Predicting Performance of Call Center Staff: The Role of Cognitive Ability and Emotional Intelligence

Abstract

This study examined the relationship between cognitive ability (IQ) and emotional intelligence (EQ) in predicting a range of different performance metrics from a call centre environment. In all, 303 call centre staff completed multi-dimensional measures of both EQ and IQ. We also had recorded nine performance data measures for each individual over a 12-month period. There were a few significant correlations with IQ (4/35) and a few more with EQ (4/28), though all EQ measures were related to “Errors Made over the year”. The performance metric that had most correlates was Average Handling Time (AHT) relating to speed of working. The number of errors an employee made was significantly positively correlated with all four EQ factors. Correlational and Structural Equation Model (SEM) analysis highlighted the importance of analysing performance metrics as distinct variables, finding contradictory evidence in the sense that some individual difference factors correlated positively with some and negatively with other outcome measures. The results are discussed in relation to the theoretical implications for researchers interested in analysing call centre performance, and also practical implications for organisations with call centres.

1. Introduction

This study looks at the relationship between personality, intelligence and work performance. Specifically, it was concerned with the relationship between the traits of emotional intelligence (EQ), cognitive ability (IQ), and objective and subjective measures of job performance in a call centre. A number of studies have examined the relationship between EQ and IQ and generally concluded that the relationship is not strong and often negative [1,2,3,4]. We were less interested in the relationship between these two factors but rather, how well they independently predicted actual work performance in a call centre. The data in this study come from a call centre which provided excellent opportunities to explore these relationships because of the amount and quality of performance data which were regularly collected and recorded. Because of problems associated with the management of call centres, there remains a great interest in the topic [5,6,7,8,9,10]. This study follows a number of recent studies that have looked at both EQ and IQ correlates and predictors of work outcomes [11,12].

1.1. Call Centres

The role of call centres in modern organisations is becoming increasingly important and prevalent [13,14]. They serve to act as the primary point of contact for customer interaction [15,16,17], being set up to deliver customer-related services remotely and to replace the need for face-to-face interactions [18]. Call centre employees can interact with as many as 60 to 250 clients during their shift [19], which means the job can be very stressful [20].

The perception that customers have of the organisation is largely influenced and determined by the interactions they have with these employees [21]. Additionally, because of the frequency of contact with customers [22,23], the frequent repetitiveness of the role and high level of performance monitoring [13,24], it is often regarded as one of the most stressful, and often poorly paid jobs in modern businesses [25,26,27]. Consequently, there has been a growing interest academically to understand how to select, motivate and manage the performance of call centre employees [28]. Most importantly, compared to many other sectors, call centres tend to keep multiple, empirical recordings of all staff’s daily behaviour, which is excellent material to examine individual correlates of their performance. It is indeed rare to have such excellent behavioural performance data for any job allowing researchers to look at individual difference correlates of work behaviour.

Previous research has looked at the relationship between employee personality and performance in call centre environments [29,30]. The job itself is widely regarded as highly emotionally labouring for employees, requiring them to deal with often difficult and angry clients, giving rise to a range of “people issues” (e.g., Hechanova [31]). Therefore, certain personality traits have been hypothesized to impact call centre performance as a function of their preference for social interaction, as well as their capacity to respond to stress and pressure [32]. These features are not exclusive to call centre staff, which would make the results generalizable to other service jobs. Equally, it should be recognized that not all call centres are the same and that some do not appear to be very stressful. Some are ‘call-in’, where staff respond to clients, while others are ‘call-out’, where staff (often cold call) to survey and sell. Both can be very stressful, particularly the former. In this study, we used a medium sized ‘call in’ centre to collect our data.

However, studies that have utilized Big Five models of personality have found little relationship between personality and call centre performance. For instance, Sawyerr et al. [33] used structural equation modelling (SEM) and found that only Openness-to-Experience was negatively predictive of job performance in call centre environments. Previous research has argued that the link between personality and call centre performance is largely moderated by the working environment; the specific working culture has an impact on which, if any, personality traits allow these employees to thrive as a function of being aligned with the outlined working practices (e.g., Bing and Lounsbery [34], Mount et al. [35], Sawyerr et al. [33]).

1.2. Emotional Intelligence

Personality traits may not fully explain performance in call centre environments, potentially because these traits are not specific enough to assess the emotion-related demands of the job. For instance, the importance of emotional adjustment, resistance to stress, and not being overwhelmed by pressure have been cited as important characteristics for better performance and lower attrition (e.g., Higgs [32]). These characteristics have been linked to the now well-known concept of emotional intelligence [36].

