Employee personality can be analyzed in terms of levels of consistency in modes of cognition, affect and behavior [16]. Among personality trait models introduced in the literature, the Big Five model developed by Costa and McCrae [17] has emerged as a popular tool for understanding the relationship between personality and various individual behaviors [18]. It has been employed by many researchers (e.g., [19-22]). The so-called “five factors” are conscientiousness, agreeableness, extraversion, emotional stability and openness to experience. The conscientiousness character of a person indicates the capacity to be responsible, achievement-striving, dependable, efficient, organized, hardworking and persevering. Agreeableness refers the tendency to be sympathetic, helpful, friendly, tolerant, trusting, good-natured, courteous and cooperative. An extraverted person is understood to be ambitious, reward-seeking, gregarious, sociable, adventurous and assertive. Emotional stability refers to the ability to remain calm, resilient, even-tempered, tolerant of stress, well-adjusted and self-confident. Finally, openness to experience indicates the capacity of a person to be broad-minded, imaginative, perceptive, intelligent, creative, curious and cultured.

Adaptive selling is defined as “the altering of sales behavior during a customer interaction or across customer interactions based on perceived information about the nature of the selling situation.” Weitz et al. [43]. In other words, to practice adaptive selling a salesperson must change or adapt his selling strategies for each customer. The Weitz et al. (1986) adaptive framework posits that an adaptive selling strategy is critical to the employee’s performance in face-to-face interactions. The concept of “adaptation” is also used by researchers in the context of service. Hartline and Ferrell [23] define it as the ability of employees in contact with customers to adjust their behavior according to the application context of clients.

In a call center environment, each employee is located within an individual cubicle and receives several heterogeneous customer calls on a given work day. In contrast to the employee who works in a B-to-B market, employee in call centers interact with customers via phone technology, with calls received through a computer system. In a call center, communication is limited to verbal interaction, thus it is critical for employee to have selling strategies to suit the needs of each customer [15]. Qualitative research conducted by Eveleth and Morris [15] in call centers shows that many employees claim that adaptive behavior significantly affects their work effectiveness.

Define Sales performance can be defined as “behavior evaluated in terms of its contribution to the goals of the organization” (Johnston and Marshall [24], p. 412). In the literature, many criteria or types of sales performance have been used (e.g., supervisory rating performance, profit, sales and performance with customers). In our research, we have used performance with customers as the criteria of performance for four reasons. First, Hunter and Perreault [25,26] highlight that performance with customers is a good criterion of performance for a technological organization. They define performance with customers “as developing an understanding of a customer’s unique problems and concerns-marketing, technology, operations, or otherwise-and recommending solutions that address those concerns” Hunter and Perreault ([25], p. 99). This definition fits with a call center environment where employees must understand each unique customer’s need or objection and be able to provide a personalized solution. In addition, many call centers in Canada deal mainly with phone and Internet companies. Hunter and Perreault’s [25,26] definition of performance with customers and the context in which they recommend it is appropriate for call centers in Canada. Second, some researchers (e.g., [3,4,6]) have highlighted the idea that the relationship between some Five-Factor traits (e.g., agreeableness) and performance depends on performance criteria. Therefore, we think Sawyerr et al.’s [22] call center study result, which found that only one trait of the Five-Factor model is associated with performance, is due mainly to the type of performance studied. Sawyerr et al. [22] have used service performance ratings composed of three items, accuracy of information provided to customers, speed of response to customer requests and the ability to solve problems. Finally, Hunter and Perreault [25] have shown that some Five-Factor traits have been significantly associated with employee performance with customers.

