SALESPERSON AUTHENTICITY AND GENDER-BASED PERCEPTIONS: AN INTERNATIONAL STUDY

Mirroring the increasing percentage of women in the professional workforce, the number of females in sales has dramatically risen in recent years. The growth of gender diversity within the world of B-to-B buying and selling requires an examination of potential gender effects associated with interpersonal communications. The latter influences perceptions of authenticity, which in turn, influence sales effectiveness. The current study uses an international sample to review interpersonal communication drivers that are associated with buyer perceptions of salesperson authenticity. Gender differences associated with the processing of body language and the spoken word are particularly examined. Results indicate that, in high-involvement sales situations, female perceptions of salesperson authenticity are highly influenced by salesperson body language. Male perceptions, conversely, are more heavily influenced by the spoken word. Implications for sales training are provided.


INTRODUCTION
In the United States, women now outnumber men in management, professional, and related occupations (Bureau of Labour Statistics, 2019). The trend is mirrored across the developed world (International Labour Organization, 2018). In sales, the number of women has grown exponentially, and women now play key roles in both buying and selling. However, potential gender differences associated with interpersonal communications and personality factors in sales have not been extensively studied (McQuiston & Morris, 2009). The fact that interpersonal communications and personality factors in sales remains understudied (Erevelles & Fukawa, 2013) is concerning since the personal interactions between salespeople and buyers affect sales effectiveness (Bush et al., 2002;Sheth & Sisodia, 2002).
In sales situations, personal interaction factors influence consumers' perceptions of salesperson authenticity, which in turn, may influence consumers' sales receptivity . Authenticity has been defined as "the unobstructed operation of one's true, or core, self in one's daily enterprise" (Kernis, 2003, p.13). Barrett-Lennard (1988) defined authenticity as "consistency between the three levels of (a) a person's primary experience, (b) their symbolized awareness, and (c) their outward behaviour and communication" (p. 82).
The current study focuses on perceptions of authenticity in high-involvement sales situations. Are male and female perceptions of salesperson authenticity driven by similar personal interaction factors? The study's research objective is to answer the question via examining the factors that drive each gender's perceptions of authenticity.
The current paper begins by reviewing personal authenticity and gender in relation to interpersonal communication factors. Next, a quasi-experimental design is presented. The design is influenced by Plouffe et al.'s (2008) call for experimental designs in sales research. The design involves an automotive sales experiment tested in a four-country convenience sample (Czech Republic, Germany, Switzerland, and the United States). Structural equation modelling (SEM) is used to analyse collected data. The paper concludes with a discussion which may be useful to sales managers who must train salespeople to operate effectively in today's climate of gender diversity.

LITERATURE REVIEW
The current section begins by discussing personal authenticity and the interpersonal communication factors that drive perceptions of authenticity. The section then reviews gender differences associated with interpersonal communications.

Authenticity
According to Freud (1938), personal authenticity results when an individual's emotional (affective), rational (cognitive), and physical (behavioural) characteristics are balanced. The physical component involves body language or other external, behavioural displays. The rational component involves displays of the mind and thinking, and the emotional component includes displays of feelings.
In sales, Stros, Heinze, and Říha (2017) determined that consumer perceptions of salesperson authenticity are primarily informed by body language (behaviour) and the spoken word (cognition). The emotional, or affective, factor does not significantly impact consumer perceptions of authenticity. The authenticity factor model (AFM) showcased that if a salesperson balances body language and the spoken word in a congruent manner, consumer perceptions of the salesperson's authenticity will be positively influenced and desirable sales outcomes will result . In contrast, if the salesperson displays a lack of balance, negative consumer judgements www.ieeca.org/journal and undesirable sales outcomes will result (see also Bauer, 2002).
In light of the general findings on authenticity, the current study explores specific interpersonal drivers of cognition and behavior and their resultant effect on perceptions of salesperson authenticity. The spoken word (cognition) is studied via the content (CO), organization (OR), style (ST), and tone (TO) of a salesperson's speech (see . Body language (behaviour) is studied via immediacy cues (IM), relaxation (RE), movement (MO), and facial patterns (FA) (see Mehrabian, 1969).