There is a vast literature on emotional intelligence and its relationship to work outcomes [37,38,39]. Although there is criticism and concern surrounding models of trait EQ, recent meta-analyses have indicated that trait EQ shows significant incremental predictive power over “bright personality” and cognitive ability, providing support that trait EQ represents a meaningful and distinct construct [40]. One taxonomy of trait EQ is the Trait Emotional Intelligence Questionnaire (TEIQue) [41,42], comprised of four factors latent factors (from 15 facets) of EQ [43]. The TEIQue has shown high incremental validity to predict emotional intelligence over and above other personality factors (e.g., Mikolajczak et al. [44]). The present study therefore uses the TEIQue.

Previous work has demonstrated the role of trait EQ in predicting job performance (e.g., Abraham [45], Sy et al. [46], Varca [47]). High emotional intelligence also predicts improved job performance [46,48,49,50], especially when the job requires emotional labour, which is when employees must alter their emotional expressions in order to meet the social norms of the organization [37]. There is indeed a growing body of evidence showing that job performance is related to EQ [51,52]. EQ in call centre employees has been associated with increased job satisfaction [46], job advancement [53], organisational commitment as well as lower turnover intentions (e.g., Law et al. [54], Wong & Law [55]) and lower levels of stress [36]. We therefore expect to find positive correlations in this study.

In an important meta-analysis, Pirsoul et al. [56] found that emotional intelligence was significantly related to career adaptability, career decision-making self-efficacy, entrepreneurial self-efficacy, salary, career commitment, career decision-making difficulties, career satisfaction, entrepreneurial intentions, and turnover intentions. Thus, EQ relates to a wide number of organizational outcomes.

This study looks to extend previous research by examining socioemotional traits, namely trait EQ, to assess whether using an EQ-taxonomy of personality is more capable of explaining call centre job performance.

1.3. Cognitive Ability

Beyond EQ, assessing the cognitive ability (IQ) of potential employees has been used for decades as an indicator for predicting future work performance [57]. There is an extensive literature that indicates individual’s cognitive ability is related to both task-related and extra-role performance (see meta-analysis by Gonzalez-Mulé et al. [58]) with fluid intelligence benefiting employees through enhancing the ability to learn novel, job-relevant information and adapt to changing task requirements [59,60]. However, recent research has also found that job complexity moderates the relationship between cognitive ability and job performance, where cognitive ability was more predictive of performance in high-complexity jobs compared to low-complexity roles [61]. This is of particular interest as the nature of call centre roles are low in complexity. However, there is very little research looking at the role of cognitive ability in call centre performance. This study will extend previous research in this area by looking at this relationship specifically within a call centre context.

By examining a facet approach, this study will also aim to examine unique effects of specific facets to provide a granular insight into how these variables impact performance. Previous research has looked at job performance as a singular construct (i.e., total IQ score) but have used different metrics to characterise performance. For instance, Sawyerr et al. [33] measured performance on a single dimension based on information provided to customers and speed of response; Higgs [32] got managers to rate employee performance on a 1 to 5 scale; whilst Brown et al. [62] measured performance based on a 1 to 5 rating of quantity and quality of work performed. The problem with this approach is that the disparity in previous results could be attributable to the non-uniform approach to measuring performance in this environment.

1.4. This Study

This study is not based on a particular theory, but rather explores a unique data base. This study extended previous research on call centre employees’ performance by specifically analysing a range of performance-related factors (rather than performance as a unified, singular construct), with metrics including quality of calls, sales per call, errors made by employees, first call resolution, average handling time, lateness and absenteeism. We explored these relationships before looking at ability and personality correlates. That is, we were concerned with performance factors all measured behaviourally over a long period of time.

We essentially examined the relationship between IQ, EQ and performance both at the facet and the higher order level. Thus, we looked at five facets of IQ and four of EQ. This is an exploratory study, although we did expect to find that different performance variables related to different antecedents. From previous studies, we expected EQ factors to be more highly and significantly correlated with the call centre behaviour measures than the IQ factors.

We were inevitably constrained by the data set we had and the tests used which, however, we believe are both valid and reliable. Based on the previous literature, we hypothesized that self-control (especially emotion regulation and stress management) as well as emotionality (especially emotion perception) would be related to a number of performance measures. Equally, we hypothesised that reasoning and word meaning would relate to many of the call centre variables.

