination of the British Crime Survey

   The primary purpose of the British Crime Survey is to estimate how many of the public in England and Wales are victims of selected types of crime over a year, describing the circumstances under which people become victims, and the consequences of crime for victims. Other aims include providing background information on fear of crime among the public and on public contact with the police. Respondents are asked a series of screening questions to establish whether they or their households had been victims of relevant crimes during the one-year reference period. They are then asked a series of very detailed questions about the incidents they reported. Basic descriptive background information on respondents and their household(s) is collected to allow analysis of the sorts of people who do and do not become victims. Information is also collected on other areas that are of intrinsic interest and that could usefully be related to experience as a victim, namely fear of crime, contact with the police, lifestyle, and self-reported offending. Being in survey form, the British Crime Survey is a qualitative method of research, where data is gathered, processed, and analysed accordingly, with graphs and charts being created in order to ascertain more effectively the implications and range of any findings.

   Within a survey, it is found that there are three primary method of questioning and answer-gathering. J. Michael Dean, Chief Investigator, National EMSC Data Analysis Resource Centre (2001) states these methods- ‘When creating your questions it is essential to consider the wording of the questions. The wording of a question may bias the results. Also, you should consider the format in which you want the responses. There are three main response formats: multiple choice, numeric open-end and text open-end.’

   The term ‘multiple choice’ is self explanatory; a number of set answers are given, and the participant is usually required to select one of those answers. There is also, when required, an option to select ‘do not know’ or ‘non applicable’. Numeric open end style surveys allows for the participant to answer any question posed with a figure, without limit, and sometimes with a limit. The problem here is that the range given can create charts and graphs difficult to analyse. (Haworth, 2000). With Text open end surveys, this applies even more so: the participant has total freedom, and is not restricted by set answers, making the potential range of answers vast. An earlier critique of the BCS was made by Lynn and Eliot (2000) in order to improve the structure and methods used in future BCS’s.

   The critique will comprise of six sections-
• The advantage of the methods used in the British crime survey
• The disadvantages of the methods used in the BCS
• The quality and reliability of the data produced
• Possible alternative research methods
• Values and ethics raised or overlooked
• Future improvements

   These have been chosen as they will demonstrate the ideas behind the methods used in the BCS. It is the opinion of this essay that suggesting the positives as well as the negatives plays a role in the creation of future BCS’s. There is a need to know what is working, in order to make preservations in the format, for future reference. However, more is learnt from mistakes made. Karl Popper (Centenary Conference 12-14 July 2002) notes that-

   ‘Not learning from our mistakes appears to be the prevalent situation, seeming to affect technology, professionals and their derivative organisations in equal proportions.’
Therefore, any criticisms made within this critique shall be constructive, and any methods for overcoming problems identified will be suggested where relevant.
The advantage of the methods used in the British crime survey

   The most recent BCS, conducted 2002/2003, is structured around the principle that change is a constant within our society. This is accommodated by the ‘move to annual sampling; a move to continuous interviewing throughout the calendar year; and a change in the reference period of the survey to the last 12 months rather than the last calendar year.’ Also, ‘…the main changes introduced in the 2001 BCS were an increase in the overall sample size to 40,000…’ The move to an annual cycle was ‘originally proposed as a means of increasing the frequency with which survey-based estimates could be published, while the increase in sample size was motivated primarily by a need to increase the precision of certain estimates, notably those of rates of violent crimes.’ (Lynn et al, 2000)

   The move from sampling every four years to annual sampling demonstrates a more gradual change in the criminal climate, thus providing more insight into what is causing crime. A fall in violent crime can perhaps be attributed to a fall in alcohol related offences, and so forth. Haworth (2000) tells us that-

   ‘Surveys are useful in describing the characteristics of a large population. No other method of observation can provide this general capability.’
This capacity to offer coverage of a wider population is reflected in the BCS, and is a characteristic required to demonstrate the number of different experiences the population has of crime. Consequently, many questions can be asked about a given topic giving considerable flexibility to the analysis.
• There is flexibility at the creation phase in deciding how the questions will be administered: as face-to-face interviews, by telephone, as group administered written or oral survey, or by electronic means.
• Standardized questions make measurement more precise by enforcing uniform definitions upon the participants.
• Standardization ensures that similar data can be collected from groups then interpreted comparatively (between-group study).
• Usually, high reliability is easy to obtain by presenting all subjects with a standardized stimulus, observer subjectivity is greatly eliminated.

   The disadvantages of the methods used in the BCS
There are a number of disadvantages in the methods used in the BCS. Haworth notes that ‘very large samples are feasible, making the results statistically significant even when analyzing multiple variables’, a suggestion followed by the BCS. However, it is the opinion of this essay that a participant may feel overwhelmed by the number of questions asked, thus rushing and offering inaccurate answers, making some data incorrect. The lengthening of the survey period, at the recommendation of Lynn and Eliot, goes some way to solving this problem. Trochim (2000), in listing the disadvantages of the interview method of surveying, argues that-
• Staff & Facilities Needs are high
• The Respondent does not have time to formulate answers
• There are difficulties with long response categories

   The opinion of Lynn and Eliot (2000), that ‘larger samples result in survey estimates having smaller variance (smaller standard errors)’ has been noted, and should be accepted.

   Lynn and Eliot note that ‘Retention of the sample of selected respondents would maximise the precision gains’. This, to an extent, is correct. Correlation over time is almost certain to be higher within persons than within an area. However, this option has some disadvantages. It is likely to be the most expensive option, owing to the need to trace movers and follow them up to new addresses, with Lynn and Eliot noting that ‘…the sample will become less geographically clustered than it was originally’. There may be a propensity to refuse, likely to be related to the experience of any earlier interview, and ‘this in turn is closely related to the respondent's victimisation experience(s)’ (Lynn and Eliot, 2000). Interviews last longer with victims of crime than with non-victims, and are considerably longer still for multiple victims (Hales and Stratford, 1999).

   Another possible negative aspect of this design is that it introduces the possibility of ‘conditioning effects’. In other words, the experience of being interviewed previously on the BCS could affect a sample member's responses on a subsequent occasion. This would have to be carefully evaluated. In addition, there is a risk of the sample becoming ‘out-of-date.’ (Lynn, Eliot, 2000). In other words, at the second survey period, the sample will be biased against recent immigrants and those under sixteen (who are too young to take part in the survey.)

   If a section of the population has a tendency to select different answers which differ dramatically from those of others, there may be an effect on the variance of estimates. In some circumstances, the probabilities can be controlled in a way which reduces standard errors, but usually ‘variation in selection probabilities has the effect of increasing standard errors.’ (Haworth, 2000)

1  2

BACK TO CRIMINOLOGY