SEARCH

SEARCH BY CITATION

Keywords:

  • sexual harassment;
  • online culture;
  • scale development;
  • organizational deviance;
  • organizational behavior

Abstract

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography

The use of computer-mediated communication continues to increase dramatically in organizations, bringing with it new avenues for sexual harassment. Despite the recognition that cybersexual harassment is an important phenomenon and that the online behaviors of employees can adversely affect the organization, there is little research on the development of instruments to measure related constructs. In the present study, scales are created to measure perceptions of the online environment that are thought to precipitate the occurrence of harassment and intentions to engage in sexually harassing behavior. It is found that perceptions of the online environment as stimulating (where risk taking is exhilarating), and as an environment in which blatant prejudice is acceptable, are related to multiple forms of cybersexual harassment.

The study of sexual harassment (SH) in organizations has a past rooted in concrete concerns regarding individual antecedents of harassment and organizational consequences (Berdahl, 2007; Fitzgerald et al., 1997; Fitzgerald, Gelfand, & Drasgow, 1995; Krings & Facchin, 2009; Willness, Steel, & Lee, 2007). Traditional definitions of SH in this context are largely guided by the legal definition provided by the Equal Employment Opportunity Commission, which includes quid pro quo SH and acts creating a hostile work environment (1984). It has long been recognized that individual difference variables (Pryor, Giedd, & Williams, 1995; Pryor & Whalen, 1997) and situational characteristics (Dall'Ara & Maass, 1999; Fitzgerald, Swan, & Fischer, 1995; Miner-Rubino & Cortina, 2007) interact to predict the occurrence of these types of SH in organizations (Nelson, Halpert, & Cellar, 2007; Krings & Facchin, 2009; Willness et al., 2007). Situational variables in particular are the focus of much research that suggests that hostile behavior is more likely to occur in a masculine environment where abuse is easy, tolerated, and penalties are minimal (Dall'Ara & Maass, 1999). Unfortunately for organizations, the use of computer-mediated communication (CMC) may allow these characteristics to emerge. Indeed, as the use of CMC continues to increase dramatically in organizations, new avenues for SH are introduced (MacKinnon, 1997).

The unique atmosphere created by the online environment led theorists to question how women would be treated and to redefine SH in order to correspond with the emerging world of cyberspace. Cybersexual harassment (CSH) includes behavior that follows the traditional definition of SH, but involves CMC (Biber et al., 2002). Despite the recognition that CSH is an important phenomenon and that the online behaviors of employees can adversely affect organizations (Glover, 2002; Porter & Griffaton, 2003; Swink & Cameron, 2004), there is little research on the development of instruments to measure related constructs. The present study addresses this problem with the development of a scale designed to measure one's likelihood of engaging in CSH. In an attempt to understand the underpinnings of online behavior, a scale is also developed to assess perceptions of the online environment that should be directly related to one's likelihood to engage in CSH. In the following review, the theoretical framework that provides the foundation on which to begin scale development is identified, addressing what types of behavior constitute online harassment, and the nature of the online environment that may perpetuate such behavior.

Theoretical Framework

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography
Defining CSH

Traditional SH research suggests that individuals differentiate from two to five dimensions (Fitzgerald et al., 1995a; Fitzgerald & Hesson-McInnis, 1989; Fitzgerald & Ormerod, 1991; Fitzgerald & Shullman, 1993). The most salient aspects of harassment are gender harassment, unwanted attention, and sexual coercion (Fitzgerald & Shullman, 1993; Gelfand et al., 1995; O'Connell & Korabik, 2000; Willness et al., 2007). Gender harassment involves misogynist behaviors that are insulting, hostile, or degrading towards women. Unwanted sexual attention corresponds closely with the legal notion of creating a hostile work environment, and may involve behaviors such as sexual comments about dress, touching, and display of sexual materials. Finally, sexual coercion is similar to quid pro quo SH in which one is bribed or threatened to perform sexual acts in exchange for some job-related benefit.

Given the unique characteristics of CMC, women are in danger of new forms of harassment. The three dimensions of SH indicated above may take on new forms with the introduction of technology. Hadler and Jaishanker (2008) suggest that there are a variety of negative behaviors targeting women online. These include e-mail harassment, cyber-stalking, cyber defamation, hacking, morphing (editing pictures), e-mail spoofing (where the origin in misrepresented), cyber pornography, cyber sexual defamation, cyber flirting, and cyber bullying. Although little empirical research has been done juxtaposing online behaviors with traditional definitions of SH, it seems that these behaviors may be perceived as gender harassment, unwanted sexual attention, and/or sexual coercion. Many of these, especially e-mail harassment, pornography, sexual defamation, and flirting, seem similar to hostile work environment SH.

It is gender harassment, however, that is seen as most common in cyberspace (Barak, 2005). Online, gender harassment has been typified as active versus passive and verbal versus graphic. Active verbal gender harassment, for example, includes unwelcome, offensive messages purposely sent via e-mail, in chat rooms or forums. Active graphic gender harassment also fits this description, but includes pictures. Passive verbal gender harassment includes intentional messages posted to many potential receivers. This may include, for example, an offensive code name, offensive information included with personal details, or flaming (acting aggressive, argumentative, and power-oriented). Passive graphic gender harassment includes exposure to offensive pictures that are not actively sought out (e.g., forced pop-up windows or redirection to an offensive site). All of these behaviors may be perceived as unwanted sexual attention rather than gender harassment, or in the form of more explicit sexual coercion.

Although much research has been done regarding face-to-face SH, and much theoretical work has been published regarding the nature of CMC, no scale has been created to empirically examine the nature of online behaviors. Hence, it is unknown if similar factors are perceived online as in the real world (gender harassment, unwanted sexual attention, and sexual coercion), or if individuals recognize a difference between passive and active online behaviors, as suggested by the literature. To address the lack of a reliable and valid measure of online behavior, the CSH scale was developed. Questions included in scale development were written based upon the theoretical dimensions posited above. In this article, the psychometric properties of this scale, including reliability, construct validity, and factor structure will be reported.

Hypothesis 1: The CSH scale is reliable and composed of multiple factors distinguishing between active and passive and verbal and graphic.

Although no formal scale has been validated to examine online SH, there is some empirical research suggesting that perceptions of online behavior differ from perceptions of face-to-face behavior. One study, for example, found that behaviors such as misogyny, the use of sexist nicknames, and comments about dress were seen as more harassing online than face-to-face. Requests for company, however, were seen as more harassing face-to-face (Biber et al., 2002). Interestingly, a follow-up study found that participants felt the organization was less responsible for online SH relative to face-to-face harassment (Ritter & Doverspike, 2002).

