“I’ve had enough!” Mental health as a mechanism in the relationship between sexual harassment and organizational commitment

Abstract

BACKGROUND:

Research on workplace sexual harassment has focused largely on its direct antecedents and consequences. In contrast, few studies unravel mechanisms through which sexual harassment is related to organizational outcomes.

OBJECTIVE:

Following affective events theory, the purpose of this paper is to investigate mental health as a mechanism in the relationship between sexual harassment and organizational commitment.

METHODS:

Quantitative data were collected using a survey design for which 249 employees responded.

RESULTS:

Structural equation modeling showed that the mediation model was partially supported for the overall sample because mental health was a partial and not full mediator. However, post-hoc analysis revealed that the mediation model was suited to men, but not women.

CONCLUSIONS:

This study contributes to sexual harassment research by (a) investigating a novel mechanism in the sexual harassment to organizational commitment relationship and (b) highlighting the importance of gender differences when attempting to unravel the sexual harassment causal chain. Implications for organizations regarding the development of sexual harassment policies, complaint procedures, and training are explained.

Keywords

Counterproductive work behaviour unwelcomed sex-related behaviors psychological distress psychological well-being affective events theory

1 Introduction

Employees’ mental health is a topic that has received much attention by researchers over the past few years, and rightfully so. According to the World Health Organization (WHO), over 300 million persons worldwide suffer from depression [1]. Depression not only affects the quality of people’s lives, but also has trickle-down effects on a country’s economy. The economic impact of depression and anxiety on productivity averages one trillion US dollars annually in the global economy [2]. One potential cause of depression in the workplace is sexual harassment [3].

In the early 2000s, sexual harassment claims in the US declined significantly from the 1990s [4]. However, in present times, sexual harassment has become a topical and pervasive issue that has been receiving considerable media attention globally [5, 6], inclusive of Trinidad and Tobago [7 –9]. Sexual harassment usually takes the form of unsolicited physical contact and unwelcomed verbal and non-verbal advances. In the workplace, these sexual harassment actions are often regarded as a form of abuse or counterproductive behavior, which can lead to hostile, intimidating, humiliating, and disruptive work environments [10].

Researchers have examined important outcomes of sexual harassment in the workplace (hereinafter referred to as sexual harassment). For instance, a meta-analysis by Willness, Steel, and Lee [11] reported that sexual harassment is related to (a) job-related outcomes, e.g., commitment, satisfaction, work withdrawal, and productivity and (b) health and well-being outcomes, e.g., mental health, physical health, and posttraumatic stress disorder. To date, most of these studies on sexual harassment examine its direct outcomes. A few studies explain the mechanisms through which sexual harassment is related to its outcomes [12 –14]. The present paper adds to the latter body of sexual harassment research by investigating a novel mechanism. Specifically, this paper draws on affective events theory (AET) [15] to explain that sexual harassment is related to employees’ affective state (mental health), which, in turn, is related to employees’ attitude towards their organization (organizational commitment). The aim of this paper is to contribute to the sexual harassment literature by identifying a what (i.e., mental health), to explain how the relationship between sexual harassment and organizational commitment unfolds [16].

1.1 Sexual harassment

In a 30-year review of sexual harassment literature from 1988 to 2008, Cortina and Berdahl [3] explain that sexual harassment can be viewed from three perspectives. First, sexual harassment can be defined from a legal perspective created by the Equal Employment Opportunity Commission (EEOC) in 1980. The EEOC defines sexual harassment as

unwelcome sexual advances, requests for sexual favors, and other verbal or physical conduct of a sexual nature constitute sexual harassment when this conduct explicitly or implicitly affects an individual’s employment, unreasonably interferes with an individual’s work performance, or creates an intimidating, hostile, or offensive work environment [17].

Second, sexual harassment can be defined from a social-psychological perspective. Unlike the legal perspective, the social-psychological perspective does not include work outcomes in its definition, e.g., interfering with an individual’s performance. Instead, the social-psychological perspective focuses on individuals’ perceptions of specific behaviors. In this view, sexual harassment is defined as “unwanted sex-related behavior at work that is appraised by the recipient as offensive, exceeding her resources, or threatening her well-being” [18]. This definition is based upon women’s experiences, but recent studies refine this definition for men as well. Third, sexual harassment can be defined from a lay perspective. The lay perspective adjusts the definition of sexual harassment according to society’s changing perceptions of sexual harassment behaviours [19].

The sexual harassment definition challenge is outside the scope of this paper. Nonetheless, the varying perspectives must be highlighted prior to explaining the definition used in this paper. For the purpose of conducting an empirical study of sexual harassment, the stance adopted in this paper is the social-psychological perspective because it (a) focuses on specific sexual harassment behaviors without including outcomes, and (b) has led to the development of a well-accepted measure of sexual harassment, and thus lends itself to measurement of the concept. In this view, researchers generally agree that sexual harassment refers to perceived implicit or explicit unwelcomed sex-related behaviors that can take the form of gender harassment, unwanted sexual attention, and/or sexual coercion [3, 20].

This definition of sexual harassment highlights its dimensional structure, including gender harassment, unwanted sexual attention, and sexual coercion. Gender harassment refers to a range of degrading, insulting and hostile behaviors (verbal and non-verbal). Gender harassment can be subdivided into sexist hostility and sexual hostility. Sexist hostility refers to experiences that are discriminatory based on one’s gender, e.g., derogatory gender-related remarks. Sexual hostility refers to experiences that are explicitly sexual, e.g., lewd sexual stories or jokes. Unwanted sexual attention refers to unreciprocated requests and unwelcoming gestures, e.g., touching, grabbing, and intrusive phone calls. Sexual coercion or quid pro quo occurs when one person offers another organizational rewards in return for sexual favours. Sexual coercion typically occurs when there are power differences between two parties, e.g., supervisor and subordinate. Overall, given the two sub-dimensions of gender harassment, sexual harassment is composed of four dimensions.