There has been very little research that has directly examined the relationship between Five-Factor personality traits and performance in call centers. Sawyerr et al.’s (2009) call center research has shown, in contrast to the author’s prediction, that only openness to experience was significantly correlated with employee performance. We think this result is attributable mainly to the performance criteria used. Sawyerr et al. [22] determined service employee performance to be measurable by three factors: 1) the accuracy of information provided to customers; 2) the speed of response to customer requests; and 3) the ability to solve problems. However, these items do not capture all aspects of employee performance with customers in call centers. For example, their performance criteria do not include the ability of employees to generate sales. This example shows how many authors [6,27] have highlighted the importance of choosing relevant performance criteria when investigating the relationship between personality traits and performance. In our research, we have chosen performance with customers as the criterion of employee performance in call centers, and we have justified the relevance of this choice below.

Many studies and meta-analyses have shown that conscientiousness and emotional stability are the most robust predictors of performance across all jobs and settings [3- 6,11,28], specifically in jobs that require interaction [28]. Therefore, we predict that conscientiousness and emotional stability are correlated with performance with customers. In addition, based on Sawyerr et al.’s [28] results, we predict that openness to experience also has a positive effect on performance with customers.

Prior research (e.g., [3-6,11,28]) has highlighted the idea that the other five-factor personality traits (agreeableness and extraversion) are correlated with some performance criteria and some jobs [4]. In addition, Mount et al.’s [28] meta-analysis reported that Agreeableness and Extraversion are associated with performance in jobs that require interpersonal interaction. Bakker et al.’s [29] research also showed that Agreeableness is an important predictor of performance in cases of negative encounters between employees and customers. The cases described above cover two aspects of employee performance in call centers. Therefore, we expect that Agreeableness and Extraversion predict employee performance with customers in call centers. In sum, we expect the following:

H₁₁. Conscientiousness will be positively associated with employee performance with customers in call centers.

H₁₂. Agreeableness will be positively associated with employee performance with customers in call centers.

H₁₃. Extroversion will be positively associated with employee performance with customers in call centers.

H₁₄. Emotional stability will be positively associated with employee performance with customers in call centers.

H₁₅. Openness to experience will be positively associated with employee performance with customers in call centers.

Headings, Very few researchers have investigated the relationship between personality traits and adaptive behavior. Spiro and Weitz [13] showed that five personality traits (self-monitoring, empathy, androgyny, being an opener, and locus of control) are positively associated with employee adaptiveness. Based on that study, Verbeke [30] investigated the impact of four personality traits (self-monitoring, interpersonal control, openness and rigidity). The results revealed that three personality traits (self-monitoring, interpersonal control and opening) are positively correlated with adaptive behavior, and, conversely, the results revealed that the personality trait of rigidity is negatively correlated with adaptive behavior. Studies also reported that Openness to Experience has a positive impact on adaptive performance [31] and on team adaptiveness [32]. Based on the above research, we expected the following:

H₂₁. Openness to experience will be positively associated with employee adaptiveness in call centers.

To our knowledge, no prior research in general sales or in call centers has specifically addressed the relationship between the remaining four Five-Factor traits (Conscientiousness, Agreeableness, Extraversion and Emotional stability) and adaptive behavior.

People who score high on Conscientiousness seem to be hardworking, self-disciplined, task focused, and achievement oriented [33,34]. Prior studies have reported that this trait is associated with service orientation [35], and it is an important predictor of interpersonal interactions in service settings [28]. Recently, Brown and O’Donnell [36] showed that conscientiousness is a significant predictor of learning goal orientation, which in turn is recognized as a determinant of adaptiveness [37]. Guided by these studies, we expected the following:

H₂₂. Conscientiousness will be positively associated with employee adaptiveness in call centers.

An individual who scores high in Agreeableness is friendly, cooperative, flexible, and has successful relationships with others [3,38]. Hogan et al. [35] have reported that Agreeableness is associated with service orientation. Mount et al.’s (1998) meta-analysis showed that Agreeableness is a determinant for positive interpersonal interactions in the workplace. In addition, Bonzionelos [39] indicates that agreeableness is related to the tendency to care for others. These qualities may be particularly relevant for employee who need to interact with each customer on a case-by-case basis. Therefore, we conclude the following:

H₂₃. Agreeableness will be positively associated with employee adaptiveness in call centers.

