Do I Say or Let it Go? Emotions Predict How People Respond to Receiving Sexual Objectification at Work

Many people are reluctant to report sexual objectification at work. We tested whether emotions determine how people respond to sexual objectification at work. In Study 1 (N = 159) women recalled a time that they had experienced sexual objectification at work. Participants then rated their emotions in this situation and how they responded. Anger positively and a shame-based emotion (rejection) negatively predicted taking action against the perpetrator (active response). In contrast, shame positively predicted women blaming themselves (self-blame). Moreover, pride positively and anger negatively predicted women viewing the action positively (e.g., as flattering, benign response). In Study 2 (N = 135) women imagined themselves receiving either a highly objectifying or ambiguous comment at work. Being objectified increased negative emotions and decreased pride. Moral outrage (i.e., anger and disgust) positively whilst shame-based emotions negatively predicted active responding. Shame-based emotions positively predicted self-blame, whilst pride positively and anger negatively predicted benign responding. Therefore, emotions determine how people respond to sexual objectification at work. Promoting moral outrage and reducing other emotions (e.g., shame, fear, and pride) may make women (a) more willing to report sexual objectification at work and (b) less likely to blame themselves or view such actions positively (i.e., benign responses).

Although sexual harassment may take different forms (e.g., gender harassment, unwanted sexual attention, and sexual coercion; Dionisi et al., 2012;Karami et al., 2019;Keplinger et al., 2019), sexual objectification is prominent in many forms of sexual harassment. Recent developments in the sexual objectification literature have suggested that whether people take action against sexual objectification depends on their emotional responses towards the sexually objectifying behaviour (Guizzo et al., 2017;Shepherd, 2019;Shepherd & Evans, 2020). We advance the growing body of research assessing sexual objectification in the workplace (e.g., Gervais et al., 2016;Szymanski & Feltman, 2015;Syzmanski & Mikorski, 2017), by testing the role of different emotions in determining whether sexual objectification at work promotes an active response (e.g., reporting the incident), passive response (e.g., ignoring the incident), self-blame response (e.g., thinking the target brough it upon themselves) or a benign response (e.g., viewing the incident as flattering).
Despite legislation being in place to protect individuals, sexual harassment is still commonly experienced in the workplace. For example, a survey showed that 20% of the UK population is estimated to have experienced sexual harassment at work within the last 12 months, with this being more common among women than men (Government Equalities Office, 2020). Similar findings can be found in other parts of the world. In the US, a survey found 59% of women had experienced sexual harassment in the workplace (Graf, 2018). Similarly, a survey in Australia found 39% of women have experienced sexual harassment within the last 5 years (Australian Human Rights Commission, 2018). Therefore, sexual harassment at work is a prominent issue across the world.

Emotional Consequences of Sexual Objectification
Objectification theory (Fredrickson & Roberts, 1997) suggests that experiencing sexual objectifying behaviours is likely to result in people internalising this experience and evaluating their self-worth based on their appearance (i.e., self-objectification), which subsequently has detrimental effects on psychological well-being. This theory is wellsupported (e.g., Calogero, 2004;Karsay et al., 2018; and has been demonstrated in a variety of populations Strübel et al., 2020). Although most studies in this area have generally focused on the influence of objectification on body shame and mental health problems, a growing body of research has looked at how experiencing sexual objectification elicits a variety of emotions, including shame, disgust, anger, and anxiety (e.g., Fredrickson et al. 1998;Koval et al., 2019;Shepherd & Evans, 2020). In line with appraisal theories (e.g., Smith & Lazarus, 1993), Shepherd (2019) argues that the type of emotion that is felt following sexual objectification is likely to depend on the target's interpretation of the action. For example, people are likely to feel anger when they think someone has treated them illegitimately (e.g., 'it is wrong to objectify me'; Guizzo et al., 2017). Disgust may be felt when people believe the behaviour harms their humanity (e.g., 'this action is treating me as an object'; Rozin et al., 1999). Similarly, people may feel shame when they think an action may relate to a specific flaw (e.g., 'this makes me feel less competent'), inferiority when they believe the action relates to a global flaw (e.g., 'this makes me feel worthless') and rejection when they are concerned about the perceptions of others (e.g., 'others may look down upon me because of this'; Gausel & Leach, 2011). Finally, pride may be felt when people believe they have met a goal (e.g., 'this action shows I am attractive to other people'; Tracy & Robins, 2007).
The emotion literature also suggests that different emotions are likely to motivate people to undertake different actions in a situation. For example, moral outrage-based emotions (e.g., anger and disgust) motivate people to take action (Bastian et al., 2013), whereas rejection motivates people to take more defensive responses, such as withdrawing from the social situation (Gausel et al., 2016). Importantly, this has also been demonstrated in the context of sexual objectification. Researchers have found that people may undertake a variety of actions following sexual objectification, such as confronting the perpetrator or reporting them (active response), ignoring the incident (passive response), viewing the action positively (e.g., as a joke or believe it is flattering; benign response) or believing they were responsible for the action (self-blame response; Fairchild & Rudman, 2008). Importantly, Shepherd (2019) demonstrated that following a sexually objectifying behaviour, anger and disgust increased active responding and reduced passive responding, shame increased self-blame and pride increased benign responding. Therefore, emotions play a strong role in determining how people respond to sexual objectification.

