Gendered wording in job advertisements: The presence of a masculine or feminine goal orientation and its effect on job appeal
Annelieke van den Berg (10545042)
Master's Thesis
Graduate School of Communication Master's program Communication Science
Supervisor: Lise van Oortmerssen Date: 29 January 2016
Abstract
Today, women make up a larger part of the student population than ever before.
Consequentially they are also an increasingly important pool of applicants for positions aimed at starters with a university background. Previous research has found, however, that women are sensitive to wordings in job advertisements with typically male or female connotations, and that the presence of masculine gendered wording can demotivate women to apply for specific positions. In this study, 99 job advertisements for applicants with a university background and limited working experience were analyzed through an inductive content analysis to investigate to which degree gendered wording is actually present in current job advertisements aimed at starters. Results showed that almost 90% of the analyzed
advertisements contained at least one example of a type of gendered wording. One of the types of gendered wording which was commonly present in job advertisements was the description of either a masculine or feminine goal orientation. To further examine this new type of gendered wording an experiment was conducted with 179 near or recently graduated students. In this experiment participants read a job advertisement which contained gendered wording (i.e. a masculine or feminine goal orientation) in either the description of the company or of the desired applicant, and were asked to indicate job appeal and anticipated belonging. The presence of a feminine goal orientation was found to positively influence job appeal for women when compared to the presence of a masculine goal orientation, but only when the gendered wording was present in the description of the company. The results did not indicate that the effect of gendered wording on job appeal was mediated by belonging. Also, the gendered wording did not influence job appeal or belonging for men.
Keywords: gender stereotypes, gendered wording, job advertisements, job appeal, belonging
Gendered wording in job advertisements: The presence of a masculine or feminine goal orientation and its effect on job appeal
Over the last decades increasingly more women have been obtaining university degrees. In The Netherlands the amount of female students has more than doubled over the last thirty years and currently there are several thousand more female students enrolled in universities than men (Van der Laan, 2013). Because of this growth, recently graduated females are becoming an increasingly important pool of applicants for starting positions. However, research has found that women are sensitive to gendered wording in job advertisements and that the presence of words with masculine connotations can discourage them from applying (e.g. Bem & Bem, 1973; Born & Taris, 2010; Gaucher, Friesen, & Kay, 2011). Although these findings underline the importance of studying gendered wording, since it can decrease job appeal for an important pool of applicants, the understanding of the presence and effects of gendered wording is far from complete.
First of all, there is a scarcity of research using content analysis to examine in what way gendered wording is actually present in job advertisements. One study reported that it is still the case that job vacancies are announced with masculine nouns such as ‘frontman’ (Horvath & Sczesny, 2015), but since this study only focused on gendered wording present in job titles no conclusions could be drawn about the presence of gendered wording in the remainder of the advertisement. In a different study performed by Gaucher et al. (2011), complete job advertisements were analyzed via an automated content analysis, in which it was counted how often words from a pre-specified list of masculine and feminine words were present in these advertisements. It was found that both masculine and feminine gendered wording was commonly present, but since an automated vantage point was used to get to this conclusion possibly important nuances in the form and function of gendered wording were not taken into consideration.
One way in which an automated content analysis is limited is that it only considers the job advertisement as a whole and does not take into account that there are several sections to a job advertisement, such as job description and company description. However, the effect of gendered wording might differ based on the section in which the wording appears. Moreover this type of analysis fails to consider that it is likely that several types of gendered wording exist and that the function of gendered wording might vary depending on how it is used and presented. The analysis also neglects to note instances of explicit references to gender. The discussions about striving towards equal representation of the sexes in the workplace have increased (for example the discussion about enforcing a quota of women in top positions, Davidson, 2015; Dekker, 2015) and may have contributed to the acceptance and usage of positive gender discrimination. Dutch legislation, for instance, allows employers to indicate a preference for women under certain conditions (Rijksoverheid, n.d.). It would therefore be interesting to investigate whether job advertisements are indeed presenting this type of explicit positive discrimination.
Overall, analyzing gendered wording in job advertisements from a more qualitative vantage point seems to be called for, so that the placement and function of gendered wording can be evaluated and occurrences of explicit gendered wording and positive gender
discrimination can be detected. The first study that is presented in this article, an inductive qualitative content analysis, is aimed at addressing all these issues. This content analysis was furthermore conducted in order to find a new vantage point from which to examine the effects of gendered wording in job advertisements. Until now, research has mainly focused on
gendered wording as typically masculine or feminine behaviors and traits (e.g. Born & Taris, 2010). However, the content analysis in this article discovered that also a different type of gendered wording was commonly present in job advertisements, namely goal orientation. It is
important to examine whether the presence of a masculine goal orientation also decreases job appeal for women, similarly to the presence of typically masculine behaviors and traits.
Finally, it remains unclear from previous studies if the effect of gendered wording differs depending on the location within the job advertisement where it is present. Gaucher et al. (2011) suggested that this would be relevant to examine since people might make different inferences based on the section of the job advertisement in which gendered wording appears. As an example, Gaucher et al. (2011) proposed that gendered wording in the description of the company might influence the perception of the company as a whole, whereas its presence in the description of the job tasks might only affect the perception of the job itself.
The second study that is presented in this article thus aims to investigate whether location moderates the effect of gendered wording on job appeal. Moreover, it examines whether the newly discovered type of gendered wording, goal orientation, has a similar effect on job appeal as the previously researched behaviors and traits. To this end, an experiment was conducted which considered how communicating a masculine or feminine goal
orientation affected job appeal and anticipated belonging depending on its location within the advertisement.
Theoretical Framework
Gender Stereotypes and Gendered Wording
From a historical perspective men have always had different roles in society than women. Compared to men, women are more likely to hold positions with lower status and authority: they are more often homemakers and less likely to have paid work (Eagly & Steffen, 1984). Stemming from this divide between gender roles are gender stereotypes, which are still prevalent in our society today. These stereotypes are widely held beliefs about typical attributes of men and women, that are very resistant to change (Heilman, 2001).
Social role congruity theory posits that people attribute role-specific qualities to the social group which they typically expect to fulfill a certain role (Eagly & Karau, 2002). As such, men are expected to be agentic, since men are more commonly seen in positions of higher status, whilst women are expected to be communal, since they are more often seen as homemakers and the prime caretakers of children (Gaucher et al., 2011; Johnson, Murphy, Zewdie, & Reichard, 2008). As a result of this divide, certain words in our language have also become characterized as typically male or female. For example, men are seen as independent, decisive and dominant, whereas women are seen as kind, helpful and sympathetic (Heilman, 2001). One of the consequences of these stereotyped words is that leadership qualities, which are often asked for in job advertisements, are more readily ascribed to men than women (Eagly & Karau, 2002; Horvath & Sczesny, 2015).