The online environment

The uniqueness of CMC has led to a multitude of questions regarding the nature of interactions. On one hand, there was speculation that women and minorities would be able to use online space to free themselves of social status and stigma restraints (Herrmann, 2007; Maier, 1995; McCormick & Leonard, 1996; Siegal, Dubrovsky, Kiesler, & McGuire, 1986; Stone, 1993; Sullivan, 1999; Taylor, Kramarae, & Ebben, 1993). An analysis of the characteristics of the online environment (OE), however, suggests that gender stereotypes do not disappear in a virtual reality.

The Internet atmosphere, in fact, has proven to be ideal for the perpetuation and exaggeration of negative behaviors based on stereotypes. Multiple research studies have demonstrated that gender roles remain in effect in an online environment, prescribing the rules of acceptable behavior based upon one's gender, allowing for the harassment of women (Allen, 2000; Barak, 2005; Bell & de La Rue, 1995; Citron, 2009; Cooper, Safir, & Rosenmann, 2006; Guadagno et al., 2011; McGerty, 2000; Rodino, 1997; Soukup, 1999; Sussman & Tyson, 2000). One study found, for example, that men tend to monopolize the conversation in chat rooms (Newmeyer, 1988). Another study found that men communicated more frequently, and had longer postings than women (Sussman & Tyson, 2000). The researchers suggested that socialization of gender stereotypes and the resulting power differentials are so strong that they emerge regardless of discourse medium. A good amount of research has supported the claim that men are more likely to act in an aggressive, argumentative, and power oriented manner online, whereas women are more likely to observe (Herring, 1994; Herrmann, 2007; Huffaker & Calvert, 2005; McCormick et al., 1996; Rodino, 1997; Shea, 1995).

Observations such as these are directly related to the underlying culture (values, norms, and expectations) of the OE. The formative years after the Internet was introduced, when men tended to make up 80%-90% of cyberspace, laid the foundation in forming the online culture (Maier, 1995; McCormick et al., 1996; Sardar, 1999). As the initial members of the OE, men had a disproportionate effect in creating online behavioral values and norms (Kendall, 2000; Kramarae, 1995; MacKinnon, 1997), resulting in a culture of elevated openness, bravado, and hegemony. Hence, the Internet environment consists of a masculine culture in which abuse is both easy and allowed, factors that precipitate harassment (Dall'Ara & Maass, 1999; Fitzgerald et al., 1995a; Fitzgerald et al., 1995b). In an era of increasing gender equality in the workplace, the online environment is a space where men can reassert their greater social status in a stronger, more insistent fashion than in the current face-to-face environment (e.g., Berdahl, 2007).

Indeed, researchers have dubbed the Internet the “Triple A Engine,” suggesting the dangers of the Internet emerge because of access, affordability, and anonymity (Cooper et al., 2006). Others refer to the OE as the “Penta A Engine” (Barak, 2005), also including acceptability and aloneness. All of these characteristics combine with a sexually charged online environment (e.g., Cooper et al., 2006) to encourage SH. The availability and affordability of the online world are the first factors that together mean easy access for individuals of all proclivities. Anonymity serves to neutralize status differences, make individuals feel invisible, reduce personal accountability, and make escape easy (Huffaker & Calvert, 2005; Rodino, 1997). Acceptability is a factor related to the hegemonic culture of the OE. Blatant sexism, not acceptable in traditional settings, is allowed online (McCormick et al., 1996). For the harasser, antiminority behavior may even be rewarded as the Internet is one of the few safe spaces remaining to openly express prejudice or sexism (Barak, 2005). Finally, aloneness is a characteristic of cyberspace that can encourage harassment because it implies a lack of negative physical social cues, eye contact, existing social norms, and personal accountability, as well as ease of escape. The juxtaposition of these characteristics create a disinhibition effect where traditional rules of behavior do not apply. Hence, individuals allow their true personality and attitudes to emerge and engage in behavior otherwise seen as risky (Barak, 2005; Cooper et al., 2006; McCormick et al., 1996; Rodino, 1997; Siegal et al., 1986; Sullivan, 1999).

Although there is a great deal written about the OE, no scale has been created to empirically examine online properties in order to gauge how they influence behaviors online. To address the lack of a reliable and valid measure of the OE, the OE scale was developed. Questions included in scale development were written based upon the theoretical dimensions posited above, in particular, anonymity, acceptability, and aloneness (affordability and access were not included as they are less strictly perceptual variables; affordability can be gauged by the reduction in the cost of Internet services and personal computers, access by the increase in the numbers of such products/services). The remainder of this article will report on the psychometric properties (reliability and factor structure) of this scale. An initial examination of the construct validity of this measure will be conducted by exploring its relationship with CSH.

Hypothesis 2: The OE scale is reliable and composed of 3 factors: (1) anonymity, (2) acceptability, and (3) aloneness.

Perceptions of the OE should be related to the likelihood of engaging in CSH. Indeed, the characteristics of the OE seem to play a large role in the occurrence of CSH. A common model explaining the propensity to harass suggests that sexually harassing behavior is best understood by not only examining the characteristics of the potential harasser, but also the characteristics of the situation (Pryor et al., 1995; Pryor & Whalen, 1997). That is, individuals possessing the personal characteristics tending toward harassing behavior (e.g., sexism) will do so only if the situation allows such behavior. One study, for example, found that men are more likely to harass when they observe others harassing (Pryor et al., 1995). Previous research has demonstrated that women are most likely to be harassed when the situation tolerates it, when the setting is male dominated, when abuse is easy, and/or when poor remedies exist (Dall'Ara & Maass, 1999; Fitzgerald et al., 1995b). Organizations generally have sought to eliminate these circumstances in the physical workplace; however, all of these factors may exist in cyberspace.

The relationship between the OE and SH intentions can also be explained by the theory of planned behavior (TPB) (Ajzen, 2001; Ajzen & Fischbein, 1977). The TPB suggests that behavioral intentions are predicted by attitudes (affective and cognitive), subjective norms, and perceived behavioral control. The preceding discussion implies that both online norms and perceived behavioral control components may especially increase intent to engage in sexually harassing behavior online versus face-to-face. For example, online norms in a sexually charged environment deem such behavior as acceptable. Factors related to perceived behavioral control, including access, affordability, anonymity, and aloneness, further encourage such behavioral intentions. Although attitudes and subjective norms are often strongly related, much research suggests that the relative impact of attitudes versus norms and perceived behavioral control depends on the context. In strong situations, subjective norms may be the best predictor of behavioral intentions (Azjen, 2001). Hence, the distinguishing characteristics of the OE may be very likely to increase intentions to engage in CSH. This hypothesis will serve as an initial test of the construct validity of the OE scale.

Hypothesis 3: Perceptions of the OE are positively related to the likelihood of CSH.

Method

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography

Procedure

The basic steps followed in the present study included item generation, scale development, and scale evaluation (Hinkin, 1995). Separate samples were used for item generation and scale development versus scale evaluation. In all phases, scales were distributed during class time by a female research assistant and collected when completed. Participants were assured that their responses were completely confidential, and asked to please consider the questions carefully and to answer as honestly as possible. When participants turned in completed surveys, the survey was coded with a participant number, and the consent form was stored separately from the completed questionnaire.