Emotional Responses to Sexual Objectification at Work
It is important to note that Shepherd (2019) assessed responses to sexual objectification specifically in a leisure setting (i.e., at a gym). However, these processes are likely to occur in other contexts. Indeed, emotions are also a key component of sexual objectification within work situations. For example, experiencing sexual objectifying behaviours during interview situations may result in targets experiencing a variety of negative emotions, such as fear, anger, and embarrassment (Gervais et al., 2016). Similarly, staff working in sexually objectifying restaurant environments may feel a variety of negative emotions towards objectification, including disgust, anger, and anxiety (Moffitt & Szymanski, 2011). Additionally, women who experience sexual harassment at work may blame themselves and feel ashamed (Keplinger et al., 2019). Based on Shepherd (2019) we argue that the type of emotion that the target experiences will determine how they respond to sexual objectification at work. Feelings of anger are likely to result in an active response (i.e., reporting the act) and reduce passive responding (i.e., ignoring the action), whilst feelings of shame are likely to cause self-blame (i.e., wishing they had not got into the situation) and pride is likely to lead to benign responding (i.e., viewing the action as a compliment).
Although Shepherd (2019) assessed a variety of emotions, fear was not considered in this study. Fear is an important part of sexual objectification experiences (Calogero et al., 2021) and has been reported following sexual objectification at work (Gervais et al., 2016;Moffitt & Szymanski, 2011). Therefore, fear is an important emotion to consider after workplace sexual objectification. Fear is a negative emotion that is felt when people believe they are threatened or in danger (Smith & Lazarus, 1993). For example, following sexually objectifying behaviours people may feel fear about their personal safety (Calogero et al., 2021) and toward men (Szymanski et al., 2021). Similarly, fear may also be felt towards the negative possible consequences of reporting an incident of sexual objectification, such as receiving retaliation that damages the targets earnings, career prospects or social standing (Cortina & Magley, 2003;Karami et al., 2019;Minnotte & Legerski, 2019;Wilson & Thompson, 2001). Regardless of the source of the emotion, fear is associated with avoidance (Frijda, 1987). In line with this, feeling fear towards sexual objectification during a realistic interview situation has been found to prevent people from confronting the perpetrator (Woodzicka & LaFrance, 2001). Therefore, fear is likely to deter people from active responses and instead promote passive responses.

Influence of Power on Sexual Objectification at Work
It is also important to consider factors within the workplace context that may influence the emotions. This is especially important considering there are differences between the leisure setting used in previous research (Shepherd, 2019) and the workplace. For example, workplaces contain more complex power structures. High-power increases the likelihood that people with objectify others (Gruenfeld et al., 2008). Power has also been found to influence sexual objectification (McLaughlin et al., 2012;Minnotte & Legerski, 2019;Raj et al., 2020). Moreover, feeling powerless in a sexually objectifying work environments is associated with poorer psychological well-being, such as greater anxiety (Szymanski & Mikorski, 2017). Therefore, it is important to consider the influence of power in these situations.
Power is likely to influence people's emotions (Keltner et al., 2003). Indeed, people who feel they have high-power are less likely to feel negative emotions and more likely to feel positive emotions (Bombari et al., 2017). Moreover, low-power individuals may be reluctant to express their feelings towards people with higher-power (Diefendorff et al., 2010). When low-power individuals do express anger, it may backfire and result in anger being experienced by the high-power individuals (Lelieveld et al., 2012). Therefore, it is likely that people's emotional responses to sexual objectification in the workplace and their reactions to this are dependent on the power of the perpetrator. Indeed, when people experience sexual objectification from a high-power perpetrator (e.g., a line manager) they may try to neutralise feelings of anger and avoid active responses.
Other emotions may be more prominent in such situations. For example, research from within the workplace suggests that experiencing incivility from a high-power target increases feelings of shame (Hershcovis et al., 2017). Similarly, targets may be likely to feel fear towards reporting sexual objectification by senior managers as this may damage their career (North, 2016). Therefore, in contrast to leisure settings which may lack the same power dynamic, it is likely that the power relations between the target and perpetrator may influence the type of emotion that is predominantly felt, with shame and fear being more likely than anger when the perpetrator is of higher power. This higher shame should subsequently increase the likelihood of people undertaking a self-blame, whilst the higher fear should increase the likelihood of a passive response.

The Present Study
As mentioned above, previous research has suggested that in a leisure setting (i.e., a gym) the type of emotion that is experienced following sexual objectification determines how people respond. Indeed, feeling anger promotes active responses, shame promotes self-blame and pride promotes benign responses. Although this research may be applied to other settings, there are major differences between leisure settings and the workplace which means further empirical research is needed in this area. For example, unlike leisure settings, there may be a power dynamic within the workplace that influences the extent to which people are likely to feel different negative emotions and thus alter people's reactions to sexual objectification. For example, people may be more likely to feel fear at work than in leisure settings, especially when the sexual objectification is undertaken by a superior colleague (e.g., a line manager). Moreover, in the workplace people are likely to interact with the same people regularly, meaning that the objectifying behaviours may be undertaken repetitively by the same individual (Karami et al., 2019). Therefore, further research is needed to test the role of different emotions in influencing how women respond to sexual objectification in the workplace. In addition, further research is needed to assess the influence of emotions that are particularly relevant in the workplace, such as fear.
Based on this, the primary aim of this research was to assess the role of emotions (i.e., anger, disgust, shame, inferiority, rejection, pride, and fear) in predicting how women respond to sexual objectification in the workplace. Based on previous research (Shepherd, 2019), we hypothesised that (a) anger would promote active responding and reduce passive responding, (b) shame would promote self-blame, and (c) pride would promote benign responding. In line with previous research (Woodzicka & LaFrance, 2001), we also hypothesised that fear would promote passive and reduce active responding. Moreover, given the influence of power on emotions (Bombari et al., 2017;Hershcovis et al., 2017), our secondary aim was to assess the role of power on the emotions and responses. We hypothesised that fear and shame-based emotions (i.e., shame, inferiority, and rejection) would be higher and anger would be lower when the perpetrator was a senior colleague than when they were of lower or equal status to the target. This higher shame should variables were the responses to sexual objectification and included passive, self-blame, benign, and active.

Materials and Procedure
This study was approved by the authors' Institutional Review Board before data collection commenced. Participants in this online study were presented with an information sheet and asked to provide consent to take part in the study. Consenting participants were then asked to provide information about themselves (e.g., age, gender identity, ethnicity, etc.). Following this, participants were asked to describe a time that they had experienced sexual objectification at work (e.g., inappropriate comments, staring or touching parts of the participant's body). Participants were asked to write down who undertook this behaviour, what the person said or did and where the incident occurred.

