RESEARCH METHODOLOGY - The Impact of Emotional Labor on Counterproductive Work Behavior: The Mo

This chapter introduces the research methodology implemented in this study to test the hypotheses and to achieve the research objectives. It outlines the different aspects of methodological discussions such as the research framework, research design, research sampling, the measurement instruments, pilot test, data collection, data analysis and the research procedures.

Research Framework

This study’s research framework is developed from a combinations of previous literature reviews and other Models and theories to accomplish the objective of this research. This study examines the impact of emotional labor on counterproductive work behavior and how mindfulness and perceived organizational support can be utilized to moderate the effect that it has on that relationship. Figure 3.1 illustrates Emotional Labor surface acting and deep acting as the independent variables; Counterproductive Work Behavior as the dependent variable; Mindfulness and Perceived Organizational Support as the moderators, which is proposed to have a moderating effect on the relationship between emotional labor and counterproductive work behavior. The hypotheses proposed in this study are as follows:

Hypothesis 1a: Emotional labor surface acting is positively related to CWB Hypothesis 1b: Emotional labor deep acting is negatively related to CWB

Hypothesis 2a: Mindfulness attention weakens the positive effect of emotional labor surface acting on CWB

Hypothesis 2b: Mindfulness present focus weakens the positive effect of emotional labor surface acting on CWB

Hypothesis 2c: Mindfulness acceptance weakens the positive effect of emotional labor surface acting on CWB

Hypothesis 2d: Mindfulness awareness weakens the positive effect of emotional labor surface acting on CWB

Hypothesis 2e: Mindfulness attention weakens the effect of emotional labor deep acting on CWB

Hypothesis 2f: Mindfulness present focus weakens the effect of emotional labor deep acting on CWB

Hypothesis 2g: Mindfulness awareness weakens the effect of emotional labor deep acting on CWB

Hypothesis 2h: Mindfulness acceptance weakens the effect of emotional labor deep acting on CWB

Hypothesis 3a: Perceived organizational support weakens the positive effect of emotional labor surface acting on CWB

Hypothesis 3b: Perceived organizational support weakens the effect of emotional labor deep acting on CWB

Research Procedure

This part provides a description of the process that this study went through. Figure 3.2 illustrates the research process for this research. First, the researcher reviewed literatures to identify a topic that would contribute to the existing literatures and also have a positive contribution towards the organizational structure in Belize. Then, the research problem and purpose of the research were identified. After the identification of a proper problem and purpose, the research measurements and items of each variable were established. The next step after the items are developed was a pilot test that was conducted to ensure the validity and reliably of the instrument. Finally, the data was collected through online questionnaires from the targeted sample.

The valid data was analyzed by SPSS and AMOS, conclusions and suggestions were then discussed.

Figure 3.2. Research procedure of this study

Conclusion and Findings Data Analysis

Sample and Data Collection Pilot Test

Development of Research Instruments and Measurments Development of Research Framework

Identification of Research Problem and Purpose of the Study Identification of Research Topic

Literature Review

Pilot Study

A pilot test was conducted prior to data collection. The researcher first conducted a pilot test to assess the reliability and face validity of the questionnaire (Akeem, 2015). A sample of thirty (34) employees working in the service industry. The designed survey included demographic questions and questions that measured the four variables of research interest. After the pilot test, the researcher made corrections to the screening and demographic items. In addition to minor grammatical errors within the survey instrument.

The questionnaire was administered from the beginning of February to Mid-March via the Online mode using Google Forms. The acceptable Cronbach’s alpha for reliability analysis is > .7 (Nunnally, 1978). For the pilot test, Cronbach’s alpha for emotional labor deep acting was .93, emotional labor surface acting was .89, counterproductive work behavior (CWB) was .92, mindfulness was .92 and lastly perceived organizational support was .95. All variables had a Cronbach’s alpha of over .89, which is interpreted as being high acceptable values.

Measurement

The online questionnaire hosted via Google Forms consisted of a total of fifty-four (54) questions, available in Appendix A: Three (3) Screening Questions; Thirteen (13) Demographic questions; eight (8) Emotional Labor questions; ten (10) Counterproductive work behavior questions; twelve (12) Mindfulness questions and eight (8) Perceived Organizational support questions. The survey questions were designed based on a well-developed instrument that had a high value of Cronbach’s alpha. The arrangement of these items in a sequential order was as follows, first the independent variable (EL- Deep Acting and Surface Acting), then the dependent variable (CWB), followed by the two moderating variables (Mindfulness and Perceived Organizational Support).