Sexual harassment research has primarily focused on authoritative figures harassing a lower-level employee, e.g., supervisor-subordinate relationships. However, sexual harassers can also be coworkers, customers, and clients [3]. The present paper investigates the workplace context in general, and thus focuses on sexual harassment by all employees in the workplace, including both supervisors and coworkers. Moreover, this paper examines employees engaging in sexual harassment in Trinidad because sexual harassment is especially likely to be problematic in this context. In Trinidad, sexual innuendos are deeply embedded in the national culture. Specifically, music and literature in Trinidad is filled with sexual imagery and metaphors, which may make it difficult for employees to discern sexual boundaries in interactions [21].

1.2 Mental health

Mental health is “a state of well-being in which every individual realizes his or her own potential, can cope with the normal stresses of life, can work productively and fruitfully, and is able to make a contribution to her or his community” [22, para. 2]. The WHO’s definition of mental health highlights two important dimensions of the concept, including psychological distress and well-being. Prevalent symptoms of psychological distress such as anxiety and depression have been the focus of research on mental health [23]. However, Veit and Ware [23] explain that the definition of mental health should also include characteristics of psychological well-being. In support of this proposition, Veit and Ware’s [23] empirical examination of mental health showed that it is composed of two key dimensions. First, psychological distress includes anxiety, depression, and loss of behavioral/emotional control. Second, psychological well-being includes general positive affect and emotional ties 1 .

1.3 Sexual harassment and mental health

An individual subjected to psychological trauma, such as physical abuse, emotional abuse, or sexual abuse, has an increased probability for developing poor mental health or a mental illness [25]. According to Cortina and Berdahl [3], the longevity of employees’ sexual harassment experiences determine degrees of depression, general stress and anxiety, posttraumatic stress disorder, and overall impaired psychological well-being. After all, sexual harassment is an event that should impact employees’ affect (this point is discussed in more detail later on). Hence, unsurprisingly, mental health is one of the most commonly studied consequences of sexual harassment, and studies show a negative relationship between the two concepts [10 , 27]. Furthermore, sexual harassment is associated with reduced mental health across a variety of occupations, such as flight attendants, nurses, and teachers [27]. Therefore, the following is proposed:

Hypothesis 1. There is a negative relationship between sexual harassment and mental health.

1.4 Mental health and organizational commitment

Meyer and Allen [28] refer to commitment as “a psychological state that (a) characterizes the employee’s relationship with the organization, and (b) has implications for the decision to continue or discontinue membership in the organization”. Meyer and Allen further explain that organizational commitment is composed of three dimensions, including affective, normative, and continuance. Affective commitment refers to an emotional attachment, identification and involvement with an organization and is expressed by employees wanting to remain with the organization. Normative commitment refers to an obligatory feeling whereby employees believe that they ought to remain with an organization. Continuance commitment involves the costs associated with leaving where employees perform a cost assessment and need to remain with an organization.

One factor that may influence employees’ overall commitment to their organization is their mental health. While the association between mental health and organizational commitment is well established [29], the direction of this relationship is a bit unclear in extant research. Meyer et al. [29] explain that researchers have yet to arrive at consensus on the causal ordering of commitment in relation to correlates such as health and well-being. Some researchers investigate commitment as a moderator in the relationship between work stressors and employee well-being [30, 31]. Other researchers show that commitment can be a precursor to well-being [32]. Few researchers suggest that well-being can be an antecedent to commitment [33]. Jain et al. [33] explain that individuals who experience a sense of well-being in their organization may become attached to said organization. These individuals may develop genuine feelings for the organization, rather than merely thinking about their employment relationship from a profit and loss perspective [33]. Following the latter stance that well-being is a predictor of organizational commitment (which is further explained in the next subsection), the following is proposed:

Hypothesis 2. There is a positive relationship between mental health and organizational commitment.

1.5 Sexual harassment, mental health, and organizational commitment

Sexual harassment is likely to have implications for employees’ organizational commitment. Willness et al. [11] explain that employees may hold the organization responsible for allowing sexual harassment to manifest itself, and thus partly blame the organization for sexual harassment. When the organization does not protect its employees from sexual harassment, employees may feel angry and disillusioned with the organization – not just the harasser [11]. In line with these arguments, meta-analyses show that sexual harassment is negatively related to organizational commitment [11, 26]. Willness et al. [11] explain that the moderate correlation between sexual harassment and organizational commitment may be better understood by the introduction of third variables. Hence, while sexual harassment may directly influence organizational commitment, researchers need to unravel the potential mechanisms between sexual harassment and organizational commitment. One theory that may help to explain how sexual harassment influences organizational outcomes is affective events theory (AET).

AET proposes that events in the work environment cause affective reactions, which in turn influence work attitudes [15]. In AET, an event is “an incident that stimulates appraisal of and emotional reaction to a transitory or on-going job-related agent, object or occurrence” [34]. For this definition, Basch and Fisher [34] use an example of an employee who is disgusted because her boss made a sexual innuendo. Here, the employee appraises her boss (a job-related agent), whose action triggers an emotional response of disgust. Similarly, in this paper, it is expected that employees would appraise their supervisors or coworkers’ sexual harassment events in the workplace (job-related agents), and one affective reaction to these events can be mental health, i.e., anxiety, depression, positive affect, and behavioral control. Moreover, if sexual harassment is considered a negative event, then it is likely to produce a stronger mental health reaction that positive events [15]. The consequence of mental health, i.e., the affective experience, is then organizational commitment, i.e., a work attitude. Empirically, few studies examine affect-related mechanisms in sexual harassment research, and show that fear and negative mood can mediate the relationship between sexual harassment and organizational commitment [35]. Hence, in line with AET, the following is proposed:

Hypothesis 3. Mental health is a mechanism in the relationship between sexual harassment and organizational commitment. Specifically, sexual harassment is negatively related to mental health, which, in turn, is positively related to organizational commitment.