The presence of gendered wording (i.e. words with stereotypically male or female connotations) in job advertisements can hence influence in how far men and women are equally attracted to the advertised position. Even if people do not personally endorse gender stereotypes, their thoughts, feelings and behavior might still be subconsciously affected when stereotypical associations are primed by gendered wording (Jost & Kay, 2005). If a job advertisement lists many typical male qualities, women may perceive a lack of fit with the advertised position (Horvath & Sczesny, 2015). This effect is partly attributable to the oppositional nature of gender stereotypes: women are seen to lack what men have and the other way around (Heilman, 2001). Moreover it generally holds that, even if one is not consciously aware of accepting the stereotypes which are primed, the members of
subordinated groups (in this case women) are complicit to their own subordination (Jost & Kay, 2005). This means that, whilst reading a job advertisement containing a lot of masculine words, women might subconsciously infer that they do not possess the advertised qualities to the same degree as men and hence feel less addressed by the advertisement.
Implicit and Explicit Gendered Wording
Within this article a distinction is made between implicit and explicit gendered wording. This distinction is used when describing previous research and it also forms the basis of an important categorization in the qualitative content analysis as presented in Study 1. This section aims to clarify how implicit and explicit gendered wording can be differentiated from each other and how it is understood in the remainder of the article.
Gendered wording is explicit when linguistic forms are used which only refer to either men or women. Many languages have grammatical gender, meaning that gender can be recognized through a difference in grammatical form (Horvath & Sczesny, 2015). A clear example of gender-marked words are pronouns (e.g. he/she). Gender marking can also be present in nouns and is dependent on whether or not a language is gender-neutral. English is an example of a natural-gender language, where human nouns do not vary grammatically based on gender (Horvath & Sczesny, 2015), whilst in Dutch human nouns are usually gender-marked. For example, in English ‘a teacher’ is both masculine and feminine, whereas in Dutch a teacher can be ‘leraar’ in its masculine form or ‘lerares’ in its feminine form. Explicit gendered wording thus refers to the presence of gender-marked nouns or pronouns.
When words are not gender-marked but do reflect cultural stereotypes about men and women, gendered wording is implicit (Gaucher et al., 2011). For example, the word
‘dominant’ can be used to describe both men and women, and one cannot derive from the statement ‘a dominant person’ whether it signifies a man or woman. It is only through society’s understanding of gender roles that ‘dominant’ is seen as typically male rather than typically female (Heilman, 2001). Words can thus be understood as implicit gendered wording when they are not inherently masculine or feminine in form but when they are more readily ascribed to members of one gender because of their connotations.
Studies on Gendered Wording in Job Advertisements
Until the mid-1970s it was common for job advertisements in newspapers to be sorted by gender and preferences towards a specific gender were often explicitly stated (Pedriana & Abraham, 2006). Against this backdrop, Bem and Bem (1973) conducted the first study about the effects of gendered wording in job advertisements. In the first of two experiments,
students were asked to rate job advertisements in terms of their attractiveness. The
advertisements were equal in salary, required work experience and other details, but they were either presented as sex-biased (i.e. specifically calling out for men as ideal candidates for typically male jobs), sex-unbiased (i.e. stating that both men and women are seen as ideal candidates) or sex-reversed (i.e. specifically calling out for women as ideal candidates for typically male jobs). It was shown that women were most interested in typically male jobs when these were presented sex-reversed and least interested if there was a sex-bias towards men. In comparison, men were most interested when the advertisements were portrayed sex-unbiased.
In a second experiment, female students were asked to rate their willingness to apply for specific jobs that were presented in two sex-segregated columns (i.e. Jobs – Male Interest, Jobs – Female interest). It was found that if the jobs were segregated by sex, 46% of the women were equally likely to apply for the ‘male interest’ as the ‘female interest’ jobs. In comparison, if the jobs were presented in alphabetical order and without a reference to sex, 81% of the women preferred the ‘male interest’ jobs to the ‘female interest’ jobs. These experiments provided the first evidence that gendered wording in job advertisements affects job appeal and willingness to apply.
Nowadays, there are legislations in place in most countries to ensure that men and women are equally treated regarding their access to employment. These legislations include that under no condition there should be discrimination on the grounds of gender, for instance
in the selection criteria (Heide, 1999). Because of these legislations, explicit references to gender have become much less prevalent. However, studies have found that implicit gendered wording can also influence the extent to which women feel addressed by job advertisements.
Born and Taris (2010), for example, conducted an experiment in which they focused on the presence of implicit gendered wording within the profile description of desired applicants and its influence on application likelihood. Moreover, they manipulated the form in which the gendered wording was presented. The personal characteristics were either presented as traits, using nouns and adjectives (e.g. ‘decisiveness’), or as behaviors, using verbs (e.g. “taking decisions on how important issues can best be tackled”; Born & Taris, 2010, p.490). A main effect was found which indicated that female subjects were less likely to apply for a job when the advertisement included a typically masculine profile compared to when it included a typically feminine profile. More importantly, the results showed an interaction effect of presentation form, which indicated that application likelihood was only influenced when the profile was presented in the form of traits. When the profile was presented in form of behaviors, female subjects indicated to be just as favorable towards the masculine profile as they were to the feminine profile. This suggests that even though women feel that they have the skills to perform masculine behavior they do not attribute the corresponding traits to themselves and thus experience a lack of fit with the advertisement if masculine traits are listed as such. In contrast, men were equally likely to apply regardless of the gender-typicality of the profile and the presentation form.
Goal Orientation
In the research of Born and Taris (2010) gendered wording was operationalized as typically masculine or feminine behaviors and traits. The content analysis as carried out in Study 1 identified that this is not the only type of gendered wording which is commonly present in job advertisements, but that gendered wording can also be present as a masculine or
feminine goal orientation. Goal orientation is related to the way in which people define success and can be used to describe differences in motivations for striving towards certain achievements (Anderson & Dixon, 2009).
In the analyzed job advertisements two distinctive goal orientations were used, of which one was related to stereotypical male achievements and the other to stereotypical female achievements. In the masculine variety, goal orientation was focused on pursuing individual / agentic advancements, with an emphasis on success defined as beating the competition. In contrast, in the feminine variety, goal orientation was focused on pursuing communal / nurturing advancements, with an emphasis on working together towards successes.