Gender
Across recent decades, researchers have debated the extent and degree to which gender variance exists. The current section reviews pertinent gender difference literature. First, macro notions of gender divergence are examined and are followed by specific reviews of gender differences related to cognition and behaviour.
During the 1960s and 70s, feminist research attempted to downplay gender differences (Grant, 1988). Empirical research in the 1980s and 90s discredited the effort, but the downplaying of gender differences resurfaced in the form of gender similarity hypothesis (Hyde, 2005;Zell et al., 2015). Relying on meta-analyses of gender difference literature, proponents of gender similarity hypothesis have argued that gender dissimilarities are nominal at best. Rather, the two genders are more similar than dissimilar.
Critics of gender similarity hypothesis respond that although the genders may be more similar than dissimilar across a broad array of variables, the point is largely meaningless. For example, the DNA of humans and bonobos is 99% similar, but this similarity does not indicate that humans and bonobos are the same or that their interactions are not dramatically influenced by the 1% (Zuriff, 2015). Critics contend that the core weakness of gender similarity hypothesis is that it relies on meta-analyses which give equal weighting to interpersonally meaningful and non-meaningful variables. For example, although the genders vary significantly in relation to interpersonally meaningful items such as assertiveness (Twenge, 1997), there is minimal gender variation in relationally non-meaningful items such as time duration judgement (Block, Hancock, & Zakay, 2000). However, gender similarity hypothesis includes both assertiveness and time duration judgement as co-equal variables. Through including a large number of relationally insignificant variables, the hypothesis dilutes many significant variance factors (e.g. assertiveness, independence, understanding of others, gentleness, etc.). The net result is a blunt instrument that does not enable precise analysis of gender-based interpersonal communication factors (Zuriff, 2015).
The use of surgical instruments in genderbased interpersonal communication analysis has resulted in the suggestion that the genders vary significantly across an array of relationally meaningful variables. These differences, across multiple decades, have been extensively documented (Hall, 2006). Two competing theories have attempted to explain the variation. Gender socialization theory (or social role model theory) suggests that gender variance is fundamentally a function of divergent socialization methods (Eagly, 1987). For example, Henley (1977) attempted to explain differences in nonverbal communications through highlighting the manner in which males and females are socialized differently in relation to power and dominance. Trait theorists, conversely, view gender variance as primarily stemming from innate, biological differences (Buss, 1995;Michalski & Shackelford, 2010;Pool, 1994); a conclusion drawn from research which showcases significant gender differences in early childhood before social acculturation has wielded a measurable effect (Else-Quest et al., 2006;McClure, 2000). Combining perspective from social role model and trait theorists, Abele's (2000) dual adaptation model suggests that both social role models and innate biological factors inform gender differences.
With respect to body language, the genders vary in the physical manner by which they express themselves and in the interpretations that they assign to body language (Hall, 1978(Hall, & 2006Koppensteiner & Grammer, 2011;McClure, 2000;Vrugt & Luyerink, 2000). Regarding the latter, some early research uncovered relatively insignificant gender differences associated with the interpretation of non-verbal cues (Coleman 1949, Davitz, 1964 www.ieeca.org/journal Maccoby & Jacklin, 1974;Wesgerber, 1956). However, methodological issues plagued many of these studies (Hall, 1978) and the majority of research indicates that females are substantially more inclined to assess non-verbal communications and are more advanced in their ability to interpret non-verbal communications (Briton & Hall, 1995;Hall, 1978Hall, & 2006Guadagno & Cialdini, 2007;Kotlyar & Ariely, 2013;LaFrance & Henley, 1994;McClure, 2000;Rosen et al., 2008). Hall (1984) speculates that females' historically submissive role has increased their needs and motives to "understand subtle interpersonal cues" (p.14). Therefore, we hypothesize that: H1: Salesperson body language influences female consumers' perceptions of salesperson authenticity more than it influences male consumers' perceptions of authenticity.
In relation to the spoken word, males and females process marketing messages differently in low involvement situations characterized by message cue congruity (Meyers-Levy & Maheswaran, 1991;Meyers-Levy & Sternthal, 1991). Women tend to utilize a detailed information processing approach which focuses on central cues, and men tend to process appeals via peripheral cues and the interpretation of those cues via pre-existing schemas (Meyers-Levy & Zhu, 2010;Papyrina, 2015;Putrevu, Tan, & Lord, 2004). However, in high-involvement situations, men outperform females in the use of central cues to identify and process subject matter content (Kiecker, Palan, & Areni, 2000). Additionally, in high involvement situations, men are more likely to resort to rational analysis of central (spoken or written word) cues (Guadagno & Cialdini, 2007). Since the current study examines a high-involvement product purchase (automobile), we hypothesize that: H2: A salesperson's spoken words influence male consumers' perceptions of salesperson authenticity more than they influence female consumers' perceptions of authenticity.