2. Method

2.1. Participants and Procedure

Participants in this study were outbound call centre employees from a North American financial institution. The data were collected as part of a research exercise with the client, who supplied the IQ and EQ tests (namely Thomas International). In all, 303 employees were initially assessed as a part of a psychological consultancy service. Participants’ assessment data were merged with their performance data for the previous 12 months. Participant data were merged on an anonymous unique identifier to maintain employee anonymity throughout the process. As a consequence, no other individual demographic data were gathered on participants. There was no selection bias, as all employees were assessed.

2.2. Materials

2.2.1. General Intelligence Assessment (GIA)

The GIA assesses individuals’ cognitive abilities, by measuring their speed and accuracy across five dimensions relevant to Reasoning (problem-solving abilities), Perceptual Speed (information checking accuracy), Number Speed (numeracy capability), Word Meaning (vocabulary-related knowledge), and Spatial Visualisation [63,64]. Its aim is to primarily measure mental speed of processing (fluid intelligence), rather than depth (crystallised intelligence). It measures the ability to process novel information and learn [62], and was derived from a battery of tests (see Collis et al. [65], Irvine et al. [66]). It has good internal reliability (test-retest correlations ranging from 0.75 to 0.86) and concurrent validity (correlations with Raven’s progressive matrices; r = 0.74 [62]).

2.2.2. TEIQue

The TEIQue is a measure of trait-based emotional intelligence on a 153-item questionnaire, measuring 15 facets of emotional intelligence [40]. The TEIQue assesses 4 factors: wellbeing, self-control, emotionality and sociability, with two independent facets which are internally reliable (i.e., Adaptability and Self-Motivation [43]). The model of the particular questionnaire has been confirmed through Confirmatory Factor Analyses (CFAs) across multiple languages and contexts [67].

2.2.3. Call Centre Performance Metrics

A range of performance metrics were gathered on the participants:

Average Handling Time [AHT]—the average time an employee spent on the phone with a customer.

Sales Per Call [SPC]—the percentage of calls on which an employee has achieved a sale.

First Call Resolution [FCR]—this represents the percentage rate that an employee is able to solve a customer query on the first call.

Quality Monitoring [QM]—an average rating of the quality of employee calls with customers, as rated by two managers in the organisation. Correlations between managers was r > 0.70, indicating good reliability.

Absenteeism—the number of days signed off that an employee has had in the past 12 months.

Late—the number of instances an employee has been logged as arriving late to work in the previous 12 months.

Total Errors—the number of errors an employee has logged the past 12 months.

Adherence—the proportion of time that, during the working day, the employee was registered as able to receive calls from customers.

Sick Days—the number of instances an employee has been logged being absent for illness in the previous 12 months.

3. Analysis

SPSS 24.0 (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY, USA: IBM Corp.) was used to clean and organize the data set. The Lavaan process was used in structural equation modelling (SEM) and various fit indices were applied: the χ²/df ratio, root mean squared error of approximation (RMSEA), Standardized Root Mean Residual (SRMR), and the comparative fit index (CFI).

4. Results

4.1. Factor Analysis

A factor analysis was conducted on the call centre performance metrics to identify any latent variables. The results indicated that three factors should be retained and which made sense. A principal axis factor analysis was chosen as the EFA technique due to the assumed non-normal distribution of the performance data. The EFA extracted a model that explained 35% of the variance with fit indices showing the model had a good fit of the data (Tucker Lewis Index [TLI] = 1.00; RMSEA = 0.056). Results of the EFA can be seen in

Table 1. Principal axis factor analysis with oblique rotation on performance metrics.

	Factors
	1	2	3
Low-priority errors	0.95
High-priority errors	0.30
Lates		0.79
Adherence		−0.49	−0.34
Sickness		0.37
Average handling time			0.56
Quality monitoring			0.44
Sales per call			0.40
First call resolution
SS loading	1.16	1.08	0.88
% Variance expl	13%	12%	10%
Cumulative % varia.	13%	25%	35%

Note: Variable in italics to indicate that loading did not exceed 0.30 cut-off for any factor.

.