In Study 2 an experiment was conducted to test how presenting either a masculine or feminine goal orientation affected job appeal. Since previous studies have shown that for women there is a negative relation between masculine wording and job appeal (e.g. Born & Taris, 2010; Gaucher et al., 2011; Horvath & Sczesny, 2015), it was hypothesized that the portrayal of a masculine goal orientation would reduce job appeal for women. At the same time, several studies found that men are not susceptive to the presence of gendered wording in job advertisements (e.g. Born & Taris, 2010; Gaucher et al., 2011) and hence it was
hypothesized that for men job appeal would not be influenced by goal orientation. H1a: Job appeal will be higher for women when the job advertisement portrays a feminine goal orientation compared to when it portrays a masculine goal orientation. H1b: Job appeal will be equally high for men regardless of whether the job
advertisement portrays a feminine or masculine goal orientation.
Belonging
In an attempt to construct a causal path as to why women find jobs advertised with masculine wording less appealing, Gaucher et al. (2011) found that this effect can partly be explained by anticipated belonging. Belonging is a fundamental social need and thus a major
motive for people to enter or abandon a certain pursuit (Stout & Dasgupta, 2011; Walton & Cohen, 2007). The presence of stereotypically masculine cues can signal to women that they do not belong within that specific social context (Cheryan, Plaut, Davies, & Steele, 2009) and the presence of masculine wording in a job advertisement can therefore demotivate women to aspire that specific job. It is thus proposed that for women the presence of a masculine goal orientation decreases the extent to which they anticipate to belonging and that this is one of the underlying mechanisms that explains the effect of gendered wording on job appeal. Again, it is hypothesized that this effect will not be found for men.
H2a: Belonging will be higher for women when the job advertisement portrays a feminine goal orientation compared to when it portrays a masculine goal orientation, resulting in higher job appeal when a feminine goal orientation is present and lower job appeal when a masculine goal orientation is present.
H2b: Belonging will be equally high for men regardless of whether the job advertisement portrays a feminine or masculine goal orientation.
Location of Gendered Wording
Gaucher et al. (2011) also indicate that it might be worthwhile to consider whether the effect of gendered wording varies depending on its location within the job advertisement. Based on the information present in job advertisements readers make inferences about organizational characteristics (Thorsteinson & Highhouse, 2003). Research has argued that corporate images, which are largely presented in the company description of job
advertisements, are very influential to the appeal of the advertised job, since these images send a strong signal of whether working for that company could enhance ones self-image and self-esteem (Feldman, Bearden, & Hardesty, 2006). The description of the desired applicant gives relatively little information about corporate images and it is only when one makes multiple inferences that it is possible to base an idea of corporate images on this description. It
thus takes more cognitive effort to make inferences about corporate images based on the applicant description than it does based on the company description. Therefore, it is expected that the presence of gendered wording has a stronger effect on belonging and job appeal when it is present in the company description compared to when it is present in the description of the desired applicant.
H3: Location moderates the mediated effect as proposed in H2a: The effect of gendered wording on belonging and job appeal for women will be stronger when it is present in the description of the company and weaker when it is present in the description of the desired applicant.
A complete figurative representation of the conceptual model as used in Study 2 can be found in Figure 1.
Figure 1. Conceptual model of the effects of gendered wording in job advertisements on job appeal via belonging, moderated by the location of gendered wording and the gender of the subject.
Study 1: Content Analysis
The objective of the first study was to find out in what way gendered wording is
currently present in job advertisements aimed at starters with a university background. 99 job advertisements were examined through an inductive qualitative content analysis guided by open questions focusing on (1) whether different types of gendered wording could be found,
(2) whether gendered wording was present at varying locations within the advertisements and lastly (3) whether besides examples of implicit gendered wording examples of explicit
gendered wording were also present.
Method
The study made use of inductive qualitative content analysis, which is a method that is used to “identify and describe patterns in a collection of texts” (Thompson, 1999, p.155). The research material was comprised of job advertisements for positions within The Netherlands. The positions had to be open for recently graduated students from university with limited working experience. Advertisements for side-jobs or internships were excluded from the analysis. In contrast to the analysis by Horvath and Sczesny (2015) which only focused on gendered wording in job titles of advertisements, this analysis was focused on the remaining content of job advertisements and did not take job titles into account.
Sampling procedure. The research material was sampled from the website
jobbird.com, which is the largest database of job advertisements in The Netherlands
(Valkenburg, 2015). Via the search function available on the website, vacancies were selected for candidates with less than a year of working experience and a university education.
Furthermore it was specified that the position had to be full-time. These specifications resulted in a selection of more than 4,000 job advertisements. With the use of random systematic sampling 1 in every 3 job advertisements was selected to be part of the sample, starting from a random position within the research material (Riffe, Lacy, & Fico, 2006). Through this method 109 job advertisements were sampled for the initial coding, of which 10 were later excluded because too much work experience was required for the advertised jobs and hence the advertisements did not meet the inclusion criteria. This resulted in a final sample of N = 99. Most of the job advertisements were in Dutch (n = 92). The remainder of the advertisements were in English.
Coding procedure. The coding procedure consisted of three different phases, namely
open coding, category creation and abstraction (Elo & Kyngäs, 2007). Before the start of the coding, the list of masculine and feminine words as used in the automated content analysis by Gaucher et al. (2011) was examined to become sensitive to the kind of words which can have gender connotations. Staying true to the qualitative nature of this analysis, however, the list was not considered as definitive or used as a reference during the coding procedure. In the open coding phase each job advertisement was carefully read and any element of the
advertisement which might contain some element of gendered wording was highlighted and summarized in a code. Then, the codes were sorted into columns so that their position reflected the section of the job advertisements that the code appeared in (e.g. description of applicant, available position, organization, atmosphere, colleagues, growth perspectives or sector).
In the second stage, categories were created for when codes seemed to be related to a similar phenomenon. During this stage it was also reevaluated whether the initial codes could indeed be seen as examples of gendered wording. The codes were retained as explicit
gendered wording whenever they contained gender-marked words (i.e. words of which the grammatical form specifically refers to either a man or a woman; Horvath & Sczesny, 2015). Other codes were retained as implicit gendered wording when they contained words with connotations to culturally embedded gender roles and stereotypes (for example, all words which communicated something about agency, communion, [Gaucher et al., 2011; Johnson et al., 2008] or status and authority [Eagly & Steffen, 1984]).
In the abstraction phase, the categories were further collapsed into main and sub categories where possible. The coding procedure was concluded with four advanced codes: two main categories were formed (‘behaviors and traits’ and ‘goal orientation’), with the same two sub categories each (‘masculine’ and ‘feminine’). Lastly, these four advanced codes were
analyzed in an Excel spreadsheet to draw conclusions about the location where these gendered wordings occurred.