1.6 Gender differences in response to sexual harassment

One of the most examined moderators in sexual harassment research is victims’ gender [3]. Specifically, researchers are often concerned with differing reactions between men and women in response to sexual harassment. Even though victims’ gender has been studied extensively in sexual harassment research, the findings have been mixed. In most studies, the relationship between sexual harassment and mental health is considerably stronger for women than for men [3]. Here, women report worse outcomes than men with respect to mood, stress, and burnout [36, 37]. On the opposite end of the spectrum, few researchers showed that the relationship between sexual harassment and mental health is considerably stronger for men than for women, i.e., men experiencing stronger depression and anxiety [38]. Vogt et al. [38] explain that when a man is sexually harassed, it is unexpected. Therefore, sexual harassment may be more of a shock to a man than a woman, thus leading to a stronger impact on men’s mental health. Finally, studies also show no differences between genders in the relationship between sexual harassment and mental health [39 –41]. Given the mixed findings and lack of consensus on the role of victims’ gender in sexual harassment research, no hypothesis is put forward for any gender differences for the proposed mediation model. Still, the model is later investigated separately for men and women.

2 Methods

2.1 Ethical considerations

According to The University of the West Indies’ (UWI) policy and procedures on research ethics, the present study was classified as a survey in which there was no likelihood that participants could be identified or revealed. For this reason, the Head of Department at the Department of Management Studies determined that the project was exempt from review, which was the practice at The UWI at the time of data collection. Still, we followed ethical protocols for research and publication as outlined by the American Psychological Association. In so doing, we provided potential participants with an information sheet that preceded the questionnaire. The information sheet described: (1) general information such as the purpose of the study, benefits of the study, monetary incentive for participation (i.e., TT$2.00), and expected duration for completing the questionnaire and (2) ethical issues such as participation being voluntary with the right to decline participation or withdraw at any point, anonymity and confidentiality of responses, how to deal with potential discomfort in responding to the questions (e.g., consulting a family doctor), and contact information for any complaints or concerns.

2.2 Participants

The sample consisted of 249 participants from workplaces throughout Trinidad. Employees were from workplaces in Toco, Sangre Grande, Arima, Maloney, St. Augustine, Curepe, Maracas, St. Joseph, Port of Spain, Maraval, Chaguanas, and San Fernando (n = 20 for each location except Port of Spain) 2 . The sample included 84 men (mode age was 20 to 29 years) and 165 women (mode age was 20 to 29 years). Overall, participants were from five different age groups including 18 to 20 years (n = 40, 16.1%), 20 to 29 years (n = 117, 47.0%), 30 to 39 years (n = 42, 16.9%), 40 to 49 years (n = 25, 10.0%), 50 to 59 years (n = 22, 8.8%), and 60 and over (n = 3, 1.2%). Participants had to be employed at any organization regardless of sector, industry, and purpose (i.e., profit or nonprofit). There were no other inclusion/exclusion criteria.

2.3 Materials

2.3.1 Sexual harassment

We used the Sexual Experiences Questionnaire (SEQ) to measure sexual harassment [42]. The SEQ is a 23-item questionnaire (e.g., “Repeatedly told sexual stories or jokes that were offensive to you?”) that uses a 5-point continuum. We used the following opening statement for the SEQ: “Between the <beginning of the year >and <the month before data collection >, an employee has ... ” 3 . Higher scores indicated higher sexual harassment. The SEQ has four subscales, including (1) gender harassment (sexist hostility) (4 items; α= 0.92), (2) gender harassment (sexual hostility) (8 items, α= 0.93), (3) unwanted sexual attention (6 items; α= 0.90), and (4) sexual coercion (5 items; α= 0.95). Cronbach’s alpha for the entire 23-item questionnaire was 0.96.

2.3.2 Mental health

We used the Mental Health Inventory (MHI) to measure mental health [24]. This 18-item questionnaire (e.g., “Have you been a very nervous person?”) uses a 6-point continuum. Items were coded so that higher scores indicated higher mental health. Participants were asked to think about how they felt during the month in which data were collected. The MHI has four subscales, including anxiety (4 items; α= 0.64), depression (4 items; α= 0.84), behavior control (3 items; α= 0.78), and positive affect (4 items; α= 0.61). Cronbach’s alpha for the entire 18-item questionnaire was 0.89.

2.3.3 Organizational commitment

We used the Organizational Commitment Scale (OCS) to measure organizational commitment [43]. This 19-item questionnaire (e.g., “I really feel as if this organization’s problems are my own”) uses a 7-point continuum. We coded all items so that higher scores indicated higher organizational commitment. The OCS has three subscales, including (1) affective commitment (6 items; α= 64), (2) continuance commitment (7 items; α= 87), and (3) normative commitment (6 items; α= 78). Cronbach’s alpha for the entire 18-item questionnaire was 0.85.

2.4 Procedures

Seven data collectors distributed 350 questionnaires at organizations located throughout Trinidad. These organizations included five public-sector organizations, four government ministries, two educational institutions, a religious organization, a media house, a construction company, and a bank. Even though we used non-probability techniques to select organizations for the present study, we used systematic random sampling within each organization. Specifically, we distributed questionnaires to every third person at these organizations with a goal of collecting 20 completed questionnaires from each organization.

When distributing the questionnaires, we asked persons to complete the questionnaire and return it immediately. If an individual was unable to complete the questionnaire immediately, we returned at a future date to collect the questionnaire. Some persons, who were unable to complete the questionnaire immediately, opted to scan and email the completed questionnaire. We also facilitated a few individuals who requested soft copies to complete. Of the 350 distributed questionnaires, 249 were successfully completed 4 and returned, and thus the response rate was 71.14%.

We used three procedural remedies to reduce the effects of common method bias. First, we minimized common scale properties by varying both scale points and anchor labels [44]. Second, we minimized social desirability bias by using the SEQ’s item wordings, which do not refer to ‘sexual harassment’, but instead asks participants to evaluate behaviors [44]. The intention here was to reduce under-reporting of sexual harassment, which may particularly be an issue in countries that lack legislation to deal with sexual harassment, e.g., Trinidad [45]. Third, we further minimized social desirability bias by assuring participants that their responses were anonymous [44].