Emotion Measures
Participants were then asked to rate the extent to which they felt a series of emotions during the incident. Participants were presented with a series of single emotion words and asked to rate the extent to which they felt this emotion on a five-point scale (1 = Not at all, 5 = Very much so). The anger words were 'angry', 'annoyed' and 'furious'. The disgust words were 'disgusted' and 'repulsed'. The shame words were 'ashamed', 'disgraced' and 'humiliated'. The inferiority words were 'inferior' and 'vulnerable'. The rejection scale contained three words: 'rejected', 'withdrawn' and 'alone'. The fear items were 'anxious', 'afraid' and 'worried'. Finally, there were two pride items: 'proud' and 'satisfied'. These emotion words were adapted from previous research (e.g., Gausel et al., 2012;Shepherd, 2019). For each individual emotion, the scale was calculated using the mean of the relevant items.

Behavioural Responses
Next, participants rated how they responded to the action from 1 (Not at all) to 5 (Very much so). These scales were based on previous research (Fairchild & Rudman, 2008;Shepherd, 2019). The passive response contained seven items, such as: 'I did not do anything', 'I tried to forget the whole thing' and 'I let it go' (α = 0.85). There were four selfblame items, such as: 'I realised I had probably brought it on myself', 'I realised they would not have done it if I had dressed differently' and 'I felt stupid for letting myself get into the situation' (α = 0.68). The five benign items included: 'I consider it flattering', 'I treated it as a joke' and 'I figured they must really like me' (α = 0.61). There were four increase self-blame responses, whilst this higher fear should increase passive responses to sexual objectification by senior colleagues.

Study 1
The aim of Study 1 was to assess the role of emotions in promoting the different responses to sexual objectification. Participants were asked to recall a time that they had experienced sexual objectification at work. They then rated their emotions in this situation and how they responded to the sexually objectifying behaviour. Moreover, we asked for information about the perpetrator (e.g., job position within the organisation relative to the participant), to test the secondary aim of this research.

Participants and Design
Participants were recruited through posts on social media and recruitment forums. The study was advertised as looking into sexual objectification in the workplace. To take part, participants had to be 18 years or older, identify as female, and have experienced sexual objectification at work. For ethical reasons, participants were asked not to take part if they were likely to feel distressed when discussing their body or sexual objectification. Initially, 346 participants started this study. However, 177 participants did not complete the study and were thus removed prior to data analysis. Furthermore, eight participants stated their gender identity was something other than female (five participants viewed themselves as male and three participants selected 'in another way'). In addition, two participants stated they have not been objectified. Given these participants do not meet the eligibility criteria, they were removed from the dataset. Therefore, the final sample consisted of 159 participants, aged 18-59 years (M = 29.89, SD = 9.86). Participants were more likely to be White (n = 143, 89.94%) than Asian (n = 4, 2.52%), Black (n = 3, 1.89%), mixed race (n = 7, 4.40%) or report another ethnicity (n = 2, 1.26%). Participants were also more likely to be from North America (n = 64, 40.25%) or Europe (n = 80, 50.31%) than Australia or New Zealand (n = 11, 6.92%) or Asia (n = 2, 1.26%; two participants did not state which country they lived in).
This online study used a cross-sectional design. The predictor variables were the emotional responses to the incident of sexual objectification and included anger, disgust, shame, inferiority, rejection, fear, and pride. The outcome (0.80). In contrast, the 'inferior' item from the inferiority scale did not load as strongly onto the fear latent variable (0.61). As such, the inclusion of the 'vulnerable' item may have caused this issue. In addition, the 'rejected' item did not load strongly onto the rejection latent variable (0.16). Given this, we repeated the analyses with both the 'vulnerable' and 'rejected' items removed from the model. This revised seven-factor model fit the data well: χ²(84) = 139.14, p < .001, comparative fit index (CFI) = 0.97 and root mean square error of approximation (RMSEA) = 0.07.

Outliers
Preliminary analysis revealed that there were potential univariate outliers for the pride and benign response variables (i.e., scores ± 3 standard deviations from the mean). The data were analysed with and without the outliers included. Excluding the outliers did not alter the main results. Indeed, only one significant finding from the regression analyses changed (anger no longer significantly predicted self-blame, p = .064) Therefore, we report the analyses with the outliers included.

Nature of Work
When stating their age at the time of the incident, some participants put a range of ages (e.g., '20-24'). In such instances, we based the calculation on their older age (i.e., 24 in this example). At the time of the incident, participant's ages ranged from 14 to 49 years (M = 23.32, SD = 6.26). We calculated the time since the incident occurred by subtracting their age at the time of the incident from their age. The time since the incident ranged from 0 to 42 years (M = 6.72, SD = 8.34). When the incident occurred, participants were most likely to be working in hospitality or retail (Table 1). active response items, including: 'I let them know I did not like what they were doing', 'I talked to someone about what happened' and 'I reported them' (α = 0.72). Each scale score was calculated using the mean of the relevant items.

Nature of Work
Finally, participants provided some information about themselves and their work at the time of the incident, including the type of sector, their contract, the gender and status of the perpetrator, and their age at the time of the incident. Participants were then thanked and debriefed.

Statistical Analyses
Initially, confirmatory factor analysis was used to assess the structure of the emotion constructs. Following this, we assessed the nature of the sexually objectifying incidents. This involved exploring the age of the participant when the incident occurred, the time since the incident occurred and the type of employment (e.g., contract type, full-time/parttime). We also assessed the nature of the perpetrator (e.g., gender, power status relative to target). Correlation analyses were then used to assess the association between the emotions and responses. Linear multiple regression analyses were used to assess the role of the emotions in predicting the different responses. A separate regression analysis was conducted on each of the four responses (i.e., active, passive, self-blame and benign). Finally, ANOVAs were used to assess the influence of the perpetrator's power on the emotions and responses.