Reliability and validity. To increase the reliability of the interpretations made during
the coding process, decisions which were being made about the development of categories were written down in memos (Flick, 2014). Furthermore, to assess the validity of the findings and limit researcher bias a peer debriefing took place. During this debriefing a colleague first independently read through the texts and noted emerging themes, after which these were compared to the original findings of the content analysis to see if the results were similar (Burnard, Gill, Stewart, Treasure, & Chadwick, 2008). The reviewer indicated to agree with the original results of this study.
Results
Explicit. 10 examples of explicit references to gender were found in the analyzed job
advertisements. These examples can be subdivided into three different categories. In the first category, the explicit reference to gender was used to specify the gender of the desired applicant. This occurred 3 times, once indicating a preference towards a woman (“because of the composition of our team a woman is preferred”) and a twice indicating that both sexes were equally preferred (“it goes without saying that this vacancy is open to both men and women” and “we are looking for enthusiastic ladies and gentlemen”). These examples
occurred as an introduction to the job advertisement or as a side note underneath the applicant description.
Within the second category, references were made to explicitly masculine or feminine nouns. These references were always made in the introduction of the job advertisement or in the description of the available job position and occurred 3 times. Twice only the explicitly masculine noun was mentioned (‘king’ and ‘frontman’) and once both the masculine and feminine nouns were presented next to each other (‘frontman/frontwoman’).
In the third and final category gender-marked personal pronouns (e.g. he/she) were used. These pronouns were always mentioned in pairs, no instances occurred where only the masculine or feminine form was mentioned. This category was present a total of 4 times, and always appeared in the task description. It has to be noted that these three categories are not yet saturated, since it is likely that new relevant data would emerge if a larger sample of job advertisements is analyzed (Strauss & Corbin, 1998). However, these results are sufficient to be able to conclude that explicit gender references were not very commonplace in the
analyzed job advertisements. The categories presented here mainly serve as illustrations.
Implicit. In roughly 90% of the examined job advertisements one or more examples of
words with implicit gender connotations were found. A summary of the different examples of these gendered wordings is presented in Table 1. The most important result of the analysis is the categorization of gendered wording into two types, namely a type describing behaviors and traits, and a type describing goal orientation.
Behaviors and traits. Occurrences were found of both the use of agentic (masculine) words such as independence, self-reliance, leadership and persuasiveness, as well as the use of nurturing (feminine) words such as caretaking, support, assistance and empathy. In 82.83% of the analyzed job advertisements at least one mention was made of either a masculine or feminine behavior and/or trait. The use of these wordings was not necessarily mutually exclusive, in 36.59% of the cases words with masculine and feminine connotations were both simultaneously present within an advertisement. For example, one advertisement on the one hand described the desired applicant by stating that “leadership is in your blood”, in which leadership is a masculine trait, and on the other hand stated in the task description that “you take care of keeping the organizational unit running”, in which caretaking is a feminine behavior.
Goal orientation. A separate category to behaviors and traits was identified, namely goal orientation. Whereas behaviors and traits were used in job advertisements to describe a certain attitude or disposition, goal orientation was focused on the motivation behind this disposition. In its masculine form, goal orientation was defined to be individualistic, stressing a motivation to excel, be the best and show a strong desire to succeed. A feminine goal orientation was defined to be communal, stressing the importance of working together towards achieving shared goals and contributing to common objectives. Some examples of both masculine and feminine goal orientations as were present in the examined job
advertisements can be found in Table 1. This category occurred in 40.40% of the analyzed job advertisements and only in 12.25% of these cases both a masculine and feminine goal
orientation were simultaneously present. Predominantly it was found that either the masculine or feminine form of goal orientation was used.
Table 1
Summary of Advanced Codes for Implicit Gendered Wording
Gendered Wording
Masculine Feminine
Behaviors and traits Agentic Nurturing
Leading, convincing, responsible, individual, autonomous, proactive
Taking care, supporting, giving training, involvement, empathetic Neutral: communication skills, able to work in a team
Goal orientation Individualistic Communal
Hungry to be the best, becoming a superstar, performing above average, reaching a higher level, belonging to the top
Working together to find solutions, believing that teamwork is essential for success, dedicated to contribute to a pleasant atmosphere in the office Neutral: possibilities for advancement, aspiring further growth and training
Location. The analyzed job advertisements did not have a single common structure
which they all followed or a similar make-up of paragraphs: it varied per advertisement what elements of the job, applicant and company were described and in which order. The most minimal advertisements at the very least included a short description of the job and a list of personal competences expected of the applicant. The majority of the advertisements, however,
included a longer description of the available job position and sketched a profile of the type of applicants they wanted to respond. Especially considering the further information which some of the advertisements disclosed, such as a description of the company, the colleagues, or the atmosphere in the workplace, there was a lot of variation in which elements were present in the advertisements and in what combination. In the following analyses this extra information is grouped together as communication about company characteristics.
Table 2 shows the occurrence of gendered wording per section of the advertisement. Making a comparison between the different categories of gendered wording, it can be noted that words from the behaviors and traits category were equally as often present in the description of the available job position as they were in the description of the desired
Table 2
Location of Gendered Wording in Job Advertisements
Behavior and traits
Masculine . Feminine . Total v n . % . n . % . n . % . Applicant 50 60.98 (50.51) 18 31.03 (18.18) 56 47.46 (56.57) Job position 28 34.15 (28.28) 37 63.80 (37.37) 55 46.61 (55.56) Company 4 4.88 (4.04) 3 5.17 (3.03) 7 5.93 (7.07) Total 82 (62) 100 (62.63) 58 (50) 100 (50.51) 118 (82) 100 (82.83) Goal orientation
Masculine v Feminine . Total v n . % . n . % . n . % . Applicant 10 30.30 (10.10) 6 33.33 (6.06) 16 34.04 (16.16) Job position 2 6.06 (2.02) - - 2 6.06 (2.02) Company 21 63.64 (21.21) 12 66.67 (12.12) 29 61.70 (29.29) Total 33 (28) 100 (28.28) 18 (17) 100 (17.17) 47 (40) 100 (40.40) Note. Percentages in parentheses indicate the percentage of total analyzed job advertisements (N = 99). In the total columns the number is corrected for co-occurrences of both masculine and feminine wordings in one advertisement. In the total rows the number in parentheses is corrected for co-occurrences of gendered wording in various sections of one advertisement. In these rows the percentage of total is calculated with the corrected number.
applicant (as is shown in the far right column). It was a lot less common for these wordings to appear in the description of the company. In contrast, gendered wording from the goal
orientation category was rarely present in the description of the job position, but mostly occurred in the description of the company or the desired applicant.