Phase 1: Item Development

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography

Participants and Materials

Participants for the item development phase were 154 graduate and undergraduate students who were currently attending a mid-sized Southeastern university. Three individuals were removed from the analysis as outliers or for failing the manipulation check, bringing the final sample size down to 151. The sample was 60% female, 79% Caucasian and 11% African American, with a mean age of 22. Participants in the sample had an average of 5.6 years of work experience, and spent an average of 2.65 hours on the Internet per day, with an average of 1.37 hours on the Internet during working hours. A student sample was deemed appropriate for a study of online behavior at work because all participants were currently employed or have been in the past, and because the age range incorporated in the current study has great experience with, and developed expectations of, the OE.

A deductive approach was used for item development. Based on the construct definitions and a thorough literature review, items for each scale were generated. For each scale, a content validity assessment was conducted by 10 subject matter experts (SMEs; 50% male) who were asked to classify items into predefined categories. Items not assigned to the proper category 80% or more of the time were removed from the scale.

CSH Scale

The initial pool of items for this scale was developed based on empirical research in this area, previously created measures adapted for the purpose of this study (e.g., Biber et al., 1998; Pryor et al., 1995; Pryor & Whalen, 1997), the theoretical literature, and a study of critical incidents based on court cases (e.g., the Chevron settlement, Tammy S. Blakey v. Continental Airlines, Inc., Coniglio v. City of Berwyn). Twenty-four items were written to include the four conceptual dimensions suggested in the literature: active verbal (e.g., ask a coworker for personal, nonwork-related information online), active graphic (e.g., send your coworkers erotic pictures to their e-mail), passive verbal (e.g., use an erotic term for a user id at work), and passive graphic (e.g., view pornographic pictures on your office computer. These items were reviewed for fit by SMEs resulting in the removal of two items, and ultimately seven items classified as active verbal, three items classified as active graphic, four items classified as passive verbal, six items classified as passive graphic, and two filler items included in an attempt to attenuate response pattern bias. The number of items in each dimension was deemed adequate as previous research has found acceptable reliability with as few as three items (Hinkin, 1995). The resulting item to response ratio is an acceptable 1:7, with the present study achieving the minimum sample size of at least 150 to conduct scale development analyses (Hinkin, 1995).

Based on the likelihood of SH scale (LSH) (Pryor et al., 1995; Pryor & Whalen, 1997), participants were asked to rate how likely they would be to perform each of several different behaviors on a scale from one to five where 1 = not at all likely and 5 = very likely. Participants were told, “Assume in each scenario that no matter what you choose to do, nothing bad would be likely to happen to you as a result of your action. Try to answer each question as honestly as you can. Your answers will be completely anonymous. No one will ever try to discover your identity, no matter what you say on the questionnaire.” As argued by Pryor et al. (1995), wording the directions in this manner ensures that responses tap the true psychology and intentions of the participant.

OE Scale

As this study was intended to examine the untested contentions of theoretical researchers, the items were developed largely based upon a thorough review of the theoretical literature. Twenty-six items were written based upon anonymity (e.g., I can treat people however I want online because they don't really know who I am), acceptability (e.g., I feel safe communicating things online that I could not communicate in the real world), and aloneness (e.g., It is easy to act however I want online because others are not physically present). The resulting item to response ratio is 1:6.

The initial items were reviewed by the SMEs for construct fit resulting in 10 anonymous items, 10 acceptability items, and 6 aloneness items. One reverse-scored item was included. Participants were asked to indicate the extent they agree or disagree to the statements on a scale from one to five where 1 = strongly disagree and 5 = strongly agree.

In the first stage of the present study, an exploratory approach was used in order to examine the nature of the constructs measured and eliminate poorly loading items. Hence, the scales were examined using an exploratory Principle Axis Factoring (PAF) analysis with an oblique rotation that allowed for correlated underlying factors.

Phase 1: Results and Discussion

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography
CSH Scale

The analysis with the 22 behavioral items resulted in six factors with eigenvalues over one. Items were deleted one by one, starting with one item that did not load on any factor, and progressing to items that highly cross loaded on many factors. Both empirical and theoretical concerns were taken into account before deleting items. The resulting scale included five factors with eigenvalues over one. Available fit indexes indicated acceptable fit after poorly fitting items were removed (χ2(df = 101, N = 150) = 217.19; RMSEA 90% Confidence Interval = .07−.10; SRMR = .05). Items were included with the factor that they loaded most highly on, and no cross loadings were over .36 (see Table 1). The first factor consisted of five items that clearly indicated behavior using e-mail. The second factor consisted of three items dealing with sending others graphic materials (active graphic). Although one item was defined as passive graphic, it is clear that individuals construed this behavior as active in that one is sending pictures or links to pictures to others (not just viewing). Item two was included with the active graphic scale (as was originally intended) even though it cross loaded on the e-mail factor. In further refinements of this scale item two was revised to read ‘send your coworkers erotic pictures’ (‘in their e-mail’ was deleted from the end of this item) so it would better fit the active graphic scale. The third factor consisted of five items dealing with passive behaviors, not intentionally directed at another individual (e.g., surfing pornography at work). The fourth factor consisted of three items constituting an active verbal dimension (e.g., asking coworkers for personal information online). This factor, although defined by active verbal items, is indicative of pressuring coworkers for a more personal relationship (similar to quid pro quo). Finally, the fifth factor was made up of four items, consistent in that the behaviors indicated by the items were purposeful action meant to offend (similar to gender harassment).

Table 1. Results of CSH Exploratory Factor Analysis
Questionnaire ItemOriginal CategoryIIIIIIIVV
  1. Note: AV = active verbal; AG = active graphic; PV = passive verbal; PG = passive graphic.

