Download Comparing Victimization Surveys & Qualitative Methods in Measuring Deviance & Crime and more Study notes Sociology of Deviance in PDF only on Docsity!
SOC 3290 Deviance Lecture 18: Measuring Deviance & Crime 2: Surveys & Qualitative Methods
Given the problems we have noted with regard to official police statistics on crime, there have been a variety of attempts to get at its incidence and characteristics through other methods. Today we will review and critique two of these: (1) the available data on victimization from victimization surveys; and (2) data from qualitative methods such as participant observation, intensive interviewing, etc. (1) Survey Research:
Self-report surveys have become a common methodological tool used by criminologists - and especially by victimologists - in the past few decades. Victimization surveys have been the preferred methodological tool of victimologists since the 1970's. These gather information directly from victims without the intermediary of the police. Generally, individuals, in a representative sample of the larger population, are sought out and questioned anonymously about their experiences of victimization, if any. Such victimization surveys essentially focus on types of crime where:(i) there is a direct and identifiable victim; (ii) a direct and potentially identifiable perpetrator; and (iii) those forms of criminal victimization for which some information is available. This, in effect, limits the focus to traditional categories of interpersonal crime, and avoids others such as, for example, corporate crime.
Victimization surveys arose for one simple reason: generally, a great deal more has been known about perpetrators of crime than about their victims. In the past, official crime statistics gave virtually no information on the victims of crime nor on the incidence of crimes not reported to the police. Because of this, little could be said about which people were more likely to be victimized by crime, or about how many people were actually victimized.
In Canada, the first attempt to solve this problem was the Canadian Urban Victimization Survey conducted in 1981. It looked at 7 major urban centers across Canada with a random sample of over 60,000 people. It found that there were more than 700,000 personal victimizations (i.e. sexual assault, robbery, assault, and theft of personal property), plus almost 900,000 household victimizations (i.e. break and enter, motor vehicle theft, household theft, and vandalism).
The survey found that the more serious the type of crime, the less likely for it to occur. Gender differences were notable. Women were 7 times more likely than men to be victims of sexual assault or personal theft. Men were almost twice as likely as women to be victims of robbery or assault. As for age, those under 25 had the highest rate of victimization in all categories of personal offences, which declined rapidly with age after this point. With regard to Income and victimization, with some qualifications, the higher the family income of urban residents, the more likely it was that they would experience some form of household victimization or personal theft. Finally, lifestyle was another important variable, with a strong positive relationship found between one’s number of nights spent outside the home and rates of
victimization.
Fear of crime was found to be a significant issue, but more so for those walking alone in their neighborhood at night. Women and the elderly were more likely to express fear in this regard (50% and 98% respectively), compared to 18% of men. For those who have been the victim of sexual assault, these numbers increased considerably, even during the daytime - even though the incidence of sexual assault was relatively low compared to other offences.
The survey found that fewer than 42% of crimes were reported to the police, indicating that many more Canadians were victimized than official crime statistics would suggest. The most likely crime to be reported was theft or attempted theft of a motor vehicle (70%); the least likely was theft of personal property (29%). Women were found to have a higher reporting rate than males for sexual assault, robbery, and assault, and that those 65 and older were more likely to report incidents than younger victims.
The most common reasons given for failure to report an offence were that the crime was “too minor” (66%), that police could do nothing about it anyway (61%), and that it was too inconvenient/they didn’t want to take the time (24%). However, when broken down by offence category, the reasons for non-reporting by sexual assault victims varied in some important respects. Two thirds of women who had been sexually assaulted did not report the crime to the police. The most common reason was that police could do nothing about it (52%), but this was closely followed by 43% who cited concern about the attitude of the police or courts towards this type of crime (compared to a mere 8% of all victims of crime). In addition, fear of revenge is common among victims of sexual assault (33%), and female victims of assault generally (21%).
Finally, the data revealed that victims were most likely to report crimes which result in a significant financial loss, rather than those resulting in pain, injury and fear. Overall, it found that property crimes occurred more frequently than crimes of violence, that most of these resulted in low financial loss, and that victims themselves do not report them because they define the incidents as being too trivial to warrant police intervention. Crimes of violence were less frequent, and did not necessarily result in serious injuries, but there were serious issues raised about the consequences of making a report.
