ºÚÁÏÍø

Today there are thousands of women who are subjected to whistling and to words, comments, or behavior of a sexual nature when they walk through public spaces, which are referred to as “street harassment.” Over time, this phenomenon has been referred to in various ways, ranging from “daily incivilities,” “daily harassment by strangers,” and “gender-based street harassment” to the current name of “sexual street harassment” (Chacón, 2019). The study of harassment against women is relatively recent, since it is approached using feminist theories from 1970 onward (Kelly, 1988); however, sexual street harassment in a specific way has been studied very little (Vera-Gray, 2016), mainly due to its low visibility and its widespread occurrence.

This phenomenon has been defined more and more precisely; in this regard the Chilean Observatory against Street Harassment (2015) indicates that it is a form of genderbased violence of a non-reciprocal sexual nature directed mainly against women. Although it is true that men are also exposed to this type of harassment, they are exposed to it to a lesser extent (Macías, 2016). On the other hand, (Arancibia, 2015, cited in OCAC, 2015) defined sexual street harassment as any practice that: a) has a sexual connotation, b) is perpetrated by strangers, c) is carried out in public or semi-public spaces, d) is unidirectional, and e) has the potential to produce discomfort (anger, fear, disgust, helplessness, stress, etc.).

Likewise, the Chilean Observatory against Street Harassment (2015) proposes the following categories of harassment according to its practices:

- Verbal/non-verbal harassment: includes gestures, sounds, and comments of a sexual nature, obscene whistles and sounds, lewd looks and gestures, inappropriate comments about a woman’s body, comments alluding to sexual acts, and sexist insults.

- Physical harassment: includes touching, brushing, or grazing in a sexual way, in non-intimate or intimate parts of the body, and pressing the body against the other person.

- Serious harassment: intimidating approaches and cornering, chasing someone down on foot or by a means of transportation, exhibitionism, or public masturbation.

- Audiovisual recording: taking photographs or recording of a person or parts of their body without their knowledge.

With the most precise conceptualization of the phenomenon, and given the magnitude and impact it has on the lives of women, sexual street harassment in the Americas is going from being invisible and accepted as natural to being conceived as a psychosocial problem; however, its measurement has been scarcely developed. Until 1999, a uniform scale had not been constructed to adequately measure the severity and extent of the experience of sexual street harassment. Lenton et al. (1999) and MacMillan et al. (2000) included three to five questions on street harassment in their national surveys on violence against women. For their part, Fairchild and Rudman (2008) modified the sexual experiences questionnaire (Fitzgeral, 1990) to measure the frequency with which university women experienced 9 types of harassment from strangers. Subsequently, Lord (2009) constructed a 7-item scale where he also asked about emotional reactions to harassment.

Seeking to broaden the understanding and complexity of the phenomenon, Sullivan (2011) created the street harassment scale that measures the frequency of a variety of specific experiences of sexual harassment, finding two main factors. The first one is called “courteous/benevolent” because it includes asking for a phone number or blowing kisses in a romantic way. The second is called threatening/hostile because it involves touching or following. The author shows an excellent reliability and a strong factorial structure although he does not report their values and the scale is not published. Later, Guillén (2014) published his doctoral thesis where he develops an instrument on experiences of victimization of sexual street harassment with two parts. In the quantitative part, thirteen situations are presented where the participants mark how frequently they have experienced such situations. This part of the questionnaire has a reliability of 0.90. The qualitative part examines the emotions, reactions, and opinions that the participants have regarding the thirteen situations of sexual street harassment mentioned in the first part. In that same year, the Observatory against Street Harassment (2014) launched at the national level a questionnaire of 25 multiple-choice items combined with open responses, to describe the experiences of Chilean women in regard to sexual street harassment.

The psychometric properties of the questionnaire are not reported. Later, Billi (2015) complemented this questionnaire by adding items and sections. In the first section, he mentions twelve prompts that refer to situations of sexual street harassment where people must indicate their degree of agreement. In the second section, twelve situations are presented where they are asked to mark how frequently they have experienced these situations and to indicate who has carried them out and the approximate age of those perpetrators. In the third section, the person is asked to think about the most shocking experience of sexual street harassment that they have experienced and to mark which one it was out of a list of twelve situations, indicate their age at the time of the experience, who carried it out, the approximate age of the perpetrator, at what time of day it occurred, where it occurred, the impact of the experience, and the emotion felt at the time. In the fourth section, nine situations are presented for the participants to mark how punishable these events are. Finally, in the fifth section, the sociodemographic data of the participants are requested. The psychometric properties of this questionnaire are not reported.

Finally, in Peru, the Sexual Street Harassment Scale prepared by Cruz-Pasos in 2017 was published, which has a Cronbach’s alpha score of α = 0.898 for sexual street harassment; α = 0.876 for expressive harassment; α = 0.766 for exhibitionism; α = 801 for chasing; α = 0.834 for physical harassment; and α = 0.648 for verbal harassment, which are acceptable coefficients of reliability. Likewise, the reliability was determined by the Spearman Brown split-half method, whereby reliability scores of α = 0.879 was obtained.