3 Results

3.1 Statistical assumptions and common method bias

The data were evaluated to test statistical assumptions and examine common method bias. The Koenker test revealed no issues for heteroscedasticity. Bi-variate scatterplots showed no issues with non-linearity. Kurtosis and skewness z-values for most of the variables exceeded±2.58, and thus the data were non-normal. For this reason, maximum likelihood with the Satorra-Bentler adjustment to the χ² for non-normality was later used [46] via the Lavaan package [47] for R [48]. Harman’s one-factor test showed that the largest factor did not account for the majority of the variance (30.4%), and thus common method bias did not appear to be problematic [49]. Table 1 illustrates a correlation matrix for all of the study’s variables.

Table 1
Summary of Intercorrelations, Means, and Standard Deviations for the Scores of Sexual Harassment, Mental Health, and Organizational Commitment

Variables 1 2 3 4 5 6 7 8 9 10 11 Mean SD

1. Gender harassment (sexist) 2.22 1.24

2. Gender harassment (sexual) 0.75^ 2.57 1.22

3. Unwanted sexual attention 0.56^ 0.78^ 2.32 1.23

4. Sexual coercion 0.61^ 0.66^ 0.75^ 1.95 1.20

5. Sexual harassment 0.82^ 0.93^ 0.89^ 0.85^ 12.06 5.66

6. Depression &anxiety –0.32^ –0.34^ –0.33^ –0.36^ –0.39^ 4.19 1.10

7. Behavioral control –0.35^ –0.33^ –0.35^ –0.43^ –0.41^ 0.69^** 4.69 1.49

8. Mental health –0.37^ –0.37^ –0.37^ –0.43^ –0.43^ 0.89^ 0.94^ 4.44 1.19

9. Normative commitment –0.21^ –0.23^ –0.18^ –0.14^* –0.22^ 0.20^ 0.20^ 0.22^ 3.05 1.54

10. Continuance commitment –0.15^* –0.12 –0.10 –0.07 –0.12 0.10 0.07 0.09 0.57^ 3.12 1.41

11. Affective commitment –0.21^ –0.22^ –0.21^ –0.16^* –0.23^ 0.22^ 0.18^ 0.21^ 0.56^ 0.41^ 3.09 1.50

12. Organizational commitment –0.23^ –0.23^ –0.20^** –0.15^* –0.23^ 0.21^ 0.19^ 0.21^ 0.87^ 0.79^ 0.80^** 3.09 1.22

Variables		1	2	3	4	5	6	7	8	9	10	11	Mean	SD
1.	Gender harassment (sexist)												2.22	1.24
2.	Gender harassment (sexual)	0.75^**											2.57	1.22
3.	Unwanted sexual attention	0.56^**	0.78^**										2.32	1.23
4.	Sexual coercion	0.61^**	0.66^**	0.75^**									1.95	1.20
5.	Sexual harassment	0.82^**	0.93^**	0.89^**	0.85^**								12.06	5.66
6.	Depression &anxiety	–0.32^**	–0.34^**	–0.33^**	–0.36^**	–0.39^**							4.19	1.10
7.	Behavioral control	–0.35^**	–0.33^**	–0.35^**	–0.43^**	–0.41^**	0.69^**						4.69	1.49
8.	Mental health	–0.37^**	–0.37^**	–0.37^**	–0.43^**	–0.43^**	0.89^**	0.94^**					4.44	1.19
9.	Normative commitment	–0.21^**	–0.23^**	–0.18^**	–0.14^*	–0.22^**	0.20^**	0.20^**	0.22^**				3.05	1.54
10.	Continuance commitment	–0.15^*	–0.12	–0.10	–0.07	–0.12	0.10	0.07	0.09	0.57^**			3.12	1.41
11.	Affective commitment	–0.21^**	–0.22^**	–0.21^**	–0.16^*	–0.23^**	0.22^**	0.18^**	0.21^**	0.56^**	0.41^**		3.09	1.50
12.	Organizational commitment	–0.23^**	–0.23^**	–0.20^**	–0.15^*	–0.23^**	0.21^**	0.19^**	0.21^**	0.87^**	0.79^**	0.80^**	3.09	1.22

Note. Composite variables were calculated for the correlation matrix. The main concepts were italicized. For the sexual harassment variables, higher scores were indicative of higher sexual harassment; for the mental health variables, higher scores were indicative of higher mental health; and for the organizational commitment variables, higher scores were indicative of higher organizational commitment. ^**Correlation was significant at the 0.01 level (2-tailed). ^*Correlation was significant at the 0.05 level (2-tailed).

3.2 Confirmatory factor analyses

Prior to testing the hypothesized mediation model, we examined the factor structures for sexual harassment, mental health, and organizational commitment. For each factor structure, we deleted items when they had poor factor loadings (i.e., <0.6) and/or problematic standardized residuals (i.e., surpassing the threshold of |2.5|) [50]. For sexual harassment, we deleted two items and estimated a second-order sexual harassment model. For mental health, we also deleted two items, and depression and anxiety were combined into one latent factor because both constructs were too highly correlated to be represented as separate latent factors. We then estimated a second-order mental health model. For organizational commitment, we deleted four items and estimated a second-order organizational commitment model. These four items were the only reverse-scored items in the OCS, and reverse-scored items were found to be less internally consistent, have lower item-scale correlations, and associated with more participant inattention and confusion than non-reversed items [51, 52]. Each of the three CFA models fitted the data well (Robust CFI ≥0.95 and RMSEA <0.08) [50].

3.3 Structural models

We followed Anderson and Gerbing’s [53] two-step process to test the path model. For the first step, we estimated the measurement model with all constructs. The fit indices for the measurement model indicated a good model fit considering the sample size and model complexity (Satorra-Bentler χ² (1058) = 1678.43, p < 0.01, Robust CFI = 0.92, RMSEA = 0.049) [50]. For the second step, we converted the measurement model to a structural model. The structural model fitted the data fairly well (Satorra-Bentler χ² (1059) = 1684.03, p < 0.01, Robust CFI = 0.92, RMSEA = 0.049) (see estimates on the left side in Fig. 1).