Confirmatory Factor Analysis
Confirmatory factor analysis was conducted using AMOS (Version 28, Arbuckle, 2021). This analysis was conducted on the data from participants who had completed all the emotion items (n = 152). Initially, we tested whether each of the emotion items loaded onto the corresponding emotion latent variable. As such, there were seven latent variables (i.e., anger, disgust, shame, inferiority, rejection, fear, and pride). This produced a model in which the covariance matrix of the latent variables was not positively definite. Inspection of the inadmissible solution suggested a high correlation between the inferiority and fear latent variables (r = .99, p < .001). Further analyses indicated that there was a strong standardised factor loading for the 'vulnerable' item from the inferiority scale on to the fear latent variable

Role of Emotions in Predicting Responses
Correlation analyses were performed to assess the association between the variables. An active response to sexual objectification was positively associated with anger, disgust, and fear (see Table 3), whereas a passive response to sexual objectification was negatively associated with anger. A self-blame response to sexual objectification was positively associated with shame, inferiority, fear, and pride. Finally, undertaking a benign response to sexual objectification was positively associated with pride, but negatively associated with the other emotion variables. Although there were some strong associations between the variables, further analyses revealed that including all the emotions into a linear multiple regression model was unlikely to pose a multicollinearity issue, as the lowest tolerance value (0.37) was greater than 0.20 (Menard, 1995).
Next, we used linear multiple regression to assess the unique predictive power of the emotions in predicting the responses to sexual objectification. A separate multiple regression analysis was performed for each of the four responses: active, passive, self-blame and benign. In each model, the predictor variables were anger, disgust, shame, inferiority, rejection, fear, and pride.

Active Response
The emotions accounted for 23% of the variance in undertaking an active response to sexual objectification, F(7, 149) = 6.40, p < .001. This was due to active responses to sexual objectification being positively predicted by anger and fear (see Table 4). In contrast, active responses to sexual objectification were negatively predicted by rejection. The other variables did not predict undertaking an active response.

Passive Response
The emotions accounted for 22% of the variance in undertaking a passive response, F(7, 149) = 6.10, p < .001. Similar to active responses, passive responses to sexual objectification were negatively predicted by anger and fear, but positively predicted by rejection. Disgust also predicted passive responses to sexual objectification. The other emotions did not predict undertaking a passive response.

Self-Blame Response
The emotions contributed to 30% of the variance in selfblame, F(7, 149) = 9.18, p < .001. This was due to selfblame responses to sexual objectification being positively predicted by shame and pride, but negatively predicted by At the time of the incident, participants were most likely to be in a permanent contract, and either working full or part time ( Table 2). The perpetrator was most likely to be male. The perpetrator was more likely to be either a superior colleague, an equal status colleague or a customer/client than a subordinate colleague.  Post-hoc Tukey tests indicated that this was due to targets feeling greater inferiority when the perpetrator was a superior to when they were a subordinate (p = .032). All other post-hoc comparisons were non-significant. The perpetrator's power did not have a significant effect on any of the other emotions. However, the perpetrator's power did have a significant effect on active responding. Further analysis indicated that there was a trend for targets to be less likely to undertake an active response when the perpetrator was their superior to when they were of equal status, a subordinate colleague, or a customer. However, the post-hoc Tukey tests indicated all comparisons were non-significant.

Discussion
Study 1 assessed the role of different emotions in predicting how people react to instances of sexual objectification in the workplace. In line with our hypotheses, Study 1 found that (a) anger positively predicted active and negatively predicted passive responses to sexual objectification, (b) shame positively predicted self-blame responses to sexual objectification, and (c) pride positively predicted benign responses anger. The other variables were non-significant predictors of self-blame responses.

Benign Response
The emotions accounted for 28% of the variance in undertaking a benign response, F(7, 149) = 8.39, p < .001. Benign responses to sexual objectification were negatively predicted by anger, but positively predicted by pride.

Role of Perpetrator's Position on Emotions and Responses
As mentioned above, our secondary aim was to assess the role of the perpetrator's power on the emotions and responses. From our data, we created four groups: perpetrator superior to target (including direct line managers and not direct line managers), perpetrator of equal status, subordinate perpetrator, and customer/client. Participants who selected 'other' were not included in this analysis. We conducted a series of ANOVAs on the emotions and responses. The perpetrator's power had a significant effect on inferiority ( Table 5). Note. *p < .05, **p < .01, ***p < .001. Note. Different subscript indicates a significant different based on Tukey's HSD. Significant difference between superior and equal status condition for benign response is non-significant when outliers are removed (p = .138).

Study 2
Study 2 aimed to replicate the findings of Study 1, whilst addressing the limitations. There were four main differences between the studies. First, we decided to recruit participants through a course credit system to reduce attrition rates. Second, we used a vignette study where we specified the nature of the incident to ensure consistency across all participants. Third, we manipulated the extent to which participants had experienced sexual objectification by including a vignette that either included an overt sexually objectifying incident (high objectification condition) or a more ambiguous incident (low objectification condition). Finally, given that the perpetrator's power did not strongly influence the emotions, this factor was not included in Study 2.

Participants and Design
Participants were recruited through a course credit system at a university in the UK. To be eligible, participants had to be 18 years or older and view themselves as female. For ethical reasons, participants were informed not to take part if they were likely to feel distress considering their body or sexual objectification. As expected, the use of a course credit system for Study 2 resulted in a low attrition rate. Of the 142 participants who started the study, only seven were removed for not completing the survey. There were three participants who did not complete the study in the low objectification condition, two participants in the high objectification condition, and two participants who left the study before being assigned to a condition. Therefore, the final sample consisted of 135 participants who identified as female, with an age range between 18 and 51 years (M = 20.96, SD = 5.15).
The study used a between-participants design, with two conditions (low versus high objectification). The dependent variables were the same responses to objectification as Study 1 (i.e., active, passive, self-blame and benign). The mediating variables were the same emotions as Study 1 (i.e., anger, disgust, shame, inferiority, rejection, pride, and fear).