1.Send your coworkers dirty jokes to their e-mail.AV0.87−0.08−0.030.05−0.09
2.Send your coworkers erotic pictures to their e-mail.AG0.560.35−0.26−0.050.11
4.Use an erotic term for a user id at work or for an account you check at workPV0.200.170.35−0.180.00
5.Ask a coworker of the opposite sex for personal (non-work related) information online.AV0.14−0.02−0.040.340.19
6.Send an e-mail entitled “25 Reasons why Beer is Better than Women” to your coworkers.PV0.85−0.030.14−0.05−0.09
7.Post the following comment on a work-related bulletin board “am I the only one with the balls to stand up”.PV0.21−0.070.180.000.38
8.View pornographic pictures on your office computer.PG0.08−0.160.900.03−0.12
9.Download pornography and request colleagues to participate in the acts pictured.AG−0.09−0.100.10−0.120.82
11.Include links to sites containing pictures of pornography on your work web site.PG−0.080.540.300.28−0.18
12.Send links to erotic sites to your coworkers via e-mail.AG0.040.96−0.09−0.01−0.03
13.Send e-mails to your co-workers that joke that women are inferior to menAV0.580.000.160.070.05
14.Download pictures from pornography sites and paste them onto your work computer as wallpaperPG−0.090.170.010.010.57
15.Put pornographic pictures on your organization's websitePG−0.050.200.55−0.100.19
16.Send a coworker an e-mail making sexually-oriented comments about the way she's/he's dressed.AV0.08−0.07−0.160.210.66
17.Send sexually stereotyped jokes to coworkers via e-mail.AV0.680.00−0.030.050.10
19.Post comments in a work-related online forum about a coworker's appearance.PV−0.020.210.410.240.15
20.Send a co-worker multiple e-mails asking her/him to go out with you.AV0.14−0.02−0.020.71−0.05
21.Send a co-worker an e-mail pressuring him/her for sexual favors.AV−0.100.000.090.710.03
22.Surf pornographic websites at work.PG−0.05−0.100.860.040.10

Scale reliability was computed using coefficient alpha for each factor. Coefficient alpha for e-mail was .85, coefficient alpha for graphics was .77, coefficient alpha for passive behavior was .78, coefficient alpha for pressuring behavior was .53 (.65 if item 5 was deleted), and coefficient alpha for purposeful action was .70. Although a minimum reliability of .70 is expected (Nunnally, 1976), the pressure scale was retained at this point for further refinement and exploration in phase two.

Hypothesis one, pertaining to the factor structure and reliability of the CSH scale was partially supported. The hypothesis suggested that factors would emerge distinguishing between active and passive and verbal and graphic online modes of harassment. Instead, participants seemed to make more careful distinctions between online behavior, taking into account the medium and the severity of the behavior. For example, one factor clearly indicated that e-mail usage was perceived separately from other types of behavior that may be seen as harassing. This may be because e-mail is an everyday occurrence for most individuals in this age range, and is framed as a less severe type of action. Personal viewing versus sending pictures or providing links was also a more important distinction to participants in defining passive versus active than the number of potential receivers. Indeed, ‘providing links to many receivers’ loaded on a factor with other active graphic behavior.

Participants did distinguish between active (involving/affecting others directly) and passive (not affecting others directly) forms of pornographic behavior, first via a separate active graphic factor, and also via an active verbal factor that represented pressuring coworkers. This result is not contrary to previous research, as studies of face-to-face SH have found that individuals distinguish between sexual coercion or pressure and other forms of harassment (Fitzgerald et al., 1995a; Fitzgerald & Hesson-McInnis, 1989; Fitzgerald & Ormerod, 1991; Fitzgerald & Shullman, 1993). Finally, the fifth factor suggests that individuals distinguish between the severity of online behavior. This category may be analogous to other behaviors meant to purposely offend (e.g., cyberstalking), but is also similar to gender harassment.

Due to the sensitive nature of the questions included in this scale, it may be expected that clear factors would be difficult to achieve. This concern was addressed by including items that served as a manipulation check, and removing outliers from the analysis. In further refinements of this scale, the items are improved with social desirability and clarity in mind in an attempt to increase the reliability of the pressure and purposeful action subscales and improve the scale.

OE Scale

The analysis with the 26 behavioral items resulted in five factors with eigenvalues over one. Items were deleted one by one, following the strategy delineated above. Although item 18 loaded on two factors, it was kept in the analysis because it theoretically fit with the acceptability factor and loaded more strongly with this factor. The resulting scale included four factors with eigenvalues over one. Available fit indexes suggested acceptable fit after poorly fitting items were removed (χ 2(df = 87, N = 145) = 111.985; RMSEA 90% Confidence Interval = .01-.07; SRMR = .04). Items were included with the factor that they loaded most highly on, and no cross-loadings were over .32 (see Table 2). The four factors included aloneness, acceptability, anonymity, and a factor indicative of the potential affective stimulation experienced from the risk-taking culture online. Hence, the initial exploration of this scale resulted in the three expected factors, acceptability, anonymity, and aloneness, with one additional factor signifying the affective stimulation experienced from the masculine (aggressive and risk taking) online culture. Coefficient alpha for acceptability was .78, for anonymity was .83, for aloneness was .71, and for affective stimulation was .83.

Table 2. Results of OE Exploratory Factor Analysis
Questionnaire ItemOriginal CategoryIIIIIIIV
2. I can say anything I want online because getting away from the situation is easy.ALONE0.400.120.13−0.10
3. It is safe to express prejudice against others online.ACCEPT0.040.83−0.14−0.06
5. There is little personal accountability for one's actions online.ANON0.170.060.73−0.25
6. It is acceptable to demonstrate anti-minority behavior online.ACCEPT−0.140.650.03−0.04
8. There are little or no consequences for online behavior.ANON−0.03−0.160.890.03
9. I have support from others online no matter what I say.ACCEPT−0.090.020.210.57
11. It is acceptable to use stereotypes on the Internet.ACCEPT0.120.68−0.110.06
13. I can allow my true personality characteristics to emerge onlineANON−0.01−0.110.170.59
15. I can engage in behavior seen as risky in the outside world when I am online.ANON0.060.050.300.44
16. There are no repercussions for my behavior online.ANON−0.050.030.780.15
18. It is safe to express prejudice against women online.ACCEPT−0.060.520.350.10
19. There is something exhilarating about sharing views online that I could not openly express face-to-face.ALONE0.200.08−0.110.57
20. I feel good when I express how I really feel online.ACCEPT0.050.00−0.210.90
21. Blatant sexism, not acceptable in traditional settings, is allowed online.ACCEPT−0.010.340.320.18
22. It is easy to act however I want online because other people are not physically present.ALONE0.15−0.030.270.41
23. There are different norms for behavior online versus face-to-face.ALONE0.460.07−0.040.20
24. I have to choose my words more carefully in a face-to-face situation than I do online.ALONE0.70−0.020.060.07
25. There is a higher standard for face-to-face behavior relative to online behavior.ALONE0.76−0.04−0.050.02

Given these findings, hypothesis two, that the OE scale is reliable and composed of three factors, was confirmed with one additional factor. This result is interesting as it reflects the theoretical discussion. Theorists suggest that the OE is characterized by a masculine culture, emerging in this scale by items demonstrating the potential for feelings of exhilaration precipitated by engaging in risky types of behavior. Further, individuals clearly distinguished between the acceptability of blatant prejudice allowed online, the anonymity that reduces personal accountability, and the aloneness that makes it easy to ignore existing social norms.

In the second phase of this study, scale development continues using Confirmatory Factor Analysis (CFA) on a separate sample. Following this procedure, the relative fit of several plausible models can be tested following theoretical and parsimony concerns.