Fig.1

Structural model of the relationships between sexual harassment, mental health, and organizational commitment. Standardized parameter estimates. Estimates to the left are for the first model, i.e., prior to adding a direct path from sexual harassment to organizational commitment. Estimates to the right are for the second model with a direct unmediated path from sexual harassment to organizational commitment. Indicator variables and error variances excluded for ease in readability. All first-order latent variable indicators were significant (p < 0.001). ^***p < 0.001. ^**p < 0.01. ^*p < 0.05, ^†p < 0.10.

The findings for the first model in Fig. 1 provided initial evidence to support all three hypotheses. H1 was supported because sexual harassment was negatively related to mental health (γ= –0.49, p < 0.001). H2 was also supported because mental health was positively related to organizational commitment (β= 0.28, p < 0.01). For H3, we further tested the mediation effects in the structural model by following two steps proposed by Hair et al. [50]. The first step was to establish the significant relationships between the three constructs. To do so, we checked the following: (a) the direct unmediated relationship between sexual harassment and organizational commitment (φ= –0.30, p < 0.001), (b) the relationship between sexual harassment and mental health (φ= –0.48, p < 0.001), and (c) the relationship between mental health and organizational commitment (φ= 0.26, p < 0.01). Having established the relationships between the constructs, the second step was to assess the level of mediation through the mediated model. Here, we added a direct unmediated relationship between sexual harassment and organizational commitment to the structural model. We then conducted a Satorra-Bentler scaled chi-square difference test, which showed that the addition of the direct unmediated path significantly improved the model (Δ Satorra-Bentler χ² (1) = 5.60, p < 0.05) (see estimates on the right side in Fig. 1).

The findings for the second model in Fig. 1 indicated that mental health might be a partial mediator in the relationship between sexual harassment and organizational commitment. Specifically, (a) the direct path between sexual harassment and organizational commitment was significant (γ= –0.22, p < 0.05) and (b) the relationship between mental health and organizational commitment was now marginally significant (β= 0.15, p < 0.10) after adding the direct unmediated path between sexual harassment and organizational commitment. Moreover, the indirect effect of sexual harassment on organizational commitment was –0.07 (p < 0.10), and thus the total effect of sexual harassment on organizational commitment was –0.30 (p < 0.001). Overall, H1 was fully supported, but H2 and H3 were only partially supported.

A multi-group structural model was used to examine potential gender differences. Given the small sample size for the men group (n = 84), a composite indicator for mental health and organizational commitment was used in order to simplify the model for a smaller sample size (see Fig. 2). The findings showed that the relationship between sexual harassment and mental health was similar for men and women (γ= –0.41 and –0.45 respectively). Furthermore, the total effect of sexual harassment on organizational commitment was similar for men and women (–0.25 and –0.27 respectively). However, there are notable differences in the hypothesized model between men and women. For men, mental health was a full mediator in the relationship between sexual harassment and organizational commitment 5 . But, for women, mental health was not a mediator in the same relationship. Specifically, the relationship between mental health and organizational commitment for women was considerably weaker than that for men. Instead, sexual harassment had more of a direct association with organizational commitment for women. These differences in findings suggested that men’s organizational commitment might be affected by their mental health resulting from sexual harassment, whereas women’s organizational commitment might be largely affected by the sexual harassment itself.

Fig.2

Multi-group structural model of the relationships between sexual harassment, mental health, and organizational commitment for men and women. Standardized parameter estimates. Estimates to the left are for men (n = 84) and estimates to the right are for women (n = 165). Indicator variables and error variances excluded for ease in readability. All first-order latent variable indicators were significant (p < 0.001). ^***p < 0.001. ^**p < 0.01. ^*p < 0.05. ^†p < 0.10.

4 Discussion

The purpose of this study was to investigate how sexual harassment is related to organizational commitment in the workplace. Using AET, we propose that mental health is a mechanism in the relationship between sexual harassment and organizational commitment. The present study’s findings show that employees who perceive sexual harassment in the workplace are more likely to experience mental health issues. Moreover, mental health appears to be a partial mediator in the relationship between sexual harassment and organizational commitment.

The main contribution of the present study is that we add to the limited body of research that investigates mechanisms through which sexual harassment is related to important organizational outcomes. Specifically, our findings show how sexual harassment is related to organizational commitment. According to the meta-analysis by Willness et al. [11], there still remains a lack of theoretical development in this field of sexual harassment, and thus our intention was to move the research in this area towards a more mature level by unravelling potential causal maps. Although the proposed mediation model is only partially supported, the findings show an interesting difference between genders.

We also contribute to sexual harassment research by showing that men and women may react differently to sexual harassment. Although we did not hypothesize any gender differences, the findings reveal an interesting difference in how men and women respond to sexual harassment. At first glance, there might appear to be little to no differences between the genders, i.e., the relationships between sexual harassment and both mental health and organizational commitment are nearly identical for men and women. However, using AET as a framework for the mediation model, the differences between genders become clear – men’s commitment is associated with their mental health whereas women’s commitment is associated with the sexual harassment itself. Therefore, sexual harassment did not directly predict men’s commitment, but commitment was still impacted for men (albeit indirectly). In contrast, women appear to hold the organization responsible for sexual harassment events, and thus sexual harassment directly predicts women’s commitment.

Women may feel that the organization is not doing enough to protect them from sexual harassment, and thus develop lowered commitment towards the said organization. As such, AET does not seem to explain women’s mental health reactions to sexual harassment in relation to organizational commitment. Instead, attribution theory may better explain why sexual harassment is directly related to women’s organizational commitment. Attribution theory is a process through which individuals use information to determine an explanation for an event, i.e., what caused an event. Individuals collect and combine information to form a ‘causal judgment’ [54]. Perhaps women may be more likely than men to use information that leads to blame being attributed to the organization. Here, women may perceive that the organization gives “the opportunity for a dysfunctional colleague to act out against a colleague who was seen as subordinated to him” [41]. In other words, women may generalize the source of sexual harassment from their supervisor or coworker to the organization as a whole, and thus direct blame to the organization. If so, then this broadening of the blame to the larger organization context may explain the direct relationship between sexual harassment and organizational commitment for women. The findings for gender differences is also interesting because most research on attribution of blame for sexual harassment focus on the victim or harasser as the one responsible for sexual harassment [55]. However, it is worth investigating the extent to which individuals, particularly women, attribute blame for sexual harassment to the organization.