Materials and Procedure
The authors' Institutional Review Board provided ethical approval before we started collecting data. After giving consent, participants were asked to provide some information to sexual objectification. In addition to these hypotheses, we also found that rejection negatively predicted active and positively predicted passive responses, which is consistent with the idea that rejection promotes withdrawal and avoidance (Gausel & Leach, 2011;Gausel et al., 2012Gausel et al., , 2016. Based on previous research (Woodzicka & LaFrance, 2001), we hypothesised that fear would negatively predict active and positively predict passive responses. However, in contrast to this, we found that fear positively predicted active and negatively predicted passive responses. Although fear is often associated with avoidance (Frijda, 1987), previous research has suggested that fear may promote action when people feel this action will resolve a threat (Peters et al., 2013;Shepherd & Smith, 2017). In this context fear may promote active responses in order to tackle objectification and prevent it from happening again.
Our secondary aim was to assess how the perpetrator's power influenced the emotions and subsequent responses. We found the perpetrator's power influenced inferiority, which is a shame-based emotion. Targets felt more inferiority when the perpetrator was a superior to when they were a subordinate colleague. However, in contrast to our hypotheses and previous research (Bombari et al., 2017), we found that the perpetrator's power did not influence anger or fear. This may reflect the fact that people may feel negative emotions, regardless of the perpetrator's power. However, power may influence people's willingness to express these emotions, as indicated by some previous research (Diefendorff et al., 2010). Therefore, it may not be the emotion, but instead the willingness to express the emotion which is influenced by the power of the perpetrator.
These findings extend previous research (Shepherd, 2019) by demonstrating the role of emotions in predicting how people respond to sexual objectification in a workplace context and the influence of different emotions (i.e., fear and rejection) on responses to objectification. However, it is important to consider the limitations of this research. First, the attrition rate for this study was high, which could potentially bias the findings (Zhou & Fishbach, 2016). Second, although asking participants to recall an instance of sexual objectification increases the ecological validity of the findings, it also resulted in a diverse range of incidents being recalled that varied based on the workplace context, perpetrator, type of incident, etc. Finally, the correlational nature of this study meant that a causal direction could not be inferred. For example, it could be the case that undertaking an active response increases anger rather than viceversa. Therefore, the aim of Study 2 was to address these limitations.
performed to assess the effect of the objectification manipulation on the emotions and responses. Correlation analyses were then used to assess the association between the variables. Finally, indirect effect analyses were conducted to determine whether there was an effect of the manipulation on the responses via the emotions. The indirect effect analyses were conducted using the Process Macro (Version 4.2, Hayes, 2022). A separate indirect effect analysis was performed for each of the responses (i.e., active, passive, self-blame and benign).

Confirmatory Factor Analysis
Using Amos (Version 28, Arbuckle, 2021), confirmatory factor analysis was conducted to assess the structure of the emotion constructs. This analysis was conducted on the data from participants who had completed all the emotion items (n = 133). We first tested a model in which the emotions were seven separate constructs. Unfortunately, this produced a solution containing a covariance matrix that was not positive definite. In contrast to Study 1, this issue remained when the 'vulnerable' item was removed from the inferiority scale. In the inadmissible solution that was produced, there were strong associations between the latent anger and disgust variables (r = .97, p < .001). Anger and disgust contribute to the concept of moral outrage (Bastian et al., 2013). Therefore, there was a statistical and theoretical rationale for combining these emotions to form a measure of moral outrage. In addition, the solution suggested that there were strong associations between the latent shame and rejection (r = .96, p < .001), inferiority and rejection (r = 1.00, p < .001), inferiority and fear (r = 1.02, p < .001), and rejection and fear latent variables (r = 1.00, p < .001). This suggested that there was a statistical rationale for combining the shame-based emotions (i.e., shame, inferiority, and rejection) with fear to form an overall construct of shame and fear. Importantly, research has indicated that shame narratives of sexual victimisation may include a fear of being blamed and a fear of public scrutiny (Weiss, 2010). Therefore, there was both a statistical and theoretical rationale for combining the fear and shame-based emotions. Based on this, we next tested a model containing three factors: moral outrage (i.e., anger and disgust), shame and fear (i.e., shame, inferiority, rejection, and fear), and pride. This model fit the data well: χ²(132) = 222.69, p < .001, CFI = 0.95 and RMSEA = 0.07.
This revised three-factor model was compared against a two-factor model in which the negative emotions loaded on the first factor and the positive emotions (i.e., pride about themselves. We then manipulated sexual objectification. Participants in both conditions were asked to imagine themselves in the following situation: Please imagine that you are at work. Whilst making a cup of coffee, you are approached by a male colleague who is of equal status to you. For a few minutes you discuss the previous day at work.
Participants in the low objectification condition were then asked to imagine the colleague said the following ambiguous comment: "You look good in that outfit. It really suits you." By contrast, participants in the high objectification condition were asked to imagine that the colleague said: "You look good in that outfit. It really shows off your great curves." Participants then completed the following measures in the order presented.

Manipulation Check
We tested the extent to which participants believed they had been objectified using four items adapted from Teng et al. (2015). These items were: 'In the scenario, I felt more like a body than a real person', 'In the scenario, I felt my body and my personality were separate things', 'In the scenario, I was viewed more as an object than a human being' and 'It was only my body, not my personality, that caught this man's attention' (α = .90). These items were rated on a 5-point Likert scale (1 = Strongly disagree, 5 = Strongly agree).

Emotions
Next, participants were asked to rate how they would feel in this situation. This was rated using the same emotion items as Study 1.