Since this pioneering survey, other surveys including information on victimization have been conducted in Canada. These include the General Social Survey (1993 and 1999) which indicated that, in 1993, at least 24% of Canadians were the victims of at least 1 crime in the previous year, the 1993 Violence Against Women survey, and the International Crime Victimization Survey. Taken together, these indicate the following about violent crime:
- People are fearful despite decreases in violent crime;
- Most violent crimes are not committed by strangers (except robberies);
- 20% of sexual assaults are committed by strangers;
- Most stranger violence does not occur in the home;
from ICVS respondents was that the incident was “not serious” (44%). This was followed by the response that “I solved it myself” (14%), and the opinion that the police couldn’t do anything (13%). The most common reason for reporting crimes was insurance reasons (35%), followed by crimes should be reported (34%), wanting the offender to get caught (27%), to recover property (27%), and to “stop it” (19%). Insurance reasons and recovering property were most commonly cited in property crimes, while the wanting the offender to get caught and “stopping it” figured much more prominently in crimes against the person.
Survey Research: Problems:
Now that we have done an overview of the survey data on crime and victimization, we will critique these data, squarely addressing the pros and cons of this methodology.
The chief advantages of surveys over official statistics are that: (i) respondents are asked about theoretically relevant concerns; and (ii) they weed out public decisions not to report or police decisions not to record, resulting in improved estimates of crime and victimization.
Yet, there are numerous problems with these methods. First, such studies depend upon victims knowing that they have been victimized and offenders knowing that they have committed a crime. For example, victims of fraud may not know that they have been cheated, and some obscure criminal acts may not be recognized as such by respondents unless brought to their attention. Moreover, what some individuals would perceive as abusive, others deny and excuse.
Secondly, standardized surveys, in which researchers attempt to ask all respondents the same questions in the same way, can sometimes be insensitive to cultural factors that affect the manner in which individuals interpret certain matters. Gomme (1993) gives the example of child abuse, which may be interpreted differently by cultures that consider a certain amount of "spanking" in the child's best interest, and not meaning the same thing as "hitting" the child. Fixed choice questionnaires may be useful, but they depend upon interpretive inferences and context-bound judgements about what is or isn't a meaningful answer to a pre-packaged question. Many respondents are unable to "get into" or hear questions in the same way as those who made them up. This is because respondents, when asked to choose an answer, are likely to be involved in an entirely different set of interpretive relevancies than researchers."
Third, respondents may not always be honest in their answers. Some may be reluctant to confess having done bad things or having experienced them. Some may be embarrassed or ashamed of having been the victim of some forms of crime, or fear revenge.
Fourth, the accuracy of data may suffer due to the faulty memories of respondents. Respondents may either forget incidents, or "telescope" prior events forward in time to the period covered by the survey. Indeed, Gomme suggests that there may be a "class bias" in survey results due to the fact that educated people are more likely to recall events and describe them accurately. This means that lower-class persons and "disadvantaged" members of certain minority groups
may give artificially low estimates of both their victimization experiences and their criminal involvements.
Fifth, many subjects will seek to give socially desirable answers or please the researcher, tempering their views in light of their beliefs about what the interviewer wants to hear.
Sixth, inquiring about only some offences limits the accuracy of overall estimates of crime based on these surveys. According to Gomme, self-report surveys enumerate mostly trivial offences. Victimization surveys do not ask questions about respondents' experiences with consensual vice crimes. Involvement in "victimless" crimes such as drug use, gambling, and prostitution remain unmeasured. Other crimes regularly omitted from victimization studies include disturbing the peace and public drunkenness. Furthermore, since victimization studies confine themselves to individuals as respondents, they provide no estimates of crimes, such as vandalism and arson, that are most often suffered by organizations.
Seventh, how researchers choose respondents for inclusion in their samples may affect generalizability. Some may use nonrandom samples based on student rosters or urban households
- thereby ignoring the often important experiences of dropouts, street youth, small town and rural residents, and transients.
Finally, because "serious" crime is comparatively rare, researchers must draw very large samples in order to get at the entire range of offences. This can be very expensive and time- consuming.
Despite these objections, however, so long as one does not reify survey data as "the Truth," it can bring some suggestive descriptive information to bear on existing theoretical formulations in the "big picture." The images of deviance and crime provided from official sources and surveys of victims and criminals can be likened to aerial photographs. While such snapshots offer, from distant vantage points, informative glimpses of broad and general patterns, they unavoidably leave obscure much of the finer detail regarding the nature and processes of involvement in crime and victimization. However, one must be cautious. To color in the rough sketches provided by such methods other, inductive, and more qualitative methods are required.