The first factor represented the errors made by the call centre staff, with High- and Low-priority errors loading. The second factor, termed Absenteeism by the authors, represented factors identifying when employees were not available for work, including Lates, Sicks, and Adherence. The final factor represented Performance, representing factors that differentiated the traditional performance of employees, including Average handling time, Sales per call, and Quality monitoring. Interestingly, Average handling time positively loaded onto factor, indicating a positive relationship between length of call and performance. One performance variable–First call resolution—did not significantly load on to any of the three factors.

4.2. Correlations

Table 2. Correlations of GIA test scores and performance metrics.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
Overall GIA
Reasoning	0.75 ***
Perceptual Sp	0.81 ***	0.75 ***
Number S & A	0.78 ***	0.81 ***	0.45 ***
Word Meaning	0.73 ***	0.78 ***	0.5 ***	0.53 ***
Spatial Vis.	0.70 ***	0.73 ***	0.51 ***	0.49 ***	0.38 ***
AHT	0.04	−0.15 *	−0.03	−0.04	0.05	0.09
SPC	0.07	0.11	−0.01	0.02	0.12	0.03	0.22 ***
QM	0.14 *	0.03	0.12	0.08	0.17 **	0.15*	0.26 ***	0.12
FCR	−0.04	−0.02	−0.02	−0.02	−0.03	−0.07	0.08	0.12	0.19 **
Adherence	−0.12	−0.04	−0.09	−0.12	−0.13*	−0.07	−0.23 ***	−0.16 *	0.01	−0.17 **
Lates	−0.03	−0.07	0.05	−0.03	−0.04	−0.05	−0.01	0.02	−0.18 **	0.05	−0.38 ***
Sicks	−0.01	0.05	0.01	0.00	−0.04	−0.09	0.03	0.09	−0.03	0.12	−0.17 **	0.3 ***
LowPriorityError	−0.09	−0.04	−0.11	−0.03	−0.02	−0.15 *	−0.01	0.1	−0.29 ***	−0.26 ***	−0.09	0.02	−0.03
HighPriorityError	0.01	0.00	0.01	0.02	0.04	−0.03	−0.08	0.03	−0.13 *	−0.06	−0.02	0.05	−0.03	0.3 ***
Total Error	−0.05	−0.02	−0.06	−0.01	0.02	−0.1	−0.06	0.07	−0.25 ***	−0.19 **	−0.06	0.04	−0.04	0.77 ***	0.84 ***
Absent	−0.03	−0.01	0.04	−0.03	0.05	−0.09	0.01	0.06	−0.14 *	0.1	−0.33 ***	0.84 ***	0.77 ***	−0.01	0.02	0.01	0.01
Performance	−0.03	−0.15 *	−0.03	−0.04	0.05	0.09	0.21 ***	0.23 ***	0.26 ***	0.08	−0.23 ***	−0.01	0.03	−0.01	−0.08	−0.06	−0.06	0.01

Note: GIA = General Intelligence Assessment; AHT = Average Handling Time; SPC = Sales per Call; QM = Quality Monitoring; FCR = First Call Resolution; *** = p < 0.001; ** = p < 0.01; * = p < 0.05.

and

Table 3. Correlations of TEIQue factors scores and performance metrics.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17
Well Being
Self-Control	0.61 ***
Emotinoality	0.74 ***	0.64 ***
Sociability	0.51 ***	0.56 ***	0.63 ***
Overall TEIQue	0.86 ***	0.83 ***	0.9 ***	0.76 ***
AHT	−0.06	0.04	0.01	0.02	0.00
SPC	0.07	0.04	0.10	0.12	0.09	0.22 ***
QM	−0.14 *	−0.06	−0.08	−0.13*	−0.12	0.26 ***	0.12
FCR	−0.07	−0.02	−0.02	−0.1	−0.06	0.08	0.12	0.19 **
Adherence	0.01	0.00	−0.02	−0.09	−0.01	−0.23 ***	−0.16*	0.01	−0.17 **
Lates	0.00	0.05	0.05	0.02	0.03	−0.01	0.02	−0.18 **	0.05	−0.38 ***
Sicks	0.00	−0.10	0.04	−0.07	−0.04	0.03	0.09	−0.03	0.12	−0.17 **	0.3 ***
LowPriorityErrors	0.20 **	0.13 *	0.18 **	0.16 *	0.21 **	−0.01	0.10	−0.29 ***	−0.26 ***	−0.09	0.02	−0.03
HighPriorityErrors	0.09	0.12	0.10	0.14 *	0.14 *	−0.08	0.03	0.13 *	−0.06	−0.02	0.05	−0.03	0.30 ***
Total Error	0.17 **	0.15 *	0.17 **	0.18 **	0.21 ***	−0.06	0.07	−0.25 ***	−0.19 **	−0.06	0.04	−0.04	0.77 ***	0.84 ***
Absent	0.00	−0.03	0.05	−0.02	0.00	0.01	0.06	−0.14 *	0.10	−0.33 ***	0.84 ***	0.77 ***	−0.01	0.02	0.01	0.01
Performance	−0.06	0.04	0.01	0.02	0.00	0.21 ***	0.23 ***	0.26 ***	0.08	−0.23 ***	−0.01	0.03	−0.01	−0.08	−0.06	−0.06	0.01