Responses
The behavioural response items matched those completed with Study 1, except that the tense was changed to ensure they were suitable for the vignette nature of this study (e.g., 'I did not do anything' was changed to 'I would not do anything'). The active (α = .74), passive (α = .80), self-blame (α = .74) and benign response scales were reliable (α = .78). After completing these measures, the participants were then thanked and debriefed.

Statistical Analyses
First, confirmatory factor analysis was undertaken to assess the structure of the emotion constructs. ANOVAs were then outliers may bias the results of this study. Therefore, below we report the analyses with the outliers removed. The outliers were regarded as missing data. Given that pairwise deletion was used, the number of participants in each analysis varied between 131 and 135 participants.

Effect of Manipulation
People felt they had been more objectified in the high (M = 3.50, SD = 0.93) than the low objectification condition (M = 2.62, SD = 0.96), F(1, 133) = 28.88, p < .001, η p 2 = 0.18. Therefore, the manipulation was successful. Importantly, we found that being objectified increased feelings of shame and fear as well as moral outrage (see Table 6), and reduced feelings of pride. As expected, active responses to sexual objectification were higher in the high compared to low objectification condition, whereas benign responses to sexual objectification were higher in the low compared to high objectification condition.

Association Between Variables
Next, we looked at the correlations among the variables. An active response to sexual objectification was positively associated with moral outrage (see Table 7). Passive responding was not associated with any of the emotions. Self-blame responses to sexual objectification were positively associated with shame and fear as well as moral outrage. Finally, benign responses to sexual objectification were positively associated with pride, but negatively associated with shame and fear as well as moral outrage. The emotions were strongly intercorrelated; however, including all these emotions in a multiple regression analysis created a lowest tolerance value of 0.33. Given this was above 0.20, the data was unlikely to be biased by multicollinearity (Menard, 1995).

Outliers
Preliminary analyses found outliers for the shame and fear, self-blame, and active response variables (i.e., scores ± 3 standard deviations from the mean). We analysed the data both with and without the outliers included. There were variations between these analyses, suggesting that including

Passive Response
Passive responses to sexual objectification was positively predicted by shame and fear (Fig. 2), thus suggesting a potential indirect effect of the manipulation via shame and fear. However, the confidence intervals for this indirect effect included zero (95% CI [-0.01, 0.41]). As such, the indirect effect was non-significant.

Self-Blame Response
Self-blame responses to sexual objectification were positively predicted by shame and fear (see Fig. 3). This suggested a possible indirect effect of sexual objectification on self-blame via shame and fear. Indeed, the confidence internals for this indirect effect did not contain zero (95% CI [0.19, 0.72]), indicating that it was significant. Objectification increased feeling of shame and fear, which subsequently increased self-blaming.

Benign Response
Benign responses to sexual objectification were negatively predicted by moral outrage, but positively predicted by pride (see Fig. 4). Importantly, further analysis demonstrated a significant indirect effect from sexual objectification to benign responding via moral outrage (95% CI [-0.52, -0.15]) and pride (95% CI [-0.33, -0.03]). The findings reflect the fact that being objectified increased feelings of moral outrage and decreased feelings of pride. Pride subsequently increased benign responding, whilst moral outrage reduced benign responding.

Indirect Effect Analysis
Following this, we tested the role of the emotions as mediators of the effect of sexual objectification on the responses. In this analysis, the sexual objectification manipulation was the independent variable, the emotions were the mediators (i.e., moral outrage, shame and fear, and pride), and the responses to sexual objectification were the dependent variables (i.e., active, passive, self-blame, and benign responses). Although passive responses to sexual objectification was not associated with the emotions in the correlation analyses (see Table 7), it was still possible for an indirect effect to be present if a relationship between an emotion and passive responding was significant after controlling for the manipulation and other emotions (i.e., in the regression analyses undertaken as part of the indirect effect analyses). Therefore, despite the nonsignificant associations, we assessed the indirect effect from the manipulation to passive responses to sexual objectification via the emotions.

Active Response
Active responses to sexual objectification were positively predicted by moral outrage, but negatively predicted by shame and fear (see Fig. 1), thereby suggesting an indirect effect of the manipulation via these emotions. Indeed, the confidence intervals did not contain zero for the indirect effect via moral outrage (95% CI [0.17, 0.66]) and shame and fear (95% CI [-0.44, -0.02]). This suggests both indirect effects were significant. Therefore, being objectified increased feelings of moral outrage as well as shame and fear. Subsequently, anger positively whilst shame and fear negatively predicted active responding.

Discussion
Study 2 supported Study 1 in demonstrating that (a) moral outrage-based emotions (i.e., anger and disgust) positively predicted an active response to sexual objectification, (b) shame-based emotions (i.e., shame and fear) negatively predicted an active response and positively predicted a passive response to sexual objectification, (c) shame-based 2 may have resulted in people being less worried about their personal safety and more concerned about the damage the incident could cause to their reputation. Given that this fear was related to their social image and reputation, the shamebased emotions and fear may have been more likely to load onto a single factor in Study 2 than Study 1.