Phase 2: Scale Development

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography

Participants and Analysis

Participants for the scale development phase were 154 graduate and undergraduate students who were currently attending a midsize university in the southeastern US. The sample was 60% female, 84% Caucasian and 14% African American, with a mean age of 22. Participants in the sample had an average of 5.8 years of work experience, and spent an average of 2.28 hours on the Internet per day, with an average of 1.34 hours on the Internet during working hours. CFA was completed in order to explore the relationships found in phase one of the current study.

Phase 2: Results and Discussion

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography
CSH Scale

Goodness-of-fit statistics indicated that the five factor model found in phase one was acceptable (χ2(df = 117, N = 154) = 255.81; CFI = .86; RMSEA = .09; SRMR = .08). Four- and six-factor models did not differ significantly in a Chi-Square difference test (a one factor model was worse fitting). Item 15 was not included in this analysis, however, as it had a variance of zero. Using this sample, coefficient alpha for e-mail was .82, for graphics was .82, for passive behavior was .72, for pressuring behavior was .58, and purposeful action was .69.

OE Scale

Goodness-of-fit statistics indicated that the four-factor model found in phase one was acceptable (χ2(df = 109, N = 153) = 191.49; CFI = .90; RMSEA = .07; SRMR = .07). Three and five factor models did not differ significantly in a Chi-Square difference test (a one factor model was significantly worse). In order to achieve this fit, item two was removed from the analysis as it cross-loaded on multiple factors. Using this sample, coefficient alpha for aloneness was .81, acceptability was .72, anonymity was .80, and affective stimulation was .80.

For both the CSH and the OE scale, confirmatory analyses indicate a plausible model structure. Concerns may still emerge regarding the reliability of the CSH scale in particular as coefficient alpha for the pressuring behavior dimension was below acceptable levels. Nonetheless, it is encouraging that the factor structure of both scales was replicated during phase 2, with a separate sample, using confirmatory factor analysis.

Phase 3: Scale Evaluation

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography

Participants and Analysis

Participants for the scale evaluation phase were 181 graduate and undergraduate students who were currently attending a midsize university in the southeastern US. The sample was 32% female, 82% Caucasian, and 13% African American, with a mean age of 22. Participants in the sample had an average of 5.9 years of work experience, and spent an average of 2.00 hours on the Internet per day, with an average of 1.57 hours on the Internet during working hours.

The scale evaluation phase involves the formation of a nomological network with other variables. Hence, the testing of hypothesis three (that perceptions of the OE are positively related to the likelihood of CSH) is included in this section as well as supplemental analyses determining the extent these two scales are related to gender and gender stereotyping.

Measures

CSH Scale

Items for the scale in phase three were randomized to counteract possible order effects from the first two phases. In addition, the scales were presented in reverse order from the previous two samples (the OE scale was presented before the CSH scale). The original five-factors present in phase one included e-mail, active graphic, passive, pressure, and purposeful action. Fit analyses for a five factor scale were acceptable. Internal consistency reliability, however, continued to lag below the cut off for the pressure scale (α = .41) (see Table 3). This result may be attributed to fact that the items included in the pressure scale are all measuring pressure put on coworkers, but range dramatically in severity from asking a coworker for personal info online to pressuring a coworker for sexual favors. Due to the repeated low internal consistency for this factor, removal from the overall scale is recommended. Without these items, fit analyses for a four-factor scale were acceptable.

Table 3. Correlations and Descriptive Statistics for Scale Evaluation
 12345678910111213141516
  1. Note: N = 146-181. All correlations above .21 are significant at the .01 level. All correlations above.15 are significant at the .05 level. Coefficient alpha is reported on the diagonal.

1. Gender                
2. Age.10               
3. Hours.06−.02              
4. E-mail−.12−.08−.02             
5. Graphic−.15−.03.03.48            
6. Passive−.27−.06.00.55.68           
7. Pressure−.10−.09.06.47.86.61          
8. Purpose−.14−.09.00.57.46.60.51         
9. CSH−.18−.09.02.82.85.83.84.65        
10. Stim−.20−.07.01.57.81.85.80.68.88       
11. Accept−.23.01.02.24.31.43.35.32.38.44      
12. Anon−.14−.14.00.14.21.22.27.24.24.27.62     
13. Alone.05−.02.10.16.09.03.16.16.14.09.32.31    
14. OE−.19−.07.04.38.48.53.54.48.56.61.88.76.57   
15. PAQF.31.11.10−.14−.23−.24−.21−.11−.23−.27−.15.00.17−.10  
16. PAQM−.20.11.01.01.04.07.07.05.05.03.12.05.03.09.06 
Mean1.3222.21.971.971.21.361.231.521.441.311.82.223.361.983.843.93
SD0.473.141.60.930.510.590.550.520.520.480.790.931.010.560.540.52
OE Scale

In phase one, the initial factors for this scale included acceptability, anonymity, aloneness, and affective stimulation. Confirmatory factor analysis with a new sample continued to confirm this result. Coefficient alpha for each of these factors was also acceptable (see Table 3).

Gender Stereotyping

The Personal Attributes Questionnaire (PAQ) was used as a measure of gender stereotyping (Spence & Helmreich, 1978). The PAQ is interpreted as a measure of an individual's standing on masculine (instrumentality) and feminine (expressiveness) traits (Cota & Fekken, 1988). Coefficient alpha for this scale was .72 for masculinity and .79 for femininity.

Phase 3: Results and Discussion

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography

The final hypothesis in the present study acts as an initial test of the predictive validity of the OE scale, and as a test of the nomological network of the CSH scale. Hypothesis three suggests that perceptions of the OE are positively related to the likelihood of CSH. Descriptive statistics and correlations are presented in Table 3.

Correlations indicate that perceptions of both affective stimulation and the acceptability of prejudice are related to all facets of CSH behavioral intentions (i.e., e-mail, passive behaviors, pressure, active graphic behaviors, and purposeful action meant to offend) (see Table 3). Correlations are generally quite strong, and range from .31 (acceptability and active graphic behavior) to .85 (affective stimulation and passive behavior). The perceptions of anonymity factor in the OE scale is significantly related to all CSH dimensions, but the correlation with e-mail was only marginally significant (r = .14, α < .06). This may be due to the fact that e-mail tends to have specific targets, and tends to be less anonymous than general postings. The perceptions of aloneness dimension is the weakest predictor of CSH as it is significantly related to e-mail and purposeful action, but not graphic and passive behaviors. Given this evidence, hypothesis three, a significant positive relationship between perceptions of the OE and intent to engage in online sexually harassing behavior, is partially confirmed.

These results support past research suggesting that SH is explained, in part, by the characteristics of the situation (Pryor et al., 1995; Pryor & Whalen, 1997). This result appears to be true in the case of CSH as well as face-to-face harassment. Those individuals that perceived the online situation as more accepting of harassing behavior were more likely to report that they would engage in harassment online. This result is particularly troubling as theorists have suggested that SH is more likely to occur in environments sharing the characteristics of cyberspace (e.g., Dall'Ara & Maass, 1999). This argument is further supported by applying the theory of planned behavior. Those individuals perceiving online norms as stimulating and accepting may be particularly likely to engage in harassing behavior online.

