Annette Nierobisz
* The research was funded in part by grants from the Social Sciences and Humanities Council of Canada (#816-95-0037) and the General Research Grant fund in the Department of Sociology, University of Toronto. An earlier version of this paper was presented at the 1996 Sociologist Against Sexual Harassment annual meetings. Please direct all correspondence to Sandy Welsh, Department of Sociology, University of Toronto, 203 College Street, Toronto, ON M5T 1P9 or e-mail: welsh@chass.utoronto.ca
Abstract: Surveys documenting the prevalence of sexual harassment in Canada are hindered by four problems: the lack of mutually exclusive, behaviourally based survey items; the lack of exhaustive categories; inappropriate time frames for items; and a lack of context for these survey items. We compare the results from the 1983 Canadian Human Rights Commission study and the 1994 Violence Against Women survey to highlight these four common problems found in sexual harassment surveys. The Violence Against Women survey overcomes several of these problems and provides reliable estimates of the prevalence of sexual harassment. According to the Violence Against Women survey, 54 percent of Canadian women experience sexual harassment over their lifetime by known men in general and twenty-three percent experience sexual harassment by known men in workplace positions. This survey though, does appear to under-represent the amount of poisoned environment harassment experienced by Canadian women.
Resume: Les enquetes qui documentent la frequence du harcelement sexuel au Canada se heurtent a quatre sortes de problemes: le manque de categories de reponse mutuellement exclusives reflettant le comportement; le manque de categories epuisant le tout du comportement sous question; l'usage d'unites de temps qui ne sont pas appropriees aux categories de comportement; et le manque de contexte pour les categories de reponse. Nous comparons les resultats de l'enquete de la Commission canadienne des droits de l'homme de 1983 avec ceux de l'enquete <
Although research indicates that women have experienced sexual harassment through much of recorded history (Bularzik, 1978; Farley, 1979; Backhouse and Cohen, 1978), it was not until 1978 that the Canadian Human Rights Commission recognized sexual harassment to be a form of discrimination prohibitable under the Human Rights Act. This recognition sparked numerous surveys which explored the prevalence of harassment in Canada (e.g., Canadian Human Rights Commission, 1983; Cammaert, 1985; McDaniel and van Roosmalen, 1991; Phillips and Schneider, 1993). Yet, even with the proliferation of new studies, a clear and consistent picture of the prevalence of sexual harassment in Canadian workplaces refuses to emerge.
There is considerable variation in the estimated proportions of women reported to have experienced sexual harassment. Depending on the study, rates of sexual harassment in the United States and Canada range from 90 percent (Brooks and Perot, 1991) to 16 percent (Wizer, 1992). While the range of sexual harassment rates sometimes reflect differences in sampled populations (e.g., non-random samples, university versus workplace populations, see Gruber, 1990; 1992; Fitzgerald and Shullman, 1993; Arvey and Cavanaugh, 1995), the rates of sexual harassment in general population surveys vary from 30 percent (Loy and Stewart, 1988) to 49 percent (Canadian Human Rights Commission, 1983). We believe much of the variation in rates can be attributed to inconsistencies in measuring sexual harassment.
Our examination of the methodological issues for the study of sexual harassment ties into recent discussions about the development of sexual harassment as a social problem. We concur with those who point out how methodological scrutiny is used to delegitimize sexual harassment research (Sev'er, 1996; Gillespie and Leffler, 1987). Yet, while access to resources limits the ability of sexual harassment researchers to carry out large scale studies, we also believe there is a need to reflect on the state of knowledge in the area and to establish some methodological guidelines to ensure reliable and useful data.
To help improve studies of sexual harassment, we outline four factors researchers must consider when designing survey measures of sexual harassment. These factors are: the use of mutually exclusive, behaviourally based questions; the use of exhaustive categories; an appropriate time frame; and the context of the harassment. Before we discuss these factors in detail, we compare the discrepant conclusions produced by two landmark studies of sexual harassment in Canada, the 1981 Canadian Human Rights Commission study (CHRC) and the 1993 Violence Against Women Survey (VAWS), to highlight the role of measurement issues.
The Canadian Human Rights Commission Survey and The Violence Against Women Survey
The 1981 CHRC survey and the 1993 VAWS provide a unique opportunity to analyse the implications of sexual harassment survey item measurements for reported rates of harassment. Both the CHRC and VAW surveys represent crucial stages in the study of sexual harassment. The CHRC study was the first national survey of harassment in Canada. Along with the U.S. Merit Systems Protection Board study of federal workers, the CHRC study set a precedent for recording women's and men's sexual harassment experiences. For its part, the VAW survey represents a concerted effort on the part of government agencies (e.g., Health Canada, Justice, Statistics Canada) and other interested parties to collect reliable estimates of violence experienced by Canadian women (VAWS 1994). Although there are several studies conducted by independent researchers, such as the well-designed survey by Michael Smith and colleagues (see Gruber and Smith, 1995; Gruber et al., 1996), we are focusing on the CHRC and VAW surveys due to their historical and national prominence for documenting the extent of sexual harassment in Canada.
Even though the CHRC and VAW surveys were collected twelve years apart, there are enough commonalities in the sample characteristics to make the comparison between the two surveys useful. First, the CHRC survey, conducted in September 1981, was the first study of sexual harassment in Canada to use a national sample. This survey randomly selected 2004 Canadian men and women, aged 18 and older to be interviewed on a variety of issues including their experiences with "unwanted sexual attention." Excluded from the sample were residents of the Yukon, Northwest Territories and full-time residents of institutions such as psychiatric hospitals and army bases. Approximately 52 percent (1034) of the sample were female and forty-eight percent (970) were male. Twenty-four specific questions on sexual harassment were included in the survey. Our discussion relies on the responses of women in the sample because this makes the CHRC study comparable with the Violence Against Women survey and because sexual harassment overwhelmingly is a problem faced by women workers (USMSPB, 1981).
The second study we examine, The Violence Against Women Survey, was conducted from February 1993 to June 1993 by Statistics Canada. Interviewed by …
Комментариев нет:
Отправить комментарий