4.1 Limitations and future research

Notwithstanding the theoretical contributions made, the present study is limited by its use of cross sectional data. Cross sectional data cannot be used to test causal maps, i.e., the reverse relationships can be true, e.g., employees with poor mental health are targeted by sexual harassers. Therefore, future research needs to examine sexual harassment over a period of time, via longitudinal research designs. Longitudinal research designs are severely lacking in sexual harassment research [56] and may be particularly useful for examining sexual harassment, which tends to be an ongoing occurrence [11]. A longitudinal design may be useful for tracking changes in employees’ outcomes, e.g., mental health and organizational commitment, when sexual harassment incidents occur. Furthermore, a longitudinal design is useful to better understand gender differences in organizational commitment reactions to sexual harassment. Specifically, does sexual harassment immediately impact women’s commitment, but lead to lagged effects for men, whereby men’s mental health is first affected?

Future studies should also examine other important organizational outcomes. This paper’s focus is organizational commitment. But, mental health may affect other organizational outcomes such as performance, organizational citizenship behavior, counterproductive work behaviors, etc. In Trinidad and Tobago, outcomes such as absenteeism and productivity are particularly problematic [57, 58], and thus worth examining in relation to sexual harassment and mental health. In examining these relationships, future research can be guided by AET and attribution theory to better understand the causal pathways through which sexual harassment influences important organizational outcomes. In so doing, these causal pathways should be compared between genders. Our findings should not be misinterpreted to mean that AET applies to men and not women. We investigate only one aspect of affect – mental health. Perhaps women feel angry with the organization when it does not protect them from sexual harassment, and this anger then translates into lower commitment. Therefore, further research is needed to better understand how AET applies to women’s reactions to sexual harassment in particular.

4.2 Practical implications

We propose two recommendations to minimize, or possibly eliminate, sexual harassment. The first recommendation is that organizations should develop sexual harassment policies along with complaint procedures. Sexual harassment policies are geared towards preventing sexual harassment, and should: (a) define sexual harassment including that the organization strongly disapproves of such behavior and (b) be in writing via handbooks and other displays around the organization [59]. In addition to taking preventative action against sexual harassment via policies, organizations also need remedial action via complaint/grievance procedures. Complaint procedures should meet the following criteria: (a) outline the steps for making a complaint, (b) ensure prompt and comprehensive investigations of complaints, and (c) ensure immediate corrective action for valid complaints [59, 60]. In addition to formal harassment policies and complaints procedures, informal approaches such as mediation may be useful for victims who wish to avoid formal and public investigations that can be confrontational. Empirically, one study showed that sexual harassment policies and procedures can strongly influence organizational commitment in the military context [61]. These policies and procedures are even more effective when disseminated and enforced [62], which leads to the second recommendation.

The second recommendation is that sexual harassment policies and procedures need to be integrated with training. Training is useful for educating employees about the meaning and policies relating to sexual harassment [59]. Sexual harassment training is typically designed using instructional videos, which have been shown to improve participants’ knowledge of sexual harassment behaviors and even reduce men’s sexually harassing behaviors [63]. Other sexual harassment training methods include lectures, skits or plays, films, case-studies, computer- and internet-based training, face-to-face group sessions, and experiential methods (e.g., role play and behavior modeling) [3, 59]. These training methods may even be used to develop conflict management and teambuilding programs. Such programs may improve relationships between coworkers and management as well as relationships between coworkers themselves, thus creating an organizational climate in which sexual harassment is not tolerated [64]. For an in-depth discussion on organizational strategies for dealing with sexual harassment, see Cortina and Berdahl [3], Buchanan et al. [59], and Perry et al. [65].

5 Conclusion

Sexual harassment is in the spotlight today, perhaps more so than at any other time in humankind’s recorded history. Hence, it is important to understand how sexual harassment can lead to adverse outcomes for organizations. In the present paper, we used AET to propose that sexual harassment is an event that leads to an affective reaction, which, in turn, influences a workplace attitude. The proposed mediation model is partially supported, but the overall findings are clear – if employees suffer sexual harassment in the workplace, so too will their mental health and organizational commitment. Therefore, organizations should attempt to eliminate sexual harassment by developing and communicating clear sexual harassment policies and complaint procedures as well as utilizing a well-designed sexual harassment training program.

Conflict of interest

None to report.

Footnotes

The emotional ties sub-dimension was relatively underrepresented in the original study by Veit and Ware [23], and has subsequently been removed [].

Up to 20 responses were collected from all but one location, in order to dedicate resources towards collecting data from numerous locations throughout Trinidad. The one location for which an exception was made was Port of Spain. In Port of Spain, an additional 9 responses were obtained because Port of Spain is the capital city of Trinidad, and is home to some of the largest organizations in the Caribbean region.

Approximately ten months in the present study.

We used an attention filter to verify that respondents were reading the questions carefully. The attention filter was, “For this question kindly circle the option that represents ‘very often”’. Three respondents answered this filter question incorrectly, and thus were removed from an initial pool of 252 respondents.

The relationship between mental health and organizational commitment for men is marginally significant (p < .10). But, given the size of the relationship (β= 0.20) for a sample size of only 84 men, this relationship was still worth noting.

References

Depression [Internet]. WHO. 2017 [cited 2017 Nov 7]. Available from: http://www.who.int/mediacentre/factsheets/fs369/en/

Mental health in theworkplace [Internet].WHO.2017 [cited 2017 Nov 7]. Available from: http://www.who.int/mental_health/in_the_workplace/en/

Cortina

, Berdahl

. Sexual harassment in organizations: A decade of research in review. Handb Organ Behav. 2008.

Cunningham

, Benavides-Espinoza

. A trend analysis of sexual harassment claims: 1992–2006. Psychol Rep. 2008;103(3):779–82.