Supplemental Analysis

Several additional analyses were performed as a first step toward establishing the nomological network of the OE and CSH scales. Based on previous literature linking masculine gender roles and the likelihood of engaging in harassing behavior (Pryor, 1987; Pryor, 1995), it was expected that CSH would be positively related to the masculine trait of instrumentality and negatively related to the feminine trait of expressiveness. The results of the present study, however, demonstrated that femininity was negatively related to e-mail, active graphic, and passive CSH (see Table 3). Masculinity was not significantly related to any CSH dimension, suggesting that the online construct is indeed a very different construct than face-to-face SH. Interestingly, ratings of the OE scale also followed the same pattern (see Table 3). Femininity was significantly negatively related to perceptions of the OE as stimulating, prejudice as acceptable, and as having different standards than face-to-face environments. This means that very expressive individuals (e.g., kind, understanding) may seek out environments that meet their needs online and avoid others. Clearly, individuals high in femininity would then be less likely to engage in CSH whether it be by e-mail or in other active or passive ways. This also means that less masculine individuals are just as likely to engage in CSH online as those with a more masculine gender ideology. This may be indicative of the power of the situation in the form of the OE perpetuating such behavior for a greater variety of individuals.

Additional analyses were done to examine gender differences in the CSH and OE scales given the ample evidence of gender differences in the perception and perpetration of face-to-face SH (e.g., Biber et al., 2002; Baird et al., 1995; Dougherty et al., 1996; Fitzgerald & Ormerod, 1991; Gohann & Thacker, 1996; Gutek, 1985; Gutek, Morasch, & Kohen, 1983; Hurt, Wiener, Russell, & Mannen, 1999; Keyton & Rhodes, 1999; Rotundo, Nguyen, & Sackett, 2001; Summers, 1996). As we have seen in the case of gender role orientation, correlates of CSH may differ from correlates of face-to-face SH, and exploring gender differences in perceptions of the OE and in the propensity to engage in harassing behaviors online can give us a better idea of how CSH compares to traditional (face-to-face) conceptualizations of this construct.

As expected, there were significant mean differences between genders in active graphic and passive CSH, with males more likely to engage in these types of harassing behaviors (t(143) = 2.245, p < .05; t(155) = 4.854, p < .01) (see Table 4). Interestingly, mean differences in e-mail and purposefully offensive behavior were only marginally significant (t(126) = 1.678, p < .10; t(167) = 1.773, p < .08), with males having the higher mean scores. This result may not be surprising in regards to e-mail as females likely participate equally in this medium and may be just as likely to send sexual jokes over e-mail (given the apparent abundance of such e-mails). At least one study, however, has found gender differences in the perception of jokes as offensive (Biber et al., 2002). Such perceptions of the act as harassing may rely more on who sent the joke, but according to this study, females are just as likely to send sexual jokes via e-mail (i.e., it remains to be seen if a female sending a sexual joke would be as offensive as a male sender). The finding that females are just as likely as males to engage in online behavior that is purposely meant to offend is surprising, however, and may be indicative of the changing nature of SH with the introduction of online technology. More attention may need to be paid to the potential for females as the perpetrators of SH and to potential victims of SH from a female perpetrator.

Table 4. Mean Differences Between Males and Females
Variable

Males (n = 115)

Mean

Females (n = 54)

Mean

T-Test for Equality of Means
CSH1.521.31t(160) = 2.911, p < .01
OE2.021.82t(147) = 2.851, p < .01
E-mail2.071.83t(126) = 1.678, p < .10
Graphic1.271.10t(143) = 2.245, p < .05
Passive1.481.13t(155) = 4.854, p < .01
Pressure1.311.19t(166) = 1.287, ns
Purpose1.571.42t(167) = 1.773, p < .08
Stimulating1.381.17t(166) = 3.261, p < .01
Acceptable1.891.52t(151) = 3.522, p < .01
Anonymous2.261.99t(167) = 1.857, p < .06
Aloneness3.313.42t(166) = −.635, ns

There were also gender differences in perceptions of the OE as stimulating and the acceptability of expressing prejudice, with males having higher mean scores (t(166) = 3.261, p < .01; t(151) = 3.522, p < .01) (see Table 4). This is an interesting finding as it suggests that individuals may seek out online environments that meet their expectations, but still can find support for prejudicial ideas if they seek it out. Notably, this idea is supported by the finding that individuals scoring high in femininity see the OE as less hostile (because they seek out differ environments). This is a good indication for the OE, as it suggests that there are now a great variety of environments available, and it may be possible to avoid those that one considers offensive (although this would not be the case with active SH). Finally, gender differences in perceptions of anonymity were marginally significant (t(167) = 1.857, p < .06), and there were no differences in perceptions of aloneness (t(166) = −.635, ns).

As an initial test of construct validity, the scales created here generally performed as expected, with the exception of the relation between gender-role orientation and CSH (which is in an expected direction, but differed from previous findings). Adding to this the evidence of concurrent validity as demonstrated by the results of hypothesis three, the CSH and OE scales, in their final iteration, seem to meet the test of both reliability in terms of internal consistency and construct validity. Additional research will be necessary to test the nomological network of CSH as compared to face-to-face SH, the construct validity in terms of convergent and discriminant relations as well as predictive validity, and the reliability of these scales in divergent populations.

Overall Discussion

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography

With the significant increase in CMC (Cooper et al., 2006; Herrmann, 2007), the results of this study pose significant challenges for organizations. The lack of legal regulations relating to online behavior leaves accountability for online behavior largely in the hands of the employer. Although it remains difficult to prosecute online perpetrators individually (Valetk, 2002), it is clear that organizations can be held responsible for the online behavior of employees (Glover, 2002; Porter & Griffaton, 2003; Swink & Cameron, 2004). Chevron, for example, settled a SH case for $2.2 million based largely on an e-mail entitled, ‘25 Reasons Why Beer is Better Than Women’. In the case of Tammy S. Blakey v. Continental Airlines, Inc. (2000), the New Jersey Supreme Court held that the airline was liable for a hostile work environment based upon defamatory messages posted on a work-related bulletin board. In Coniglio v. City of Berwyn (2000), the city was held liable for failing to respond to the behavior of a supervisor that repeatedly downloaded pornography on his work computer. All of these cases transuded the cyberworld to create a hostile working environment in which it was determined that the employer was aware or should have been aware of such behavior and failed to act. Given that the second most common technology-related claim brought against organizations involves SH (Mills et al., 2000), it is clear that online SH should be considered a serious threat.