Since many of the scales or questionnaires that have been constructed to measure sexual street harassment are unpublished, and since they all belong to cultures other than Mexico’s, the objective of this work was to build and establish the evidence of validity and reliability of the Mexican Sexual Street Harassment Questionnaire (CMASC).

PARTICIPANTS

To explore the internal consistency of the Mexican Sexual Street Harassment Questionnaire, we worked with 250 volunteers from different regions of Mexico who answered the online questionnaire, whose ages ranged from 15 to 30 years with an average of 22.6 (SD 3.1) years.

The sample to carry out the confirmatory factor analysis of the questionnaire was made up of 400 volunteers from different regions of Mexico who answered the questionnaire online, whose ages ranged from 15 to 30 years, with a mean of 20 (SD 2.1) years.

TOOLS

The Mexican Sexual Street Harassment Questionnaire (CMASC) was specifically designed in Spanish for the present study, taking into account the stages for the development of the measurement instruments (Muñiz & Fonseca-Pedrero, 2008). Regarding the content of the items, once sexual street harassment was operationalized, the published literature was used to deepen the description of the phenomenon and thus design the prompts.

In terms of format, the self-reporting questionnaire was selected following the recommendations of Hamby and Finkhelor (2001) who recommended this because harassment is a sensitive subject and victims are reluctant to speak on their own behalf; additionally, due to its simplicity and ease of use, self-reporting allows a large number of people to be evaluated simultaneously, is low-cost and unobtrusive, may be done anonymously, and is easy to implement online.

Two procedures were used to verify the validity of the contents and the applicability of the first version: (a) 3 experts were asked to participate and analyze the effectiveness of the items according to theoretical dimensions and to eliminate confusion; (b) a pilot study was done using a sample of 50 women in order to eliminate prompts that were unclear.

The final version was a 35-prompt questionnaire divided into two parts. The first part consisted of 20 items with five response options ranging from “It has never happened to me” to “It has happened to me very frequently in the last year,” and it explores exposure to various forms of sexual street harassment. The second part consisted of 15 dichotomous items that explore reactions to sexual street harassment.

The researchers, after obtaining approval from the ethics commission of Autonomous University of the State of Mexico (ref. 6290/2020CIF), proceeded to publish the questionnaire online using Google Forms in which in the first part the consent of the participant is requested and the confidentiality of the data is ensured. Once the response data had been captured, two samples were sectioned to precede with the exploratory and confirmatory factor analyses.

The KMO test is equal to 0.912, and Bartlett’s sphericity test is statistically significant χ2 = 2291.485 (153), p < 0.001, which indicates that the data are appropriate for the exploratory factor analysis.

After performing the exploratory factor analysis, seventeen items were grouped into three factors that together explain 58.53% of the variance with a total Alpha of 0.92. The first factor was named “non-physical sexual harassment” and consists of 8 items with an Alpha of 0.89. The second factor was named “physical sexual harassment” and consists of 7 items with an Alpha of 0.81. And the third factor is called “explicit sexual harassment” and consists of two items with an Alpha of 0.73 (Table 1).

Table 1. Summary of exploratory factor analysis.

In order to contrast the exploratory factor structure, a confirmatory factor analysis was carried out using the program AMOS 23. The maximum likelihood method was used to calculate the goodness of fit parameters. The goodness of fit indices that were used are those mentioned by Bentler, (1990), Bentler and Bonnet, (1980), and Levy and Varela (2008), which can be seen in Table 2. Figure 1 shows the model with standardized scores.

Table 2. Expected fit indices for a structural equations model and indices obtained by confirmatory factor analysis.

Figure 1: Confirmatory factor analysis model.

For the second part, which explores reactions to sexual street harassment, since they are dichotomous items, their factorial structure was explored and confirmed through the program Factor Analysis with which tetrachoric correlations can be carried out, which are indicated for dichotomous scales (Hoffmann et. al., 2013).

The KMO test is equal to 0.892, and Bartlett’s sphericity test is statistically significant χ2 = 1590.9 (400), p < 0.001, which indicates that the data are appropriate for the exploratory factor analysis.

After performing the exploratory factor analysis, 14 items were grouped into three factors that together explain 47.30% of the variance. The first factor was named “negative reactions.” It consists of 6 items with an Orion index of 0.96. The second factor was called “neutral and positive reactions” and consists of 4 items with an Orion index of 0.90. And the third factor was called “protective reactions” and consists of four items with an Orion index of 0.86 (Table 3).

Table 3. Summary of the exploratory factor analysis for reactions to sexual street harassment

The confirmatory factor analysis using the robust method confirmed the factor structure found previously, with goodness of fit indices very appropriate to the model (Table 4).