Note: TEIQue: Emotional Intelligence Questionnaire; AHT = average handling time; SPC = sales per call; QM = quality monitoring; FCR = first call resolution; *** = p < 0.001; ** = p < 0.01; * = p < 0.05.

shows the results of correlation analysis between the GIA tests, TEIQue factors, and call centre performance. Latent factors created to represent the three-factor analysis variables were also included in the correlations. Inter-test correlations for the GIA ranged from 0.35 (Word Meaning and Spatial Visualisation) to 0.56 (Number Speed and Accuracy and Spatial Visualisation). TEIQue factor score correlations ranged from 0.52 (Well Being and Sociability) to 0.75 (Well Being and Emotionality). Employee AHT was significantly negatively correlated with Perceptual Speed. Employee call quality scores (QM) was significantly positively correlated with Reasoning score. Additionally, Spatial Visualization was positively correlated with SPC and QM. The number of errors an employee made was significantly positively correlated with all four TEIQue factors.

Table 4. Correlations between overall IQ, overall EQ, and performance factors.

	1	2	3	4
Overall IQ (GIA)
Overall EQ (TEIQue)	−0.07
Error	−0.03	0.21 ***
Absent	0.07	0.00	0.01
Performance	0.14 *	0.00	−0.06	0.01

IQ = cognitive ability; EQ = intelligence; *** = p < 0.001; * = p < 0.05.

and

Table 5. Correlations between overall IQ, the four TEIQue factors, and performance factors.

	1	2	3	4	5	6	7
Overall IQ (GIA)
Well Being	−0.10
Self-Control	0.00	0.61 ***
Emotionality	−0.08	0.74 ***	0.64 ***
Sociability	−0.04	0.51 ***	0.56 ***	0.63 ***
Error	−0.05	0.17 **	0.15 *	0.17 **	0.18 **
Absent	−0.03	0.00	−0.03	0.05	−0.02	0.01
Performance	−0.03	−0.06	0.04	0.01	0.02	−0.06	0.01

*** = p < 0.001; ** = p < 0.01; * = p < 0.05.

show the overall facet findings. Table 5 shows that the total IQ factor is positively significantly correlated with the Performance Factor and Emotional Intelligence with the Error factor. Table 5 shows the only significant correlations was between the four EQ facets and the Error factor. Overall, what these tables also show is that essentially there was no statistical relationship between EQ and IQ, which has been found in many other studies.

4.3. Structural Equation Modelling

SEM was used to as a way to analyse the role of cognitive ability and emotional intelligence on a range of employee performance metrics. All variables were entered as observed variables, including the five GIA subtest (Reasoning, Perceptual Speed, Number Speed & Accuracy, Word Meaning, and Spatial Visualization), the 15 TEIQue facets, and the seven employee performance metrics (Average Call Handling Time [AHT], First Call Resolution [FCR], Quality Monitoring [QM], Absenteeism, Lateness, Errors, and Sales per Call [SPC]). Non-significant regressions were removed in a backward-elimination fashion. The model was re-tested until only significant terms remained. Lateness was entirely removed from the model as a result.

The results of the model can be found in Figure 1. The chi-square statistic was non-significant, indicating the model did not significantly differ from the data: χ²(50) = 34.3, p = 0.956. Additionally, other fitness indices suggested that the model was an excellent fit of the data: χ²/df = 0.69; CFI = 1.00; RMSEA = 0.00; SRMR = 0.025.

5. Discussion

The results of this study highlight the importance of both fluid and emotional intelligence on the performance of call centre employees, offering a novel insight by using SEM to analyse a range of core job metrics. However, it should be acknowledged that, overall, neither IQ nor EQ was closely related to all the performance variables, suggesting that other factors such as work history, corporate culture or social support may play an important role in determining individual performance.