Not only is it in the best interest of the organization to avoid costly litigation, but also to maintain a constructive and productive workplace. In delineating how individuals perceive the OE, the present study found that the characteristics of CMC may run counter to this goal. Notably, previous research has suggested that the organizational climate for SH is the best predictor of face-to-face harassment (Fitzgerald et al., 1995a; Welsh, 1999). Although an organization may take steps to decrease face-to-face SH with clear policies and procedures, the organization may not have policies regarding CMC. Future research is ripe with opportunities to determine to what extent the OE can impact perceptions of the organizational climate for SH.

Nonetheless, it is clear that it is in the best interest of the organization to take direct steps to counteract the online culture. Organizations can demonstrate acceptable and unacceptable behavior, for example, by providing a clear policy that defines such behavior and training employees to identify such behavior. An organizational culture that is not accepting of SH will lay the foundation for an OE that coincides with the external work world. Organizations can eliminate anonymity, by requiring all online user names correspond to given names, and all posts require identification. Aloneness may be counteracted by the sharing of office space, strategic positioning of windows or doorways, or the monitoring of employee behavior.

The present research serves to further augment the importance of this issue to organizations as online norms and factors related to perceived behavioral control are significantly related to the intention to engage in harassing behavior. This study lays the groundwork for further research on this topic by providing scales to assess both perceptions of the OE and intent to engage in harassing behaviors online, however, additional research must be done on this issue to be assured of construct validity and to establish the nomological network surrounding these variables. Additionally, this study suffers from several limitations that can be remedied by further research, such as the use of self-reported behavioral intentions on a sensitive issue and correlational (rather than causational) findings. It is particularly important to note that causation can not be inferred from these findings. Indeed, a dissonance perspective would suggest that behavior may affect perception (i.e., that the relations posited in the current study happen in the opposite causal direction). In this initial study, however, conclusions must rest on the available body of theoretical literature until future research addresses this question.

CSH is a phenomenon that continues to grow, both due to the increasing use of CMC in the workplace, and the nature of cyberspace itself. The present study supports the literature in suggesting that the OE is related to the occurrence of harassment because it is characterized by easy access, anonymity, acceptability and aloneness in concert with experiencing affective stimulation from the underlying hegemonic culture, where abuse is easy and allowed. Unfortunately, the culture of the Internet easily crosses over to the work environment. Although the practical implications are clear, online behavior at work is still not fully understood. Because of the vast implications of employee behavior online for organizations, it is important to continue research on this topic. The present study is meant to be the first step in clearly defining and measuring the constructs relevant to online behavior at work.