(2) Qualitative Methods:
(i) Participant Observation:
Participant observation is basically a non-quantitative method that involves techniques for examining how individuals normally think and act in their natural environments. By watching, listening, asking questions, and carefully documenting interactions and behavior, those employing such techniques try to understand particular social worlds as they are understood by inhabitants. Major advantages of this method are the detail of description and depth of understanding it provides, and that, unlike the above methods, researchers are supposed to enter
(ii) Intensive Interviewing:
Intensive interviewing is another qualitative research method, that may be employed alone or in combination with other methods. These informal, unstructured interviews are often like inquisitive conversations. Researchers do not construct strictly standardized questions in advance, as in surveys, but modify lines of inquiry "on the spot" in response to fruitful items. As such interviews are relatively unstructured, they are more relaxed, natural, and may seem more akin to conversation than an interview.
There are a number of advantages to this method. First, there is less chance of the researcher and the respondent misunderstanding each other with this method compared to the structured interviewing utilized in survey research. Second, this method has the potential for providing more accurate responses than survey research on certain sensitive issues, due both to the greater opportunity to develop rapport and the possibility of carefully approaching the matter from a variety of directions. Third, this method enables question and response categories to be adjusted to fit the respondent's way of looking at the world, which reduces the problem of researchers imposing a set of irrelevant categories on the respondent. Fourth, this method generates more intimate and comprehensive pictures of individual respondents than survey research by probing for underlying motives and personal experiences that can be linked to specific attitudes and beliefs. Fifth, certain categories of the population can be reached more readily by this method than by either participant observation or standard survey research, especially when the group of interest is widely scattered. Finally, a great advantage of intensive interviewing, relative to many participant observation studies, is the speed with which in-depth data may be gathered. When this is coupled with the often cheaper cost of interview studies relative to survey research, such methods appear attractive indeed.
Inevitably, however, this method has its share of problems. First, there is the old problem of generalizability of conclusions due to small, nonrandom samples, and lack of standardization in asking and emphasizing questions in the interview procedure. Secondly, this makes it very hard to replicate an intensive interview study. Third, this method is highly vulnerable to interviewer bias, either through selective choice of questions, selective probing, leading questions, subtle cues - either verbal or nonverbal - as to the interviewer's stand, the desire of respondents to give socially desirable, though untruthful answers, and similar desires to please the interviewer. Fourth, the fact that it is usually only the researcher who goes over the tapes or interview transcripts makes it very unlikely that such biases will be noticed. Finally, there has been a general lack of standardization in data analysis procedures, which, given a single body of data, makes it possible for several analysts to come up with very different interpretations. Moreover, the traditionally heavy dependence on the judgements of one or two investigators in interpreting the data opens the door to the researcher's theoretical perspective and personal ideology significantly affecting reported findings.
In the end, intensive interviewing can be a useful and thrifty way to get a wealth of detailed information on deviant individuals and behavior. Yet, as in participant observation, the
seriousness of many of these problems largely depends on researchers' skills.
Conclusion:
The last 2 lectures have reviewed the relative strengths and weaknesses of utilizing official statistics, survey research, participant observation, and intensive interviewing for the sociological study of deviance. Now that this has been done in some detail, some concluding comments are in order.
First, it is plain that none of the above methods is perfect. Not only do each have strengths and weaknesses, but, in fact, the first two and the last two have different goals. Official statistics and survey research, traditionally associated with positivism, are more geared to looking at the macro-level, static descriptive associations between variables, and broad causal relationships. Participant observation and intensive interviewing, conversely, are more oriented towards the micro-level, ongoing social processes, and the meaning-centred goal of interpretive understanding.
Secondly, if one is to get the best possible understanding of the empirical world, one cannot simply take the side of one of these methods, use it and it alone, and attack the others as flawed. Only by combining the imperfect descriptive outlines and associations provided by official statistics and survey research, with the rich, but imperfectly gained detail of participant observation and intensive interviewing, and asking questions about discrepancies between them, can we approach a comprehensive account of deviance. Researchers disregarding this "logic of triangulation", severely limit their ultimate ability to understand the world."
Third, as pointed out by Jack Douglas, whatever methods are used by researchers, it is important to be critical of what information is provided, or what people say, due to the myriad problems of misinformation,^1 evasion, 2 lies, 3 fronts,^4 taken-for-granted meanings,^5 problematic meanings,^6 and self-deceptions.^7 It is vital to be critical in "checking out" the information received,^8 and Douglas suggests general strategies, 9 that, along with the "logic of triangulation," may go far in helping us achieve the most comprehensive accounts of deviance.
Finally, it is vital to be epistemologically humble. We can never know the "really real," as Kant pointed out two centuries ago. All methodologies "bracket" reality in different ways, and all result, even when utilized together, in socially constructed, partial accounts of an ultimate reality we can never really know. We can never achieve "the Truth" about deviance no matter what we do. It is our job, however, to carefully utilize and coordinate the available, albeit imperfect methods at our disposal to give the most comprehensive accounts possible.