Examining the separate categories in more detail, an interesting difference can be noted about the occurrence of the masculine and feminine variety of behaviors and traits. Masculine behaviors and traits were more commonly mentioned in the description of the desired
applicant compared to how often they occurred in the description of the job position. In contrast, feminine behaviors and traits were more often present in the description of the job position and were less commonly present in the description of the desired applicant. Neither the masculine nor feminine form of behaviors and traits was regularly used in the
communication of company characteristics. Looking at goal orientation, it can be noted that the masculine variety occurred a little more regularly in job advertisements than the feminine variety. For the masculine and feminine variety of goal orientation the distribution of its occurrence across different sections of the job advertisements was roughly equal.
Discussion
In approximately 90% of the analyzed job advertisements at least one example of the use of gendered wording was found. These examples were rarely explicit references. For instance, explicitly masculine nouns (‘king’ and ‘frontman’) were only used in two
advertisements. Moreover, a linguistic bias via gender-exclusive language (e.g. using only the personal pronoun ‘he’; Stout & Dasgupta, 2011) was prevented because personal pronouns always appeared in pairs (‘he or she’) and the advertisements addressed the desired applicant via the second person (‘you’). Just one example was found where positive gender
Implicit gendered wording was present more regularly. For instance, in more than 80% of the advertisements typically masculine or feminine behaviors and/or traits were mentioned. The use of feminine behaviors and traits was most common in the description of the job position, whereas masculine behaviors and traits were most commonly mentioned in the description of the desired applicant. This is an indication that feminine traits occur less often in job advertisements than feminine behaviors, since the description of the job position gives more information about the required skills (and hence behaviors) whilst the applicant
description not only asks for skills but also emphasizes personality traits that applicants should possess. In contrast, words with masculine connotations often appeared in the applicant description, indicating that the occurrence of masculine traits was a lot more common.
Born and Taris (2010) reported that women are more sensitive to masculine traits than they are to masculine behaviors and hence it can be problematic that masculine traits occur more regularly in job advertisements. The presence of masculine traits can have a strong negative effect on job appeal for women, because although women do generally consider themselves to possess the skills to carry out typically masculine behaviors they do not attribute the corresponding traits to themselves (Bosak & Sczesny, 2008). The regularity in which masculine traits occur in job advertisements might thus discourage women to apply to a part of these advertisements.
Another important finding of this content analysis is a new type of gendered wording, namely goal orientation. This type of gendered wording occurred in over 40% of the analyzed job advertisements. The masculine variety was mentioned in roughly 30% of the
advertisements and the feminine variety in roughly 15% of the advertisements. Goal
orientation was rarely mentioned in the description of the job position but commonly occurred in the description of the desired applicant or of the company. In contrast to the behaviors and
traits category where it was more common for both masculine and feminine words to be present within the advertisement simultaneously, the mentioned goal orientation was almost always either masculine or feminine. As this type of gendered wording has not been
researched before, it would be interesting to investigate if the presence of a masculine or feminine goal orientation in job advertisements has a similar effect on job appeal as gendered wording of the behaviors and traits category.
Study 2: Experiment
An experiment was conducted to test how communicating a masculine or feminine goal orientation in a job advertisement influenced job appeal and belonging for both men and women. It was hypothesized that goal orientation would influence job appeal for women, meaning that women would show higher job appeal for job advertisements with a feminine goal orientation compared to advertisements with a masculine goal orientation, and that this effect would be mediated by belonging. It was not expected that a similar effect would be found for men. Moreover the experiment tested whether this effect was dependent on the location where the goal orientation was present. Since the earlier performed content analysis showed that goal orientation was most often present in the description of the company or of the desired applicant, it was decided to make a comparison between these two sections. It was hypothesized that gendered wording in the company description would have a stronger effect on belonging and job appeal for women than gendered wording in the applicant description.
Method
Research design and sample. To test the hypotheses, an experiment was conducted
with a 2 (gendered wording: masculine goal orientation vs. feminine goal orientation) x 2 (wording location: company vs. applicant description) between-subjects design. The experiment was conducted online using Qualtrics software. Participants were recruited
through friends as well as via appeals on social media and no incentive was given for
participation in the experiment. The sample was then selected on the basis of educational level (minimally a university (WO) or applied sciences (HBO) degree) and moment of graduation (less than three years ago, or within the next two years). Moreover, participants had to be at least 18 years old, but younger than 30. In total, 179 participants completed the experiment. The mean age was 23.63 (SD = 2.15) and 53.1% of the participants was female. 89.9% of the participants studied at university and the remaining 10.1% were obtaining or had obtained an applied sciences degree. 48.0% of the participants indicated to graduate in 2016 and 22.9% to have graduated in 2015. The remaining 6.7% of the participants had graduated before 2015 and 22.4% would graduate after 2016. The sample was thus sufficiently highly educated and since they were near their graduation or recently graduated they should have been able to imagine themselves as job-seekers for the purpose of this study.
Manipulation. The stimulus material consisted of four job advertisements
representing the different experimental conditions (see Appendix A-D for the full
advertisements). The text of the job advertisements was compiled using phrases based on various real examples found within the content analysis. Gendered wording was manipulated by describing either a masculine (individualistic) goal (e.g. “the organization wants to be the best and remain leading within the sector”) or a feminine (communal) goal (e.g. “the
organization wishes to work together towards the professional growth of our employees”). Wording location was manipulated by either presenting the gendered wording in the description of the company (e.g. “the organization wants to be the best and remain leading within the sector”) or in the description of the applicant (e.g. “you’re hungry to be the best and to become a superstar”). To make the gendered wording more dominant, there were two instances of gendered wording present in each advertisement. For the full manipulation, see Table 3 below. The description of the available position (a two-year management traineeship)
and the required skills (good command of spoken and written English) were held constant across all conditions.
Table 3
Design and Manipulation Overview
Location
Goal orientation (gendered wording)
Individualistic (masculine) Communal (feminine) Company description Experimental condition A Experimental condition B
“Our organization is determined to stand apart from the competition by delivering superior performance all
across the board”
“The organization wants to be the best and remain leading within the sector”
“Our organization is a warm community of motivated employees who are committed to help each other
reaching their full potential” “The organization wishes to work
together towards the professional growth of our employees” Applicant description Experimental condition C Experimental condition D
“You’re hungry to be the best and become a superstar” “You’re ambitious and highly motivated to perform above average”
“You enjoy investing in warm relationships and are an excellent
team-player”
“You’re motivated to contribute to the growth of the organization and a pleasant atmosphere in the office”
Procedure. Upon receiving an invitation to partake in the study the participants could
follow an URL linking them towards the online experiment. Here the participants were first shown a screen with information regarding the purpose and the duration of the experiment, which also ensured their anonymity. After agreeing to informed consent the participants were asked some questions about their studies (level of education, study subject, (expected) year of graduation) and their current employment status. Also, they were asked to indicate their gender and year of birth. It was chosen to ask about these demographic variables before the manipulation so that Qualtrics could separately and evenly divide the male and female participants into the different experimental conditions. The participants were then shown an instruction screen, explaining that they would have to attentively read a job advertisement whilst imagining that they were looking for a job. To ensure the participants would evaluate the advertisement whilst reading they were told that they should consider whether they wanted to learn more about the advertisement afterwards. After reading the job advertisement
they were allocated to, the participants were asked two questions about the content of the job advertisement as a check for their reading. Then they had to answer questions measuring job appeal and belonging. The survey closed with two questions to check for the manipulation and finally with a screen including a small debriefing which also provided the option to leave comments and to enter an e-mail address to be informed about the results of the study.