Extroversion is characterized by sociability, activeness, energy, and assertiveness [17]. Onyemah (2008) [40] showed that extraversion can be linked to proactive behavior. Furthermore, some studies (e.g., [33,40,41]) have reported that extraversion is positively related to learning goal orientation. Finally, Judge et al. [42] highlighted that extraversion helps individuals to effectively communicate interpersonally. This disposition has the potential for increased success in adaptiveness selling. Based on these findings, we expect the following:

H₂₄. Extroversion will be positively associated with employee adaptiveness in call centers.

Mount et al. [28] reported that people who score high in Emotional stability seem to have good interpersonal interactions in service settings and are able to fit their personalities to any particular context. Prior studies have also reported that Emotional stability is associated with service orientation [35]. Therefore, we expect the following:

H₂₅. Emotional stability will be positively associated with employee adaptiveness in call centers.

In sum, in this study we expect the following:

H₂. Five-Factor traits will be positively associated with employee adaptiveness in call centers.

The impact of adaptive behavior on sales performance has been studied widely in sales literature. According to Weitz et al. [43], the adaptive selling framework and the practice of adaptive selling have a positive impact on employee performance. However, mixed results have emerged from empirical studies. Several salesforce researchers (e.g., [13,14,44,45]) have shown that adaptive behaviors positively influence performance. However, a few authors (e.g., [46,47]) have found no relationship between adaptiveness and performance [44].

In this study, there are at least two reasons to conclude that employee adaptiveness in call centers will positively affect performance with customers. First, the qualitative call center research of Eveleth and Morris [15] highlighted the idea that many employees consider adaptive behavior to be critical for effectiveness in a call center environment. Second, Hunter and Perreault [25] showed that practicing adaptive behavior improves performance with customers. Therefore, we expect the following:

H₃. Adaptive behavior will be positively associated with employee performance with customers in call centers.

Despite the many studies that have been conducted between personality traits and performance and between adaptive behavior and performance in both marketing literature and psychology, our knowledge about the relationship between those constructs in call centers is until recently quite limited. Based on the theorizing of the previous hypotheses, we expect that adaptive behavior mediates the relationship between Five-Factor traits and employee performance with customers at call centers. Specifically, we postulate that Conscientiousness, Agreeableness, Extraversion, Emotional Stability and Openness to Experience lead to greater employee adaptiveness in call centers, which in turn affect employee performance with customers. This mediating relationship is consistent with prior studies (e.g., [3,4,6]), which noted that mediating variables can explain uncounted variances between personality traits and performance. This mediating relationship is also consistent with recent developments in personality traits literature [20,48]. These studies suggest that a hierarchical model can better explain performance variability than a direct model between personality traits and employee performance. Therefore, we expect the following:

H₄. Adaptive behavior will mediate the relationship between Five-Factor traits and employee performance with customers in call centers.

H₄₁. Adaptive behavior will mediate the relationship between conscientiousness and employee performance with customers in call centers.

H₄₂. Adaptive behavior will mediate the relationship between agreeableness and employee performance with customers in call centers.

H₄₃. Adaptive behavior will mediate the relationship between extraversion and employee performance with customers in call centers.

H₄₄. Adaptive behavior will mediate the relationship between emotional stability and employee performance with customers in call centers.

H₄₅. Adaptive behavior will mediate the relationship between openness to experience and employee performance with customers in call centers.

The Respondents were drawn from an important call center in New Brunswick, Canada. This call center addresses the principal Canadian internet and telephony organizations. A cover letter and questionnaire were sent to 200 employees. Completed questionnaires were received from 108 employees. Female respondents constituted 58.3% of the sample. 41.7% of respondents were young (less than 24 years old), and only 42.1% of employees held BAA degrees.

The questionnaire was developed by integrating questions used in previous researches. Table 1 summarizes the scale items. The scale for performance with customers comprises five items developed by Hunter and Perreault [25]. The Five-Factor traits (Conscientiousness, Agreeableness, Extraversion, Emotional Stability and Openness to Experience) were measured using 16 items used by Sawyerr et al. [22] and developed by Goldberg [49]. Finally, adaptive behavior was determined based upon the four items used by Hunter and Perreault [25] and developed by Spiro and Weitz [13].