Abramson

. All the politicians recently accused of sexual harassment. Time [Internet]. 2017 Nov 22 [cited 2017 Nov 23]; Available from: http://%3A%2F%2Ftime.com%2F5033751%2Fsexual-harassment-politicians-roy-mooreal-franken%2F

Sweney

. Claims of sexual harassment at BBC spike after Weinstein scandal. The Guardian [Internet]. 2017 Nov 7 [cited 2017 Nov 8]; Available from: http://www.theguardian.com/media/2017/nov/07/claims-sexual-harassment-bbc-spike-after-weinstein-scandal

Chamber wants stronger sexual harassment laws. Trinidad and Tobago Newsday [Internet]. 2017 Dec 4 [cited 2017 Dec 8]; Available from: http://newsday.co.tt/2017/12/04/chamber-wants-stronger-sexual-harassment-laws/

EOC head calls for laws against sexual harassment. The Trinidad Guardian Newspaper [Internet]. 2017 Dec 6 [cited 2017 Dec 8]; Available from: http://www.guardian.co.tt/news/2017-12-06/eoc-head-calls-laws-against-sexual-harassment

La Rose

. Sexual harassment laws coming, says Rowley. Trinidad and Tobago Newsday [Internet]. 2017 Dec 6 [cited 2017 Dec 8]; Available from: http://newsday.co.tt/2017/12/06/sexual-harassment-laws-coming-says-rowley/

10.

McDonald

. Workplace sexual harassment 30 years on: A review of the literature. Int J Manag Rev. 2012;14.

11.

Willness

, Steel

, Lee

. A meta-analysis of the antecedents and consequences of workplace sexual harassment. Pers Psychol. 2007;60:127–62.

12.

Harned

, Fitzgerald

. Understanding a link between sexual harassment and eating disorder symptoms: A mediational analysis. J Consult Clin Psychol. 2002;70(5):1170–81.

13.

Jiang

, Hong

, McKay

, Avery

, Wilson

, Volpone

. Retaining employees through anti-sexual harassment practices: Exploring the mediating role of psychological distress and employee engagement. Hum Resour Manage. 2015;54(1):1–21.

14.

Xin

, Chen

, Kwan

, Chiu

, Yim

. Work–family spillover and crossover effects of sexual harassment: The moderating role of work–home segmentation preference. J Bus Ethics [Internet]. 2015. Available from: http://link.springer.com/10.1007/s10551-015-2966-9

15.

Weiss

, Cropanzano

. Affective events theory: A theoretical discussion of the structure, cause and consequences of affective experiences at work. Res Organ Behav. 1996;18:1–74.

16.

Colquitt

, Zapata-Phelan

. Trends in theory building and theory testing: A five-decade study of the Academy of Management Journal. Acad Manage J. 2007;50(6):1281–303.

17.

EEOC. Guidelines on discrimination because of sex. Fed Regist. 1980;(45):74676–7.

18.

Fitzgerald

, Swan

, Magley

. But was it really sexual harassment? Legal, behavioral, and psychological definitions of the workplace victimization of women. In: Donohue

, editor. Sexual harassment: Theory, research, and treatment. Boston: Allyn and Bacon; 1997, p. 315.

19.

Quick

, McFadyen

. Sexual harassment: Have we made any progress? J Occup Health Psychol. 2017;22(3):286–98.

20.

Fitzgerald

, Shullman

, Bailey

, Richards

, Swecker

, Gold

, et al. The incidence and dimensions of sexual harassment in academia and the workplace. J Vocat Behav. 1988;32(2):152–75.

21.

Thomas-Felix

. One-day workshop on sexual harassment and sextortion [Internet]. Trinidad: Industrial Court of Trinidad and Tobago; 2015 Jan [cited 2017 Nov 22] p. 5. Available from: http://industrialcourt.org.tt/Portals/0/Remarks%20-%20TTAWJ%20Workshop%20on%20Sexual%20Harassment%20%20Sextortion%2022-01-2015.pdf

22.

World Health Organization. Mental health: Strengthening our response [Internet]. WHO. 2016 [cited 2017 Dec 31]. Available from: http://www.who.int/mediacentre/factsheets/fs220/en/

23.

Veit

, Ware

. The structure of psychological distress and well-being in general populations. J Consult Clin Psychol. 1983;51(5):730–42.

24.

Ritvo

, Fischer

, Miller

, Andrews

, Paty

, LaRocca

. Multiple Sclerosis Quality of Life Inventory (MSQLI) [Internet]. New York, NY 10017: NationalMultiple Sclerosis Society; 1997 [cited 2017 Nov 21]. Available from: https://eprovide.mapi-trust.org/instruments/multiple-sclerosis-quality-of-life-inventory

25.

Adams

. Workplace Mental Health Manual for Nurse Managers. New York: Springer Publishing Company; 2015.

26.

Topa

, Morales

, Depolo

. Perceived sexual harassment at work: Meta-analysis and structural model of antecedents and consequences. Span J Psychol. 2008;11:207–18.

27.

Gunnarsdottir

, Sveinsdottir

, Bernburg

, Fridriksdottir

, Tomasson

. Lifestyle, harassment at work and self-assessed health of female flight attendants, nurses and teachers. WORK J Prev Assess Rehabil. 2006;27(2):165–72.

28.

Meyer

, Allen

. A three-component conceptualization of organizational commitment. Hum Resour Manag Rev. 1991;1(1):61–89.

29.

Meyer

, Stanley

, Herscovitch

, Topolnytsky

. Affective, continuance, and normative commitment to the organization: A meta-analysis of antecedents, correlates, and consequences. J Vocat Behav. 2002;61(1):20–52.

30.

Begley

, Czajka

. Panel analysis of the moderating effects of commitment on job satisfaction, intent to quit, and health following organizational change. J Appl Psychol. 1993;78(4):552–6.

31.

Reilly

. Exploring a paradox: Commitment as a moderator of the stressor-burnout relationship. J Appl Soc Psychol. 1994;24(5):397–414.

32.

Panaccio

, Vandenberghe

. Perceived organizational support, organizational commitment and psychological well-being: A longitudinal study. J Vocat Behav. 2009;75(2):224–36.

33.

Jain

, Giga

, Cooper

. Employee wellbeing, control and organizational commitment. Leadersh Organ Dev J. 2009;30(3):256–73.