Pretest. A pretest was conducted with 32 test subjects to verify that the manipulation
had its intended effects. To check for the manipulation of masculine (individualistic) and feminine (communal) wording, the participants were asked on a bipolar scale to rate to what extent they thought the job advertisement was 1 = focused on individual advancements or 7 = focused on collective advancements. Participants who saw an advertisement with feminine wording rated the advertisement as slightly more focused on collective advancements (M = 3.65, SD = 2.10) than participants who saw an advertisement with masculine wording (M = 3.60, SD = 1.54), however this difference was not significant t(30) = -.07, p = .942.
To check whether the participants had noticed that the extra sentences containing gendered wording were present in different sections of the advertisement, they were asked to rate on a bipolar scale whether they thought they were presented with 1 = more information about the desired applicant or 7 = more information about the company. Participants who saw an advertisement containing more information about the company scored higher (M = 2.94, SD = 1.64) compared to participants who saw an advertisement containing more information about the applicant (M = 2.27, SD = 1.10), but although these results were in the assumed direction again the difference was not significant, t(30) = -1.35, p = .188.
Besides the pretest, the job advertisements were also evaluated by two colleagues. Based on the results of the pretest and the recommendations of the colleagues, the job advertisements were made more specific by adding information about the sector of the company. The manipulation in the advertisements was not altered because the gendered
wording was already present in 2 out of 13 sentences (15% of the advertisement) and hence quite dominant. Rather it seemed that the manipulation testing was not optimal. Because the experiment manipulates implicit gendered wording, the manipulation is mainly expected to have an effect at a subconscious level. It is thus difficult to measure this effect by asking explicit questions. Moreover, the questions which were used in this pretest were considered to be particularly vague and it was therefore decided to alter the questions so they would be easier to answer for the participants.
Variables.
Job appeal. Following Gaucher et al. (2011), job appeal was measured on a 7-point (1 = strongly disagree, 7 = strongly agree), 6-item Likert scale (“This job is appealing”; “I think I could enjoy this job”; “This is not a job I would want,” reverse coded; “This company would be a good employer”; “This job looks interesting”; and “This company seems like a great place to work”). These items all loaded on one factor that proved to be reliable (α = .85). A scale was then created by taking each participant’s average score on all six items (M = 4.56, SD = 1.02).
Belonging. Belonging was measured on a 7-point (1 = strongly disagree, 7 = strongly agree), 6-item Likert scale (“I could fit in well at this company”; “I’m similar to the other employees who work at this company”; “My values and this company’s values are similar”; “The type of people who would apply for this job are very different from me,” reverse coded; “The employees of this company would accept me as a person”; and “I feel that I would not fit in at this company,” reverse coded). The first four items of this scale are similar to the items used by Gaucher et al. (2011). To extend the scale and improve its reliability, two more items were added based on the measurement of sense of belonging as created by Hagerty and Patusky (1995). These items also all loaded on one reliable factor (α = .85). Again a scale was created by taking the average value of all six items for each participant (M = 4.21, SD = .94).
Control variables. Although the advertised job was neutral and could appeal to participants regardless of the subject of their studies, study background (e.g. law, business, medicine) was measured to be able to account for its impact on job appeal and belonging. Also employment status was added as a control measure.
Results
Test of homogeneity. The demographic characteristics of the participants in the
different experimental conditions were homogenous. Male and female participants were equally distributed over the various groups, χ2(3, N = 179) = .04, p = .998, and there was no significant difference found between the mean age of participants in these different groups, F(3, 175) = .31, p = .822. Moreover, the groups were comparable considering the study
background of the participants, χ2(18, N = 179) = 20.99, p = .280, and their employment status χ2(18, N = 179) = 16.40, p = .565. If there would be an effect of any of these demographic characteristics on the relationship between the independent and dependent variables, it would therefore be expected that this effect is equally large across the conditions.
Manipulation check. The altered manipulation check consisted of two questions to
verify that the variables of gendered wording and wording location were successfully
manipulated. The first question tested for the manipulation of gendered wording and focused on whether participants had gotten an impression that the organizational climate was
‘individualistic (with a focus on being the best)’ or ‘communal (with a focus on working together)’. The descriptions in parentheses were added because they contained words which were explicitly present in the job advertisement, making it easier for participants to identify the correct impression. The question also included a ‘don’t know’ option, which was coded as missing values. Cross-tabulations revealed that 69.9% of the participants correctly stated that the advertisement with feminine wording gave the impression of a communal climate and
52.6% correctly recognized the masculine wording to give an impression of an individualistic climate, χ2 (1, N = 157) = 6.78, p < .01. Gendered wording was thus successfully manipulated.
To test whether the manipulation of wording location was successful, the second question focused on whether the respondents had formed this impression of an individualistic or communal climate based on the description of the applicant or of the company. Again, a ‘don’t know’ option was present, which was coded as missing values. 57.7% of the
respondents correctly formed this impression based on the description of the company and 58.1% correctly formed this impression based on the description of the applicant, χ2 (1, N = 145) = 6.78, p = .056. Since this difference is only marginally significant, it is not entirely certain whether the manipulation of wording location in the experiment was successful. If any significant effects are found during the hypothesis testing, these need to be interpreted with caution since they might not be attributable to this manipulation.
Hypotheses testing. To test whether belonging mediates the relationship between
gendered wording (i.e. masculine or feminine goal orientation) and job appeal, whilst considering the conditional effects of the location of gendered wording and the participants’ gender, a moderated mediation analysis was conducted using the PROCESS macro with 5000 bootstrap samples as developed by Hayes (2013). Model 10 was used as it corresponds to the conceptual model of this study (see Figure 1), with gendered wording as the independent variable, job appeal as the dependent variable and belonging as the mediator. Gender and location were added as potential moderators for the relationship between gendered wording and belonging, and gendered wording and job appeal.