AHT is often used as a key metric for performance in call centre employees (e.g., Koole and Mandelbaum [68]); as employees are estimated to have contact with between 60 to 250 clients per shift [19], the average amount of time spent with each customer offers a proximal assessment of efficiency. The results suggest that both fluid and emotional intelligence were significant contributors to the AHT of the employees in this organization. It was found that employees that held more positive perceptions of the future (Optimism) and were better at controlling their emotional states (Emotion Regulation) had lower AHTs and were more efficient. Previous research has highlighted the important role that regulating one’s emotion has on job performance (e.g., Shamsuddin and Rahman [52]), as employees are able to calmly but effectively respond to client queries; this would be especially important for call centre employees frequently handling frustrated and annoyed customers.

Higher optimism scores suggest that employees will approach each call with positive expectations as to how it will go, giving them the persistence required to “bounce back” when they are unsuccessful. Additionally, dispositional Optimism usually gives employees the ability to more quickly see a positive outcome to customer concerns. In this sense, it is self-fulling.

Interestingly, employees that were interested and enjoyed creating and maintaining new interpersonal relationships (Relationships) and were better able to cope with stress (Stress Management) had higher AHTs. It could be that lower stress management would be related to a greater sense of urgency for the employee to remove themselves from a stressful situation. This could explain the negative impact of stress management in this study. Additionally, the Relationships finding suggests that employees with higher relationships scores were too invested in building personal relationships rather than resolving calls quickly.

High fluid intelligence, particularly in terms of employees’ ability to problem solve quickly (Reasoning) was found predictive of quick AHT, as staff were more quickly able to understand novel queries and find adequate resolutions. Of note, the ability to articulate oneself (Word Meaning) actually had a positive relationship with AHT, perhaps suggesting that people scoring highly in this trait were using vocabulary over the register of the average customer. Additionally, Spatial Visualisation—the ability to mentally conceptualise and manipulate problems—was related to higher AHT, potentially suggesting that these employees overcomplicate issues for customers and spend additional time trying to explain their rationale.

Employees that had higher levels of fluid intelligence—particularly those who were quicker at evaluating vocabulary and word-related knowledge (Word Meaning), and who were more willing and capable to talk about and utilise their emotions (Emotion Expression)—were found to have a higher rate of sales. This suggests that employees, to be successful, need to be able to quickly respond to client queries and concerns, and respond in an accurate manner. They also need to be competent at articulating and emoting the value of a sale to customers in order to be successful. Additionally, previous research has found that personality traits such as agreeableness are linked to sales performance [69]. This study extends our understanding of this effect, finding that employees who are able to communicate with, and use, emotion effectively are most successful.

As with sales, employees that had higher Word Meaning-related fluid intelligence had the quality of their calls rated higher by their managers. Additionally, employees had higher QM ratings if they were more diplomatic and soft in their communication styles (lower Assertiveness). This extends previous literature findings that employees with the higher trait of Social Boldness have lower sales performance and results [69]. These results suggest that a reason for this could be the quality of these calls are lower, having a negative impact on overall sales performance.

The results of the SEM also found that increased errors in the role were predicted by higher levels of Social Awareness; having a preference and capacity for networking and socialising. The results suggest that these employees emphasise the socialising aspect of their role over following protocol or the ‘right way’ of doing the job, and therefore end up making more errors. However, as little research has looked at the predictors of error in a call centre environment, further research is needed to explore these theoretical links. Thus, social awareness and skills may be a double-edged sword in the sense that staff are as happy and skillful in dealing with colleagues as clients.

It has been previously argued that the strongest predictor of future absenteeism is a previous history of absenteeism. However, these results extend a growing literature that there is a potential dispositional basis for absenteeism [70], with EQ predicting absenteeism rates in this call centre. Absenteeism was higher in employees with more positive perceptions of the future (Optimism) but they also make decisions quicker and without as much forethought (lower Impulse Control). Previous research has found a link between lower emotional intelligence [71] with presenteeism and absenteeism at work as a function of boredom. The results of this study offer an initial insight into this relationship, but further research is needed to understand the potential negative impact of emotional intelligence on call centre absenteeism. Future research could look at the moderating impact of workplace repetitiveness or proneness to boredom [72] on absenteeism, or the potential curvilinear relationship between EQ and absenteeism at work, proposing a potential ‘optimal’ level of EQ for the call centre environment.