34.

Basch

, Fisher

. Affective events–emotions matrix: A classification of work events and associated emotions. Sch Bus Discuss Pap. 1998.

35.

Barling

, Rogers

, Kelloway

. Behind closed doors: In-home workers’ experience of sexual harassment and workplace violence. J Occup Health Psychol. 2001;6(3):255–69.

36.

Barling

, Dekker

, Loughlin

, Kevin Kelloway

, Fullagar

, Johnson

. Prediction and replication of the organizational and personal consequences of workplace sexual harassment. J Manag Psychol. 1996;11(5):4–25.

37.

Magley

, Cortina

, Kath

. Stress, withdrawal, and gender in the context of sexual harassment: A longitudinal analysis. In: Annual meeting of the American Psychological Association, Washington, DC, 2005.

38.

Vogt

, Pless

, King

. Deployment stressors, gender, and mental health outcomes among Gulf War I veterans. J Trauma Stress. 2005;18(2):115–27.

39.

Richman

, Rospenda

, Nawyn

, Flaherty

, Fendrich

, Drum

, et al. Sexual harassment and generalized workplace abuse among university employees: Prevalence and mental health correlates. Am J Public Health. 1999;89(3):358–63.

40.

Richman

, Shinsako

, Rospenda

, Flaherty

, Freels

. Workplace harassment/abuse and alcohol-related outcomes: The mediating role of psychological distress. J Stud Alcohol. 2002;63(4):412–9.

41.

Bildt

. Sexual harassment: Relation to other forms of discrimination and to health among women and men. WORK J Prev Assess Rehabil. 2005;24(3):251–9.

42.

Fitzgerald

, Magley

, Drasgow

, Waldo

. Measuring sexual harassment in the military: The Sexual Experiences Questionnaire (SEQ—DoD). Mil Psychol. 2009;11:243–63.

43.

Meyer

, Allen

, Smith

. Commitment to organizations and occupations: Extension and test of a three-component conceptualization. J Appl Psychol. 1993;78(4):538–51.

44.

Podsakoff

, MacKenzie

, Podsakoff

. Sources of method bias in social science research and recommendations on how to control it. Annu Rev Psychol. 2012;63(1):539–69.

45.

Renne

. Sexual harassment and inadequate laws. Daily Express [Internet]. 2017 Jan 21 [cited 2017 Nov 7]; Available from: http://www.trinidadexpress.com/20170121/news/sexual-harassment-and-inadequate-laws

46.

Tabachnick

, Fidell

. Using multivariate statistics. 5th ed. Boston: Pearson/Allyn & Bacon; 2007, p. 980.

47.

Rosseel

. Lavaan: An R package for structural equation modeling. J Stat Softw. 2012;48(2):1–36.

48.

R Core Team. R: A language and environment for statistical computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2013.Available from: http://www.R-project.org/

49.

Podsakoff

, MacKenzie

, Lee

J-Y

, Podsakoff

. Common method biases in behavioral research: A critical review of the literature and recommended remedies. J Appl Psychol. 2003;88(5):879–903.

50.

Hair

JFJ

, Black

, Babin

, Anderson

. Multivariate data analysis. 7th ed. Upper Saddle River, NJ: Prentice Hall; 2009, p. 816.

51.

Carlson

, Wilcox

, Chou

C-P

, Chang

, Yang

, Blanchard

, et al. Psychometric properties of reverse-scored items on the CES-D in a sample of ethnically diverse older adults. Psychol Assess. 2011;23(2):558–62.

52.

Sonderen

E van

, Sanderman

, Coyne

. Ineffectiveness of reverse wording of questionnaire items: Let’s learn from cows in the rain. Baradaran HR, editor. PLoS ONE. 2013;8(7):e68967.

53.

Anderson

, Gerbing

. Structural equation modeling in practice: A review and recommended two-step approach. Psychol Bull. 1988;103(3):411–23.

54.

Fiske

, Taylor

. Social cognition. McGraw-Hill; 1991, p. 717.

55.

Jensen

, Gutek

. Attributions and assignment of responsibility in sexual harassment. J Soc Issues. 1982;38(4):121–36.

56.

Pryor

, McKinney

. Research on sexual harassment: Lingering issues and future directions. Basic Appl Soc Psychol. 1995;17:605–11.

57.

Balwant

. Socio-demographic predictors of secondary school teacher absenteeism in Trinidad. Int J Employ Stud. 2016;24(1):6–24.

58.

Singh

. Mouttet: Absenteeism a major drag on productivity. The Trinidad Guardian Newspaper [Internet]. 2017 Sep 28 [cited 2017 Nov 7]; Available from: http://www.guardian.co.tt/business/2017-09-27/mouttet-absenteeism-major-drag-productivity

59.

Buchanan

, Settles

, Hall

, O’Connor

. A review of organizational strategies for reducing sexual harassment: Insights from the U.S. military. J Soc Issues. 2014;70(4):687–702.

60.

EEOC. Policy Guidance on Current Issues of Sexual Harassment [Internet]. 1999 Jun [cited 2017 Dec 13]. Report No.: N-915-050. Available from: https://www.eeoc.gov/policy/docs/currentissues.html

61.

Williams

, Fitzgerald

, Drasgow

. The effects of organizational practices on sexual harassment and individual outcomes in the military. Mil Psychol. 1999;11(3):303–28.

62.

Gruber

. The impact of male work environments and organizational policies on women’s experiences of sexual harassment. Gend Soc. 1998;12(3):301–20.

63.

Perry

, Kulik

, Schmidtke

. Individual differences in the effectiveness of sexual harassment awareness training. J Appl Soc Psychol. 1998;28(8):698–723.

64.

Snyder

, Scherer

, Fisher

. Social organization and social ties: Their effects on sexual harassment victimization in the workplace. WORK J Prev Assess Rehabil. 2012;(1):137–50.

65.

Perry

, Kulik

, Field

. Sexual harassment training: Recommendations to address gaps between the practitioner and research literatures. Hum Resour Manage. 2009;48(5):817–37.