General Discussion
In recent years there has been a growing interest in the emotions that are experienced following sexual objectification (e.g., Koval et al., 2019;Shepherd & Evans, 2020). Across two studies, we tested whether emotions determine how people respond to sexual objectification at work. Based on previous research (e.g., Shepherd, 2019), we hypothesised that (a) anger would positively predict active responses to sexual objectification, (b) shame would positively predict self-blaming responses to sexual objectification, and (c) pride would positively predict benign responses to sexual objectification. These hypotheses were supported in Study 1. Although the emotion variables varied in Study 2, we found support for these hypotheses as (a) moral outragebased emotions (i.e., anger and disgust) positively predicted active responses to sexual objectification, (b) the shamebased emotion (i.e., shame and fear) positively predicted self-blame responses to sexual objectification, and (c) pride positively predicted benign responses to sexual objectification. Therefore, these studies contribute to the growing body of research assessing the emotional responses to sexual objectification by strongly demonstrating that the emotions felt following sexual objectification at work influence how people are likely to act. In addition to these hypotheses, across both studies we found that shame-based emotions negatively predicted an active response and positively predicted a passive response to sexual objectification. Indeed, Study 1 found that active responses to sexual objectification were negatively and passive responses to sexual objectification were positively predicted by rejection, which is a shame-based emotion. Similarly, Study 2 found that the combined shame and fear construct negatively predicted active and positively predicted passive responses to sexual objectification. In the workplace, a barrier to reporting sexual harassment is the concern that it may be detrimental to people's social standing (Cortina & Magley, 2003;Foster & Fullagar, 2018). Importantly, potential damages to social bonds have been found to promote shame-based emotions (Løkkeberg et al., 2021). As such, these shame-based emotions are likely to be contributing to the unwillingness to report instances of sexual objectification at work. Rejection stems from the belief that the target's social image may be damaged in the emotions positively predicted self-blame in response to sexual objectification, (d) moral outrage-based emotions negatively predicted benign responding to sexual objectification, and (e) pride positively predicted benign responding. Importantly, these findings were replicated despite the different methodologies used between these two studies. This suggests that the high attrition rates in Study 1 did not bias the results. Moreover, the use of an experimental approach in Study 2 ensured that a causal direct could be inferred, thereby showing that the emotions are likely to predict subsequent responses to a sexually objectifying behaviour.
In Study 2 we found that anger and disgust loaded onto a single factor that we labelled as moral outrage. Similarly, in Study 2 the shame-based emotions (i.e., shame, inferiority, and rejection) and fear loaded onto a single factor, which we labelled shame and fear. Research has suggested that anger and disgust contribute to the emotion of moral outrage (Bastian et al., 2013). Similarly, research has indicated that shame-based narratives of sexual victimisation, may include the fear of being blamed for the incident and the fear of public scrutiny (Weiss, 2010). Given this and the results of the confirmatory factor analysis, there was both a statistical and theoretical rationale for combining these emotions.
It is important to consider why the structure of the emotions in Study 2 did not match Study 1. This may have been due to a variety of different reasons. For example, in Study 1 participants provided a description of a time when they had been objectified. However, in Study 2 participants were asked to imagine themselves in a situation where they received a comment from a man. Given that we presented participants with the scenario in Study 2, the extent to which people felt they had been objectified may have been more subjective than when they provided their own scenario in Study 1. This subjectivity may have made it more difficult for the participants to differentiate between the emotion constructs, especially given that the items varied based on a single word (see Gausel & Salthe, 2014), thereby leading to stronger correlations between the items. In addition, the nature of the sexually objectifying incidents may have contributed to the discrepancies in the studies. In Study 1, participants were asked to describe and evaluate a time they had been sexually objectified. As such, participants could report incidents in which the perpetrator had stared at their body, made sexual remarks about their body, engaged in sexual coercion, or touched them. As a result, in Study 1 people may feel fear towards their personal safety (see Calogero et al., 2021) or potential damage to their reputation through being blamed for the incident or public scrutiny (Weiss, 2010). In contrast, in Study 2 participants only rated their feelings in a situation where they received an appearance-related comment from a man at work. The lack of any physical contact or overtly sexual behaviour in Study flattering or a joke. Observing such benign responses to sexual objectification may result in the perpetrator regarding their actions as flattering and appropriate, thereby increasing the likelihood of the action being repeated in the future to the target and other women.
These findings contribute to the growing body of literature demonstrating the importance of assessing women's emotions towards objectification and the consequences of these emotions (Koval et al., 2019;Shepherd, 2019;Shepherd & Evans, 2020). Importantly, this research also makes a substantial contribution to the sexual harassment at work literature. Much of the sexual harassment at work literature has focused on the influence of organisational factors, such as a power dynamic (McLaughlin et al., 2012;Minnotte & Legerski, 2019) and the ratio of males to females (Raj et al., 2020). The current studies add to the growing body of research looking at processes that occur within women following sexual objectification and supports other research in demonstrating the effectiveness of applying sexual objectification theories and research into the work context (e.g., Gervais et al., 2016;Moffitt & Szymanski, 2011).