Table 4. Expected fit indices for a structural equations model and indices obtained by confirmatory factor analysis

The results obtained reveal adequate psychometric properties of the items and evidence that support the reliability and validity of the scores given by the Mexican Sexual Street Harassment Questionnaire (CMASC). The analyses provide evidence of the document structure, with three sub-scales that are consistent with theoretical substantiations. Also worthy of mentioning is the questionnaire’s overall satisfactory internal consistency index and its sub-scales. The CMASC questionnaire shows advantages over other instruments used to measure sexual street harassment (Fitzgerald, 1998; Lenton et al., 1998; MacMillan etal., 2000; Fairchild and Rudman, 2008; Lord, 2009; Sullivan, 2011; Guillen, 2014; OCAC, 2014; Billi, 2015; and Cruz-Pasos, 2017), firstly because it is a questionnaire constructed ex profeso to explore the phenomenon, and secondly because it shows better internal consistency indices than other questionnaires.

The internal consistency was confirmed by a confirmatory factor analysis in which the values suggested by Hooper, Coughlan, & Mullen (2008) were reported to assess the goodness of fit of the questionnaire; Adjusted Goodness of Fit Statistic (AGFI) ≥ 0.90, Normed-Fit Index (NFI) ≥ 0.90, Root Mean Square Error of Approximation (RMSEA) ≤ 0.06 , Comparative Fit Index (CFI) ≥ 0.90, Goodness of Fit Index (GFI) ≥ 0.90, Root Mean Square Residual (RMR) ≤ 0.08, Non Normed Fit Index ≥ 0.90. The CMASC attains the 7 indices of goodness of fit, thus showing that it is a questionnaire with very good psychometric properties to be used in the Mexican population. Hu and Bentler (1999) suggest using these indices to evaluate the goodness of fit of a model since a Chi-square analysis can easily be statistically significant if the sample is large or the data are not normal, as is the case with the present questionnaire.

The first factor of the questionnaire called “non-physical sexual street harassment” comprises items that, due to their content, indicate the use of language and symbols to make sexual allusions to women. The second factor called “physical sexual street harassment” contains items that refer to friction, touching the intimate parts of women, and even street chases. Finally, the third factor called “explicit sexual street harassment” includes items on exhibitionist acts directed at women. In general, the first part of the questionnaire explores how a man, through whistling, insults, blocking someone’s path on the street, obscene gestures, looks, or words, propositions, touching, and exhibitionism, among others, manages to attract the woman’s attention by defining her as a sexual object and forcing her to interact with him (Di Leonardo, 1981; Martínez, 2018).

The second part of the questionnaire that refers to reactions to sexual street harassment after the exploratory factor analysis shows three factors whose evidence of reliability and validity is adequate. The first two factors refer to emotional reactions, and the third refers to instrumental reactions. The first factor, called “negative reactions,” includes prompts that refer to emotions of disgust, anger, and helplessness in the face of actions. The danger of experiencing these emotions lies in the fact that the numerous effects of street harassment, when viewed as a whole, operate as a form of socialization through which women learn, through constant objectification, sexualization, and evaluation, to associate feelings of shame, disempowerment, and victimization with their identity as women (Kearl, 2010).

On the other hand, the second factor shows emotions such as indifference, confusion, or pleasure, which shows that some women still consider experiences of harassment as something innocuous, normal, and even romantic (Spaccatini et al., 2019).

Finally, the third factor captures a series of actions that women carry out before acts of harassment with the intention of protecting themselves. They include changing routes or schedules, changing their manner of dress, or requesting company to go out. In this sense there is a restriction of movement due to fear of free movement. Falú (2011) indicates that, in women, social participation is hampered (physical activity, leisure, study, work, etc.), and that a lack of self-confidence can set in, along with isolation and perceiving the outside world as threatening.

One of the limitations of the present study was that it only covered young women under thirty years of age, taking into account that there is no evidence that sexual street harassment is exclusive to young women. An area of opportunity is to expand the sample to older ages, to explore whether harassment acquires other characteristics in older women.

In Mexico, in 2007, the “General Law on Women’s Access to a Life Free of Violence” was established, which defines violence against women as any action or omission based on gender that causes them psychological, physical, patrimonial, economic, or sexual harm or suffering, or death in both private and public spheres. However, it is not until 2018 that the revision of articles 16 and 17 begins to include verbal harassment, violence in public or semi-public spaces, and the generalized use of misogynistic expressions by strangers as crimes (Pacheco & Malak, 2019). Despite the above, a great disparity can be observed in the laws of the states of the republic in regard to the classification of this crime. This classification difficulty is largely due to the characteristics of sexual street harassment since, when it occurs in a public space, without a relationship between the victim and the aggressor and with a very short duration, it is usually disguised as seduction or courtship (Sastre, 2018). Thus, while in the workplace or places of learning there are policies and sanctions against sexual harassment, there are very few legal measures against sexual street harassment, and measures to protect the victims are almost non-existent (Baptist and Coburn, 2019). The implication for future study (and action) is that once the construct has been operationalized, the instrument will be used for national surveys in addition to legislation on the subject.