METHODOLOGY
The current study utilized a quasiexperimental design which included video observations and subsequent questionnaires.
Four two-minute sales videos were produced . Professional actors, playing the roles of male and female automotive salespeople, were used in the videos. Each video presented a sales situation in which the salesperson displayed positive or negative affect (emotional expressions), behaviour (body language), and cognition (spoken word). The videos utilized "over-the-shoulder" camera shots so that viewers could only observe the torso and head of the salesperson. Each video included multiple elements of a typical personal sales interaction (welcome, needs assessment, product presentation, pre-close, etc.).
The four videos were shown to a sample of 1,421 university students who were potential automotive consumers. Sampled students were drawn from public universities in California, USA (443 participants); Prague, Czech Republic (478 participants); Ravensburg, Germany (703 participants); and Zurich, Switzerland (154 participants). After watching each video, student participants were asked to complete a 26-item survey. Survey responses were recorded both electronically and on hardcopy questionnaires. Approximately 65% of respondents were male. For a detailed discussion of the experiment and survey design/content, please see .

DATA ANALYSIS
1 Collected data was assessed for outliers, missing values, skewness and kurtosis. All tested values were within the acceptable range, according to Bortz and Doering (2006). Sample size was sufficiently large (Hair et al., 1995;Tabachnick & Fidell, 2007), and the data was normally distributed (Fabrigar et al., 1999). Missing values (less than 2% of responses) were replaced by mean values of the variable in question (Hair et al., 2009;Howell, 2007;Little & Rubin, 2002). The data was then factor analysed. Factor analysis is a multivariate statistical method which can be used in applied research for data reduction, instrument development, and trait identification.
2 Structural equation modelling (SEM), also known as covariance structural analysis or causal modelling, was used to test the hypothesized relationships in the proposed www.ieeca.org/journal 141 model (see Figure 1). SEM enables the simultaneous estimation of multiple regression equations in a single framework and allows for the investigation of relationships between evident (observed) and hidden variables. The current study's model was fitted using lavaan version 0.6-4 (Rosseel, 2012) in R version 3.6.1 (R Core Team, 2019). The model was estimated using the maximum likelihood method. The latent factors were standardized, allowing for a free estimation of loaded factors.
3 Model fit statistics indicated a good model fit (see Table 1).  Standards for an "acceptable" level of model fit may differ across disciplines (Hu & Bentler, 1999). The values of the current study's models are within commonly accepted ranges.
www.ieeca.org/journal  Table 2 displays the results associated with the "male" data set. The first column (Estimate) contains the parameter value (estimated or fixed) for each model parameter. The second column (Std.err.) contains the standard error for each estimated parameter, and the third column (z-value) contains the Wald statistic (obtained by dividing the parameter value by its standard error). The fourth column (p(>|z|)) contains the p-value for testing the null www.ieeca.org/journal hypothesis that the parameter equals zero in the population. The last column (Std.all.) presents the completely standardized relevant coefficient. Table 3 displays the results associated with the "female" data set. Factor loadings and regressions are shown in Table 3. The model assumes authentic perception as a central criterion for the sales decision (see also . Results indicate two relevant paths for both data sets (i.e. "male" and "female"). Despite the fact that a difference in the weighting of the factors relevant to a purchase decision was identified for both genders, factor relationships were generally similar. The data also revealed that body language (BL) has a negative influence on perceptions of salesperson authenticity (AUT). A positive significant relation was uncovered between perceptions of authenticity (AUT) and the spoken word (SW), which means that the salesperson's spoken words positively influence perceptions of authenticity (AUT).
Additionally, the relevant factors of the latent variables (AUT, BL and SW) were identified.
To identify potential gender differences, the standardized coefficients (Std.all.) for "male" and "female" data were compared by determining the actual differences as a percentage of the values (see Table 4).