Of particular note on this study were the seemingly contradictory findings. Employees who displayed high verbal articulacy (Word Meaning) had slower AHT, but higher SPC and higher QM. While they were not getting through as many calls, they were making more sales and were rated more highly by their managers. AHT is frequently used as a measure of performance in call centres [68], though perhaps this is not a reliable proxy. It is likely that making sales and ensuring good customer service takes longer and so an emphasis within a call centre on call volume may inadvertently promote behaviours that are counterproductive to company goals. This result has important implications for how organizations should prioritize performance metrics. Clearly, staff could “trade-off” speed for quality and need to think about how they weight performance variables, which inevitably effects how they select staff.

Another interesting finding was that an ability to stay focused and remain calm (Emotion Regulation) predicted quicker AHT, but poorer FCR. People lower on this trait were terminating calls quicker, but they were less likely to resolve them the first time. It is likely that being less able to control one’s own emotions resulted in employees being less able to resolve calls from more challenging customers.

What these contradictions highlight is the complexity of a multitude of ways to track and measure workplace performance. This is true in most jobs, but perhaps most critical in call centre-type work. Defining success in a job role is nuanced and care should be taken to ensure this is done with due rationale. The literature on call centre performance has typically been assessed as a unified but unspecified construct in research. This study shows the importance of looking at distinct aspects of performance separately. For instance, the contradictory benefit of Word Meaning on AHT and compared to sales or call quality shows the potential trade-off that organizations need to make when using this information to inform selection criteria. Organizations utilising the type of research exhibited in this study should look to focus on the core aspects of performance that are important to those roles and investigate these metrics in a distinct capacity.

The theoretical implications of this paper are two-fold. Firstly, the results of this study reinforce the notion that emotional intelligence has a significant impact in job roles that are regarded as emotionally labouring. Additionally, this study has provided a more nuanced insight by specifying specific traits that have a contributing role. Secondly, this study has demonstrated the important link between cognitive ability and the multitude of job performance metrics, showing how these constructs can play both a beneficial and hindering role. Previous work has theorized the relationship between cognitive ability and performance as positive due to the additional capacity, as well as the ability to quickly process and accurately apply knowledge and expertise to a task. However, this study demonstrates the need for a nuanced understanding of this relationship, finding that there are instances were these additional capabilities can negatively impact core job performance indicators. Clearly, additional research is needed to explore this concept, testing how moderating factors (e.g., job complexity, organizational culture) would have an impact on exacerbating or ameliorating this negative relationship.

The results of this study have used SEM to show the important, unique roles that both cognitive and fluid intelligence play on a range of call centre performance metrics. Organizations involved in recruiting call centre employees that are looking to make hiring decisions that best impact performance can do this by introducing measures of both emotional and cognitive intelligence into their selection criteria. Introducing these may prove to be beneficial to organizations in light of the high amount of cost associated with staff in call centres (e.g., Richardson and Marshall, [72]) as well as the increasing importance placed on managing people issues as a part of modern business.

6. Implications

There are a number of implications of this research for call centre managers. First, it is important to prioritise the behavioural outcome variables that are assessed. Often, there has to be a trade-off between speed and accuracy, as well as client satisfaction and economic return, which are often determined by different personal call centre staff factors. For instance, some may be happy to trade off AHT for accuracy, though all would, no doubt, prefer staff not to be regularly late or absent. Second, it would seem that EQ is a more relevant predictor of call centre staff behaviour than IQ. Whilst our results highlighted the importance of verbal IQ, the job clearly does not usually involve higher-order processing. In this sense, it may be more beneficial to invest more in personality than ability assessment at selection. Third, given the size of the correlation in this study there seem to be other more important factors that account for call centre output. These may include other personality factors such as Tolerance of Ambiguity, Adjustment and Conscientiousness, all of which have found to be related to success in different jobs [73,74].

7. Limitations

This study is not without limitations. Firstly, this study only collected data on one organisation, meaning there are caveats to the applicability and generalisability of the results to other industries, other call centres and in different countries. To test this, future research is needed, particularly by looking at whether these results remain consistent from differing industries with similar job performance metrics. It would also be desirable to repeat the study with different measures of both IQ and EQ, as these may show subtly different results. In addition, it would have been desirable to have much more data on the participants, such as their career history and any performance data.