The results of the moderated mediation analysis can be found in Table 4. No significant main effect of gendered wording on job appeal was found (p = .134). When male and female participants were taken into account simultaneously the presence of either a masculine or feminine goal orientation did not affect job appeal. The interaction effect of participants’
gender and gendered wording was also not significant (p = .456), meaning that the effect of gendered wording on job appeal did not vary between men and women. These initial results do not find support for Hypothesis 1a, which posed that job appeal would be higher for women if the job advertisement portrayed a feminine goal orientation compared to a masculine goal orientation. The results do support Hypothesis 1b, which posed that no significant effect would be found for men.
Table 4
Regression Coefficients of the Moderated Mediation Analysis
Outcome
Belonging Job appeal
Predictor Coefficient Coefficient
Intercept 4.46 (.17)** 1.62 (.33)** Belonging - .64 (.07)** Gendered wording -.14 (.25) .33 (.22) Gender -.10 (.20) .08 (.17) Gender x G.W. .30 (.28) .17 (.24) Location -.11 (.20) .17 (.17) Location x G.W. -.17 (.28) -.57 (.24)* Overall F .80 18.77 Overall R2 .02 .40**
Note. Standard errors in parentheses. *p < .05. **p <.001.
In H2a it was hypothesized that the effect of gendered wording on job appeal for women would be mediated by belonging. As can be seen in Table 4, the results provide no evidence for a mediation effect. Although belonging did significantly predict job appeal, b = .64, t(172) = 9.78, p < .001, gendered wording was not shown to have an effect on belonging (p = .571). Moreover, seeing as the interaction term of gender and gendered wording is again insignificant (p = .281), there is no evidence that the effect of gendered wording on belonging varies between male and female participants. Hypothesis 2a is thus rejected seeing as
belonging is not shown to mediate an effect of gendered wording on job appeal for women. In contrast, Hypothesis 2b posed that no similar mediation effect would be found for men and can thus be accepted.
Finally, it was hypothesized in H3 that wording location would moderate the mediated effect of gendered wording on job appeal via belonging for women. To test for this
hypothesis, first an interaction between gendered wording and wording location on job appeal and belonging has to be established. As a next step, it can then be examined whether this interaction varies between male and female participants. A significant and negative interaction was found of gendered wording and location on job appeal, b = .58, t(172) = -2.37, p < .05, but no significant interaction was found of gendered wording and location on belonging (p = .541). From this result it can already be concluded that Hypothesis 3 is not fully supported by the data, because even when taking the moderating effect of location into account there is again no evidence that belonging mediates the effect of gendered wording on job appeal. It remains relevant, however, to further examine the significant interaction effect of gendered wording and location on job appeal to possibly find partial support for this hypothesis.
The significant interaction effect of gendered wording and location on job appeal has a negative coefficient. Because of the way in which location was dummy coded (0 = gendered wording in the description of the company, 1 = gendered wording in the description of the applicant), this means that the effect of feminine versus masculine gendered wording on job appeal is more negative when it is present in the description of the applicant (where job appeal decreases by .25 when a feminine rather than masculine goal orientation is present) compared to when it is present in the company description (where job appeal increases by .44 when a feminine rather than masculine goal orientation is present). This is graphically
represented in Figure 2a, where the two slopes are significantly different from each other. To examine the effects underlying the significant interaction of gendered wording and location on job appeal in more detail, one can look at the statistics for the conditional effects
across the various experimental conditions which the PROCESS macro also provides (see Table 5). When examining the direct effects, a significant and positive effect is found for women in the experimental condition with gendered wording present in the company
description, b = .51, t(172) = 2.44, p < .05. As can also be seen in Figure 2b, in this condition women show significantly more job appeal when the company description presents a feminine Figures 2a and 2b. Mean scores on job appeal for variations in gendered wording, dependent on location. *p <.05.
Table 5
Conditional Effects of Gendered Wording on Job Appeal via Belonging at Different Levels of Gender and Location
Direct Indirect Coefficient (SE) 95% CI Coefficient (SE) 95% CI Males Company .33 (.22) -.10 to .76 -.09 (.16) -.43 to .21 Applicant -.25 (.21) -.66 to .17 -.20 (.17) -.56 to .11 Females Company .51 (.21)* .10 to .92 .10 (.15) -.18 to .40 Applicant -.07 (.20) -.47 to .33 -.01 (.16) -.34 to .31 Note. *p < .05. 3.8 4.3 4.8
Masculine wording Feminine wording
Male and female participants
Company description Applicant description
3.8 4.3 4.8
Masculine wording Feminine wording
Female participants only
Company description* Applicant description
goal orientation (M = 5.08, SD = .95) compared to when it presents a masculine goal orientation (M = 4.45, SD = .84). In the other experimental conditions no significant effects are found of gendered wording on job appeal. These results thus find partial support for Hypothesis 3: for women it holds true that the effect of gendered wording on job appeal is stronger when it is present in the description of the company and weaker, or in this case non-significant, when it is present in the description of the desired applicant. Also, this finding provides conditional support for Hypothesis 1a: gendered wording does affect job appeal for women but only when present in the description of the company.
Follow-up tests indicated that, surprisingly, presenting a masculine goal orientation in the company description did not seem to have a negative effect on job appeal for women when compared to men. No significant difference was found between the mean scores on job appeal for male and female participants who read this advertisement, F(1, 47) = .09, p = .767. Women who read this advertisement indicated approximately as much job appeal (see above) as men who read the same advertisement (Mmen = 4.52, SDmen = .87). Surprisingly, men also evaluated the advertisement with feminine wording in the company description more
positively (M = 4.73, SD = 1.01) than the advertisement with masculine wording, but only for women the positive effect of feminine rather than masculine gendered wording on job appeal was significant. This suggests that for women the presence of a masculine goal orientation does not necessarily decrease job appeal, but rather that presenting a feminine goal orientation can stimulate job appeal.
Discussion
These results present some evidence that gendered wording in job advertisements indeed influences job appeal for women but not for men. Under certain conditions, women indicate higher job appeal when a feminine goal orientation is present compared to when a masculine goal orientation is present.This effect is conditional since it is moderated by
location. The positive effect of feminine wording on job appeal for women was only found when the gendered wording was present in the description of the company and not when it was present in the description of the applicant.
Contrary to what was hypothesized, the effect of gendered wording on job appeal was not shown to be mediated by belonging. This study thus does not provide evidence that masculine gendered wording signals to women that they do not belong in the job, which is in disagreement with the findings of Gaucher et al., (2011). Since this experiment tested for a different type of gendered wording than the experiment of Gaucher et al., (2011), it could be that the causal reasons explaining the effect of gendered wording on job appeal differ
depending on its type. Alternatively, it could be that this mediation effect was not found in this experiment because the job advertisements contained relatively little information about the company. Seeing as the participants read this job advertisement as part of an experiment and not out of necessity to find a job, they were likely not very motivated to make inferences about the organizational climate based on the little information they were presented with. It is therefore possible that instead they indicated more neutral opinions towards belonging.