RESULTS AND DISCUSSION
Personal interaction factors affect perceptions of authenticity, which, in turn, affects sales effectiveness. The current study highlighted that male and female perceptions of salesperson authenticity are driven by variant personal interaction factors. Male consumers' perceptions of authenticity are, in high involvement product situations, more heavily influenced by the spoken word. Female consumers' perceptions of authenticity are, in high involvement product situations, more heavily influenced by body language.
Specific results associated with the study's hypotheses are as follows: H1: For male customers (as compared to female customers), the body language (BL) of the salesperson has a relatively lower influence (-27.7%) on perceptions of salesperson authenticity (AUT). Therefore, Hypothesis H1 is supported. Body language (BL) exerts greater influence on female perceptions of salesperson authenticity (see Table 4).
Additional body language results include the finding that male perceptions of body language are more heavily influenced by body movement (15.5%) and facial expressions. For female customers, facial www.ieeca.org/journal 145 expression (FA) did not significantly impact perceptions of body language.
H2: For male customers (as compared to female customers), the spoken word (SW) has a relatively higher influence (14.4%) on perceptions of salesperson authenticity (see Table 4). This finding supports Hypothesis H2. The spoken word has a smaller influence on female perceptions of salesperson authenticity than it does on male perceptions of authenticity.
Regarding body language, the current study found that not only do the genders differ in relation to the relative influence of body language on perceptions of salesperson authenticity, but they also differ in relation to the interpretation of body language. Male interpretations of body language are more heavily influenced by large-scale body movement and facial expressions. Females, on the other hand, are more holistic in their interpretation of body language. The current study's finding accords with past research regarding gender-based interpretations of body language. Females, on average, are much likelier to perceive and understand physical nuance or micro-signals, while males are usually less adept at interpreting these displays (Briton & Hall, 1995;Hall, 1984;Guadagno & Cialdini, 2007;Kotlyar & Ariely, 2013;LaFrance & Henley, 1994;McClure, 2000;Rosen et al., 2008).
There are several implications associated with the current study's body language findings. First, sales managers and salespeople should be taught that body language substantively impacts consumer perceptions of authenticity. Salesperson training should include crossgender feedback in which male and female trainers or co-workers provide interpretive body-language feedback. Filmed role-plays are a particularly useful format for enabling circumspect body-language assessment and feedback (Hughes, Ginnett, & Curphy, 1993;Parker, Pettijohn, & Luke, 1996). Feedback from both male and female reviewers should be provided so that salespeople develop an understanding of the manner in which the same gestures or body-language expressions may elicit variant responses.
A second implication of the current study's body-language finding is that males and females should not assume that they understand the body language of the other gender correctly. The fact that body language varies in its effect on gender-based perceptions of authenticity implies that the genders do not interpret body language in the same manner. Further research is required to better understand the specific nature of this difference, but the fact that a difference exists should be consciously recognized and considered by sales managers and salespeople.
Regarding the spoken word, the current study confirmed prior findings that males are generally more reliant on the spoken word during high involvement product purchase situations (Guadagno & Cialdini, 2007). In the current study's automotive purchase situation, male consumers relied more heavily on central cues (the spoken word) to inform their perceptions of salesperson authenticity. The implication for female salespeople in high-involvement product situations is that they must refrain from dismissing central cues (the spoken word) in favour of peripheral cues (body language). Conversely, male salespeople must be careful to avoid an overreliance on the spoken word when dealing with female customers.
Although the current study's methodology resulted in acceptable reliability and validity, limitations exist. First, the utilization of a convenience sample resulted in a non-normal age and gender distribution of sampled participants. Second, the study's location may limit its generalizability. The experiment was conducted using an automotive dealership in the United States. International variations in the automotive-buying process may impact whether the study's findings can be generalized to other business markets and cultural environments (see also Stros, Říha & Möslein-Tröppner, 2018;. Finally, only a small minority of the sampled university participants were current automobile customers. However, considering that most of the sampled participants were potential automotive customers, the authors do not view this limitation as significant.