References

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography
  • Ajzen, I. (2001). Nature and operation of attitudes. Annual Review of Psychology, 52, 2758.
  • Ajzen, I., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84(5), 888918.
  • Allen, A. (2000). Gender and privacy in cyberspace. Stanford Law Review, 52(5), 11751200.
  • Baird, C.L., Bensko, N.L., Bell, P.A., Viney, W., & Woody, W.D. (1995). Gender influences on perceptions of hostile environment SH. Psychological Reports, 77, 7982.
  • Barak, A. (2005). SH on the internet. Social Science Computer Review, 23(1), 7792.
  • Bell, V., & de La Rue, D. (1995). Gender harassment on the Internet [Electronic version]. Retrieved July 21, 2008, http://www.gsu.edu/˜lawppw/lawand.papers/harass.html.
  • Berdahl, J. L. (2007). Harassment based on sex: Protecting social status in the context of gender hierarchy. Academy of Management Review, 22, 641658.
  • Biber, J. K. Doverspike, D., Baznik, D., Cober, A. B., & Ritter, B. A. (2002). SH in online communications. CyberPsychology and Behavior, 5, 3342.
  • Citron, D.K. (2009). Law's expressive value in combating cyber gender harassment. Michigan Law Review, 108, 3, 373415.
  • Cooper, A., Safir, M.P, Rosenmann, A. (2006). A preliminary look at online sexual activities at the office. CyberPsychology and Behavior, 9(1), 2229.
  • Dall'Ara, E., & Maass, A. (1999). Studying SH in the laboratory. Sex Roles, 41, 681704.
  • Dougherty, T., Turban, D., Olson, D., Dwyer, P. & LaPreze, M. (1996). Factors affecting perceptions of workplace SH. Journal of Organizational Behavior, 17, 489501.
  • Equal Employment Opportunity Commission. (1984). Discrimination because of sex under Title VII of the CRA of 1964, 29 CFR 1609.1. Federal Register 45, 2502425025.
  • Fitzgerald, L. F., Drasgow, F., Hulin, C. L., Gelfand, M. J., & Magley, V. J. (1997). Antecedents and consequences of SH in organizations. Journal of Applied Psychology, 82, 578589.
  • Fitzgerald, L. F., Gelfand, M. J., & Drasgow, F. (1995a). Measuring SH: Theoretical and psychometric advances. Basic and Applied Social Psychology, 17, 425427.
  • Fitzgerald, L.F., & Hesson-McInnis, M. (1989). The dimensions of SH. Journal of Vocational Behavior, 35, 309326.
  • Fitzgerald, L.F., & Ormerod, A.J. (1991). Perceptions of SH. Psychology of Women Quarterly, 15, 281294.
  • Fitzgerald, L.F., & Shullman, S.L. (1993). SH: A research analysis and agenda for the 1990s. Journal of Vocational Behavior, 42, 527.
  • Fitzgerald, L. F., Swan, S., & Fischer, K. (1995b). Why didn't she just report him? Journal of Social Issues, 51, 117138.
  • Gelfand, M.J., Fitzgerald, L.F., & Drasgow, F. (1995). The structure of SH. Journal of Vocational Behavior, 32, 152175.
  • Glover, R. (2002). Privacy in the workplace. Metropolitan Corporate Counsel, 10(6).
  • Gohann, S., & Thacker, R. (1996). Emotional and psychological consequences of SH a descriptive study. Journal of Psychology, 130, 429441.
  • Guadagno, R.E., Muscanell, N.L., Okdie, B.M., Burk, N.M., & Ward, T.B. (2011). Even in virtual environments women shop and men build. Computers in Human Behavior, 27(1), 304308.
  • Gutek, B., Morasch, B., & Kohen, A. (1983). Interpreting social-sexual behavior in a work setting. Journal of Vocational Behavior, 22, 3048.
  • Gutek, B.A. (1985). Sex and the workplace. San Francisco: Jossey-Bass Publishers.
  • Hadler, D., & Jaishankar, K. (2008). Cybercrimes against women in India: Problems, perspectives, and solutions. TMC Academic Journal, 3(1), 4862.
  • Herring, S. (1994). Gender differences in computer mediated communication: Bringing familiar baggage to the new frontier. Paper presented at the meeting of the American Library Association, Miami, FL.
  • Herrmann, A. F. (2007). “People get emotional about their money:” Performing masculinity in a financial discussion board. Journal of CMC, 12(2), 499522.
  • Hinkin, T.R. (1995). A review of scale development practices in the study of organizations. Journal of Management, 21(5), 967988.
  • Huffaker, D.A., & Calvert, S.L. (2005). Gender, identity, and language use in teenage blogs. Journal of CMC, 10(2).
  • Hurt, L.E., Wiener, R.L., Russell, B.L., & Mannen, R.K. (1999). Gender differences in evaluating social-sexual conduct in the workplace. Behav. Sci. & the Law 17, 413433.
  • Kendall, L. (2000). “Oh no! I'm a nerd!” Hegemonic masculinity on an online forum. Gender and Society, 14, 256274.
  • Keyton, J., & Rhodes, S.C. (1999). Organizational SH: translating research into application. Journal of Applied Communication Research, 27, 158173.
  • Kramarae, C. (1995). A backstage critique of virtual reality. In S. Jones (Ed.), CyberSociety: CMC and Community. Thousands Oaks, CA: Sage Publications.
  • Krings F., & Facchin, S. (2009). Organizational justice and men's LSH: The moderating role of sexism and personality. Journal of Applied Psychology, 94(2), 501510.
  • MacKinnon, R. (1997). Virtual rape. Journal of CMC, 2, 4, 0.
  • Maier, F. (1995). Cyberspace: The next gender gap. Upside, 7, 88.
  • McCormick, N., & Leonard, J. (1996). Gender and sexuality in the cyberspace frontier. Women & Therapy, 19, 109119.
  • McGerty, L. (2000). Gendered subjectivities and domestic use of the Internet. CyberPsychology and Behavior, 3(5), 895899.
  • Mills, J.E., Clay, J.M., & Mortensen, M. (2000). Managing e-mail liability. Hotel and Restaurant Administration Quarterly, 10, 6471.
  • Miner-Rubino, K., & Cortina, L. M. (2007). Beyond targets: Consequences of vicarious exposure to misogyny at work. Journal of Applied Psychology, 92, 12541269.
  • Nelson, C.G., Halpert, J.A., & Cellar, D.F. (2007). Organizational responses for preventing and stopping SH. Sex Roles, 56 (11-12), 811822.
  • Newmeyer, F. J. (1988). Extension and implications of linguistic theory. In Newmeyer, F. J. (Ed.), Linguistic theory, p.114. NY, NY: Cambridge University Press.
  • Nunnally, J.C. (1976). Psychometric theory, 2nd Edition. NY: McGraw Hill.
  • O'Connell, C.E., & Korabik, K. (2000). The relationship of personal vulnerability, work context, perpetuator status, and type of harassment to outcomes. Journal of Voc. Beh., 56, 299329.
  • Porter, W.G. & Griffaton, M.C. (2003). Monitoring the electronic workplace. International Association of Defense Counsel Journal [Electronic version]. Retrieved August, 2008, http://home.millsaps.edu/fiserhl/.
  • Pryor, J. B., Giedd, J. L., & Williams, K. B. (1995). A social psychological model for predicting SH. Journal of Social Issues, 51, 6984.
  • Pryor, J. B., & Whalen, N. J. (1997). Characteristics of harassers and the social circumstances under which SH occurs. In W. O'Donohue (Ed.), SH: Theory, research, and treatment (pp. 129151). Needham Heights, MA: Allyn & Bacon.
  • Ritter, B.A., & Doverspike, D. (April, 2002). The changing nature of SH. Presented at the Society for Industrial/Organizational Psychology Conference, Toronto, ON.
  • Rodino, M. (1997). Breaking out of binaries. Journal of CMC, 3(3).
  • Rotundo, M., Nguyen, D., & Sackett, P.R. (2001). A meta-analytic review of gender differences in perceptions of SH. Journal of Applied Psychology, 86, 914922.
  • Sardar, Z. (1999). Cyberspace as the darker side of the west. Futures, 27 (7), 777794.
  • Shea, V. (1995, August 21). Not afraid to flame. Computerworld, 29, 8182.
  • Siegal, J., Dubrovsky, V., Kiesler, S., & McGuire, T. W. (1986). Group processes in CMC. Organizational Behavior and Human Decision Processes, 37, 157187.
  • Soukup, C. (1999). The gendered interactional patterns of computer-mediated chatrooms. Information Society, 15(3), 169176.
  • Stone, A.R. (1993). Violation and virtuality [Electronic version]. Retrieved March 18, 2002, from The University of Texas Web site: http://www.en.utexas.edu/˜slatin/opera
  • Sullivan, A. (1999). The innernet. Forbes, 10(4), 157158.
  • Summers, R.J. (1996). The effect of harasser performance status and compliant tolerance on reactions to a complaint of SH. Journal of Vocational Behavior, 49, 5367.
  • Sussman, N.M., & Tyson, D.H. (2000). Gender differences in computer-mediated interactions. Computers in Human Behavior, 16, 381394.
  • Swink, D. & Cameron, E. (2004). Employee use of Internet. The Academy of Legal Studies in Business Journal of Employment and Labor Law, 10(1).
  • Taylor, H.J., Kramarae, C., & Ebben, M. (Eds.). (1993). Women, information technology, and scholarship. Urbana, IL: University of Illinois Centers for Advanced Study.
  • Valetk, H.A. (2002). Cyberstalking: Navigating a maze of laws. New York Law Journal [Electronic version]. Retrieved March 13, 2007, www.valetk.com/pdf/cyberstalking.pdf
  • Welsh, S. (1999). Gender and SH. Annual Review of Sociology, 25,169190.
  • Willness, C. R., Steel, P., & Lee, K. (2007). A meta-analysis of the antecedents and consequences of workplace SH. Personnel Psychology, 60, 127162.

Biography

  1. Top of page
  2. Abstract
  3. Theoretical Framework
  4. Method
  5. Phase 1: Item Development
  6. Phase 1: Results and Discussion
  7. Phase 2: Scale Development
  8. Phase 2: Results and Discussion
  9. Phase 3: Scale Evaluation
  10. Phase 3: Results and Discussion
  11. Overall Discussion
  12. References
  13. Biography
  • Barbara A. Ritter is an Associate Professor and Department Chair of the Management and Decision Sciences Department in the E. Craig Wall Sr. College of Business Administration at Coastal Carolina University. Barbara earned a Ph.D. in Industrial and Organizational Psychology from the University of Akron in 2004. Barbara's research interests include perceptions and legitimacy of organizational leaders, justice perceptions, and sexual harassment. She has been teaching and studying these and other diversity-related issues for over 10 years and has led training exercises for MBA students, local organizations, and national and international conferences. Her recent publications appear in The Journal of Applied Psychology, The Journal of Business Ethics, and Human Relations. Barbara can be contacted at the E. Craig Wall Sr. College of Business Administration, Coastal Carolina University, Conway, SC, 29528.