General Discussion
This research has found that gendered wording is commonly present in job
advertisements aimed at starters with a university background, predominantly in its implicit form. Explicit gendered wording occurred more rarely and only one example of positive gender discrimination towards women was detected. Implicit gendered wording could be categorized into two different types: gendered wording was present either as masculine or feminine behaviors and traits, or as a masculine or feminine goal orientation. Considering the location of gendered wording, it was found that masculine behaviors and traits were mostly mentioned in the description of the desired applicant, whereas feminine behaviors and traits
were mostly mentioned in the description of the available job position. Behaviors and traits were rarely mentioned in the description of the company. Goal orientation, in contrast, was most commonly present in the description of the company and also occurred regularly in the description of the desired applicant.
In an experiment it was found that women showed more job appeal when a feminine goal orientation was mentioned in a job advertisement compared to when a masculine goal orientation was mentioned, but only when the gendered wording was present in the company description. This indicates that the effect of gendered wording on job appeal is indeed
dependent on the location of the advertisement where it is present, as was suggested by Gaucher et al. (2011). No evidence was found that a masculine or feminine goal orientation affected job appeal when it was present in the description of the desired applicant. In agreement with previous studies on gendered wording in job advertisements (e.g. Born & Taris, 2010; Gaucher et al., 2011), it was found that the presence of gendered wording did not influence job appeal for men.
A possible explanation for why there was no significant effect of gendered wording on job appeal for women when goal orientation was mentioned in the applicant description is that younger women have had less experience with gender biases towards men. As stated by organizational psychologist Belle Derks, female students often do not yet recognize that men get different opportunities than women to climb the work ladder, but only realize this later in their working lives (De Bruin, 2016). When a job advertisement portrays a masculine goal orientation in the applicant description it signals an organizational bias towards employees striving for stereotypically male successes. Because stereotypes not only describe what men and women are like but also prescribe how men and women are expected to behave,
characteristics which are highly valued and celebrated for men are often regarded as unattractive when portrayed by women (Heilman, 2001). In organizations which value
employees with a masculine goal orientation, men are thus especially likely to advance more quickly since they are generally applauded for being achievement-oriented. In comparison, women are more likely to be penalized for having a masculine goal orientation, because that means they deviate from the expectation of being more service-oriented (Heilman, 2001). Since female students have not yet had many experiences with biases of this kind, they might not be as sensitive to gendered wording that portrays such a bias. In comparison, women with some work experience are likely to be aware of gender inequalities in the work place and hence could be more attentive to biases portrayed in the description of desired applicants.
The check for successful manipulation of wording location was only marginally significant and it is thus not entirely certain that participants correctly recognized that the gendered wording described either the goal orientation of the company or of the desired applicant. It is thus necessary to consider that there might also be different reasons why the effect of gendered wording on job appeal varied depending on where it was presented. An alternative explanation could be that the company description was presented in the first paragraph of the job advertisement and the applicant description further on in the third paragraph. Primacy effects explain that, when making judgements, information that is presented early on can influence all the conclusions that people subsequently draw (Kruglanski & Freund, 1983). When gendered wording is present in the beginning of the advertisement it could influence the judgement of all the text that follows and make a more lasting impression compared to when it is present later on in the advertisement. It is thus possible that the difference in effect is merely based on the order in which the two descriptions appeared and not on the different inferences which were made depending on these descriptions.
In a previous study, Gaucher et al. (2011) found that the effect of gendered wording on job appeal was mediated by belonging. Contrary to the expectations, this research did not find
similar evidence of this mediation effect. Walton and Cohen (2007) state that people especially try to evaluate whether or not they would belong in a certain pursuit when they perceive themselves to be a minority. A possible explanation why the mediation effect was not found in this study could be that because women are no longer a minority in universities they are less afraid to be underrepresented, and are therefore less attentive to cues that might signal whether or not they belong (Cheryan et al., 2009). Alternatively, as was stated before, it could be that no significant effects were found because participants were not motivated to make many inferences in order to be able to form an impression of the organizational climate and evaluate belonging, but instead indicated neutral opinions towards belonging.
Limitations of the Research
It has to be recognized that this study is not without limitations. First of all, the job advertisements which were analyzed in Study 1 were sampled from the website jobbird.com, meaning that strictly speaking the results of the content analysis are only generalizable to the advertisements placed on that website. That being said, Jobbird.com can be seen as a
representable website for other online databases of job advertisements. It is likely, however, that job advertisements presented in different media, such as newspapers, company websites or social media, differ from those presented in these online databases.
A second limitation is that the job advertisement which the participants had to read in the experiment was purposively kept very general so that it could appeal to all participants regardless of their study background. There was for instance relatively little information present about the actual tasks that the job entailed. If the participants had been instructed to imagine themselves to be qualified for the position, rather than to evaluate it against their own study backgrounds, the advertisement could have been made more specific.
A second way in which the job advertisement was kept more general was by asking for relatively few personal competencies in order to limit the amount of text surrounding the
gendered wording. Because of this generality it could be that even though the advertisement was based on and looked comparable to actual job advertisements, participants might not have felt that the job advertisement was very realistic. A final disadvantage of the generality of the advertisement was that it elicits less strong opinions and draws participants more towards middle-response options.
Directions for Future Research
Aside from its limitations this research does provide suggestions for future studies. It was found that, under certain conditions, gendered wording had an effect on job appeal for women but that this effect was not explained by a difference in anticipated belonging. It remains relevant to test for alternative mediators which might explain why women value job advertisements portraying a feminine goal orientation in the company description more than those portraying a masculine goal orientation. Since research has argued that corporate images present in the company description send strong signals about whether working for that
company could enhance self-image (Feldman et al., 2006), it would be worthwhile to test for a mediating effect of self-image enhancement (e.g. ‘working at this company would make me feel good about myself’).
Furthermore, it was posed that primacy effects could be an alternative explanation why an effect of gendered wording on job appeal was only found when it was mentioned in the description of the company, seeing as this description was presented in the first paragraph of the job advertisement. It would thus be interesting to conduct an experiment with the same example of gendered wording present in either the first, middle or last section of the job advertisement and compare to what extent job appeal is then similarly affected by those statements.
Lastly, it was suggested that women with working experience might be more sensitive to gendered wording than female students with limited working experience. Hence it would