In this study, the method of structural equations is adopted as an analysis strategy, as recommended by Roussel et al. [50] for numerous assumed relations between variables. In the literature, there are primarily two types of statistical approaches based on structural equations: Linear Structural Relationships (“LSR”) and Partial Least Squares (“PLS”). In this research, we adopted the PLS approach for three reasons. First, the size of our sample is insufficient to render an adequate analysis by the Lisrel approach. Second, the PLS approach tolerates non-normality and especially multicollinearity [51]. Finally, beta PLS path coefficient estimation is more accurate than the Lisrel approach [51].

To test mediation, this study followed the procedures suggested by Baron and Kenny [52]. This method involves establishing four equations of regression that link up three types of variables. The first regression determines whether the independent variable has a direct effect on the dependent variable. The second regression examines whether the independent variable has a hitting action on the mediating variable. The third regression analyzes whether the mediating variable has an impact on the dependent variable. Finally, the last regression examines whether the influence of the independent variables on the dependent variable is considerably reduced when the mediating variable is added. In this respect, the test of Sobel will be used to prove if the reduction of the effect of independent variables in the dependent variable in the presence of the mediator variable is significant. This test involves calculating the z value (z = (a*b)/√(b²* s_a² + a²*s_b²); a and b: designed the non-standardized regression coefficients due respectively to the effect of the independent variable on the mediating variable and the effect of the mediating variable on the dependent variable; s_a and s_b: standard errors of a and b).

As recommended by Anderson and Gerbing [53], we developed a measurement model before estimating the structural paths to test the hypothesized relationships between constructs. Unidimensionality was assessed prior to examining reliability and validity [54]. To assess the unidimensionality of the constructs, all independent and dependent variables were analyzed by using varimax rotation, taking three criteria into consideration: 1) all items having a lower community under 0.4 will be eliminated; 2) the factors extracted are those that have a value higher than 1; and 3) Cronbach’s exceeded the minimum acceptable values, which was 0.7. The results illustrate that all constructs are unidimensional. As Table 1 reports, all items had a significant loading on corresponding constructs because the lowest l was 0.71 (>0.7) and the lowest t-value was 6.53 (>1.96). Using Chin’s [51] criteria, the convergent validity of the constructs is acceptable. All composite reliabilities, ranging from 0.90 to 0.93, are also above the cut-off value of 0.70, and the extracted variances [55] were higher than 0.5. Therefore, reliability is also acceptable.

The study then assessed the discriminant validity of the key constructs by using the procedures recommended by Fornell and Larker [55]. It consisted of checking whether the square roots of the AVE values were consistently greater than all corresponding correlations as shown in Table 2. As reported in Table 2, the analysis showed that each construct shares more variance with its corresponding measures than it shares with other constructs in the model. Therefore, all constructs reveal an acceptable discriminant validity.

The hypotheses were tested by simultaneously testing the proposed relationships using Partial Least Squares (PLS) path modelling. PLS is a prediction-oriented, variance-based

Table 1. Survey items, reliability and validity for measures in the study.

Table 2. Discriminant validity.

Diagonal entries show the square roots of average variance extracted, others represent correlation coefficients.

approach to Structural Equation Modelling (SEM), which makes very few assumptions about the distribution of the variables, and requires relatively few observations, compared to more traditional Maximum Likelihood (ML) SEM techniques such as LISREL. PLS has been chosen over LISREL because PLS is highly appropriate for complexes predictive models [52]. Furthermore, our sample size (N = 108) is small, then it is inadequate for LISREL analysis.

We used specifically SmartPLS, to estimate both the main and mediation effects stated in the hypotheses shown in Figure 1. The results are reported in Table 3.

Table 3. Main and mediating effects.

^**p < 0.05, ^*p < 0.1.