Limitations and Future Research Directions
It is important to consider the limitations of this research. First, although we were able to infer a causal direction of the relationships, we were not able to infer a causal effect of the emotions on the responses. In order to do this, we would need to manipulate the emotions and assess their effect on the responses (for a discussion, see Mackinnon & Pirlott, 2015). Therefore, further research is needed to determine the causal effects of the emotions on the responses. Second, this research was conducted on people who view themselves as female. This group commonly experiences sexual objectification at work (Raj et al., 2020). However, research has demonstrated the harmful consequences of sexual objectification in both cisgender and transmen and transwomen (Strübel et al., 2020). Therefore, it is important for future studies to extend these findings to other populations. Third, this research focused on the role of the emotions on the responses. However, it is also important to assess the factors that influence the emotions. Although Study 1 assessed the role of the perpetrator's power, further factors need to be considered. For example, the type of emotion that people may feel may vary depending on whether the instance of sexual objectification was a one-off event or a repeated incident, the objectification occurred in a private or public situation, and whether bystanders supported the target. As such, further research is needed to assess these factors.
Additionally, it is worth evaluating how the emotions were measured in this research. The emotions were eyes of other people and this motivates people to withdraw from others (Gausel et al., 2012(Gausel et al., , 2016. Moreover, people may feel fear towards the damage that sexual victimisation may have on their image (Weiss, 2010). These shame-based emotions are, therefore, likely to stem from the potential damage of the sexual objectifying behaviour to the target's social image and social bonds. Given that these emotions are based around the negative perceptions of other people, it is likely that the target will try to hide the sexually objectifying incident in order to protect their image and social bonds. This desire to hide the incident may have made people unlikely to take an active response to sexual objectification and instead take a passive response. As such, these shame and fear-based emotions may contribute to people's reluctance to report sexual objectification in the workplace.
Interestingly, Shepherd (2019) hypothesised that rejection would negatively predict active responses, but found that this relationship was non-significant. The difference between the findings of the current studies and Shepherd (2019) may be due to the different contexts used in these studies. Rejection may be more likely to predict active and passive responses in the workplace than the leisure setting used in this previous research (Shepherd, 2019). This is because the concern about damaging one's social standing may make rejection more likely to deter active and promote passive responses in the work context in which the target is interacting with colleagues on a regular basis. In contrast, people are less likely to know or interact with others in a leisure context, such as a gym. This may make people less concerned about how reporting instances of sexual objectification may influence their social image, thereby attenuating the relationship between rejection and active responding.
Although moral outrage, shame-based emotions and fear are more likely to be experienced following sexual objectification than pride, it is also important to consider the positive relationship between pride and benign responses to sexual objectification. Objectification theory states that society evaluates women based on their appearance and that the frequent exposure to sexual objectification may result in women internalising this perception and evaluating themselves based on their appearance (i.e., self-objectification). Based on this self-objectification process, women may hold the goal of looking attractive to other people. As mentioned above, pride is felt when people believe they have achieved a goal (Tracy & Robins, 2007). As such, some women may feel pride when they receive some forms of sexually objectification (e.g., comments about their appearance) as they may perceive that it indicates they are meeting their goal of looking attractive to others. This positive feeling may result in these women interpreting such sexually objectifying behaviours positively, thereby resulting in a benign response to sexual objectification and the target viewing this action as sexually objectifying behaviour. As a result, targets may be less likely to view others as a source of criticism and more as a source of support. This is likely to reduce the extent to which targets feel they may be blamed or negatively scrutinised by others for reporting instances of sexual objectification. As a result, this intervention should reduce feelings of shame-based emotions (e.g., rejection) and fear towards their social image being damaged, thereby increasing the likelihood of sexual objectification at work being reported. Bystander intervention training may alter other emotions as well. For example, this training may increase the perception in colleagues within an organisation that sexual objectification at work is illegitimate and harmful to the target. This may mean that people are more likely to view this action as wrong, leading people to be more likely to experience moral outrage-based emotions and less likely to feel pride. This should, in turn, increase the likelihood of an active response and reduce benign responses to sexual objectification.
In addition to bystander intervention training, it is also important to reduce the likelihood of people experiencing sexual objectification in the first place. Recent research has demonstrated the effectiveness of sensitising web campaigns in reducing harassment and sexual coercion (Guizzo & Cadinu, 2021). Therefore, it may be useful to undertake a two-pronged approach. Using sensitising web campaigns may reduce the likelihood of sexual objectification occurring at work. By also applying bystander intervention training, targets may feel supported by others when objectification occurs.

Conclusion
In conclusion, across two studies we demonstrated that emotions play a role in determining how people are likely to respond to sexual objectification at work. Feeling moral outrage-based emotions (e.g., anger) is likely to promote active responses whereas feeling shame-based emotions is likely to deter them. Moreover, shame promotes self-blame and pride promotes benign responses. It may be effective to harness emotions to encourage responses to sexual objectification that are likely to be most effective for combatting sexual objectification, such as active responses. As such, it may be useful to find ways to promote moral outragebased emotions and reduce the likelihood of people experiencing emotions that promote non-optimal responses to sexual objectification (e.g., shame-based emotions, fear, and pride). This may help to reduce the negative consequences of sexual objectification and the likelihood that the perpetrator will undertake such actions in the future.
Funding The authors did not receive funding to conduct this research. measured using items that varied based on a single word. Although this ensures uniformity in how the constructs were measured, it may create subjectivity around the meaning of the items as they do not place the emotion within a specific context (Gausel, 2014). Including the emotion word into a sentence that gives the construct meaning (e.g., 'I would feel rejected because others may look down on me') can increase the extent to which participants rate the intensity of the emotion and reduce associations between different emotion constructs (Gausel & Salthe, 2014). As such, adding meaning into each of the emotion items may increase the likelihood of people endorsing the items and their ability to differentiate between the different emotions. This may reduce the likelihood of items measuring different emotions loading onto a single construct, as they did in Study 2.

Practice Implications
Previous research has identified numerous barriers that prevent people from reporting instances of sexual objectification at work, including concerns of damaging their career prospects, earnings, or social standing (Cortina & Magley, 2003;Foster & Fullagar, 2018;Minnotte & Legerski, 2019). However, the present research suggests that another important barrier to reporting objectification may be the emotion that the target feels. Although anger and moral outrage may promote active responses (e.g., reporting the perpetrator), people may be reluctant to report the incident when they feel shame-based emotions. This suggests that reducing shame-based emotions may be beneficial in motivating people to undertake an active response towards sexual objectification at work. Moreover, given that pride promotes benign responses to sexual objectification and given that such responses may encourage perpetrators to repeat sexually objectifying actions in the future, it may also be useful to reduce feelings of pride following sexual objectification. Therefore, it is important to find ways to harness emotions in order to promote active responses to sexual objectification (i.e., anger and moral outrage), as these are likely to demonstrate to the perpetrator and to others that such actions are harmful to the target and will not be tolerated.
An effective way to adapt people's emotions may be to apply interventions that change the perceptions of the situation. For example, bystander intervention training has been found to be effective for changing people's attitudes about sexual violence and increasing their likelihood of helping (Mujal et al., 2021). Therefore, it may be beneficial to provide bystander intervention training for sexual harassment within the workplace. This should increase the likelihood of colleagues supporting and helping targets of Data Availability The data is available from the corresponding author on request.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law. Zhou, H., & Fishbach, A. (2016). The pitfall of experimenting on the web: How unattended selective attrition leads to surprising (yet false) research conclusions. Journal of Personality and Social Psychology,111(4), 493-504. https://doi.org/10.1037/ pspa0000056.
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.