Research Sample and Data Collection

The participants of this study consisted of the frontline employees from Belize that work in various service industries. Frontline employees are often seen as the spokesperson for the companies that they work and are expected to meet customers’ requests (Cambra-Fierro et al., 2014). The present research focuses on this position because of two reasons. First, the frontline

employee has direct contact with customers and would be more likely to experience emotional labor. Second, the repercussions of such an environment would impact the customers, the coworkers or upper management.

To achieve the research purpose, the research methodology applied in this study has been developed from the previous research used in quantitative approach. The online self-administered survey instrument was developed to collect data for all the variables including emotional labor, mindfulness, perceived organizational support and counterproductive work behavior. The purpose of research and guarantee of confidential were clearly determined in the introduction part on the first page. To avoid confusion and to filter out respondents that met the criteria, three screening questions were put in place.

The convenience sampling together with snowball sampling method were used to allow the author to reach a specific group of population (i.e. frontline employees) in certain numbers.

By convenience sampling method, the link of the google form survey was sent to several frontline employees from different sectors in the service industry. Later, by way of snowball sampling, those respondents assisted with gathering more responses by forwarding the link to their network members in the same career.

To guard against the multiple participation responses, the individual is allowed to access the questionnaire only once. The questionnaires contain some items which are to be reverse scored in order to cross-check answer validity.

Emotional Labor

According to Diefendorff et al. (2005), the measure contained eight items, with responses on a 7 point Likert scale ranging from ‘Strongly Disagree’ to ‘Strongly Agree.’ A sample item is ‘I put on a show at work.’ (Coefficient alphas for the two dimensions of surface acting and deep acting are both .85), see Table 3.1.

24 Table 3.1.

Emotional Labor Measurement

Code Questionnaire Component

EL_DA1 I try to actually experience the emotions that I must show to customers EL_DA2 I make an effort to feel the emotions that I need to display towards others EL_DA3 I work hard to feel the emotions that I need to show to customers

EL_DA4 I work at developing the feelings inside of me that I need to show to customers EL_SA1 I put on an act in order to deal with customers in an appropriate way

EL_SA2 I fake a good mood when interacting with customers

EL_SA3 I put on a “mask” in order to display the emotions I need for the job EL_SA4 I fake the emotions I show when dealing with customers

Note. Adapted from “The dimensionality and antecedents of emotional labor strategies,” by Diefendorff, J. M., Croyle, M. H., and Gosserand, R. H. (2005), Journal of Vocational Behavior, 66, 339–357.

Counterproductive Work Behavior

The ten items developed by Spector et al. (2010) was used to measure CWB. The response format is a 5-point Likert-type scale ranging from “1= Never” to “5=Every Day.” (Coefficient alphas for the two dimensions CWB- Organizational and CWB- Personal were both .86), see Table 3.2.

Table 3.2.

Counterproductive Work Behaviour Measurement Code Questionnaire Component

CPWB1 I purposely waste my employer’s materials/supplies CPWB2 I complain about insignificant things at work

CPWB3 I tell people outside the job what a lousy place I work for CPWB4 I come to work late without permission

CPWB5 I stay at home from work and said I was sick when I wasn’t CPWB6 I insult someone about their job performance

CPWB7 I make fun of someone’s personal life CPWB8 I ignore someone at work

CPWB9 I start an argument with someone at work CPWB10 I insult or make fun of someone at work

Note. Adapted from “Measurement artifacts in the assessment of counterproductive work behavior and organizational citizenship behavior: Do we know what we think we know?” by Spector, Bauer, & Fox, (2010), Journal of Applied Psychology, 95(4), 781-790.

Cognitive and Affective Mindfulness Scale

The twelve items scale developed by Feldman et al. (2007) was used to measure mindfulness. The response format is a 6-Point-Likert type scale ranging from “1= Almost Never” to “6= Almost Always”. Coefficient alphas is .77, see Table 3.3.

Table 3.3.

Cognitive and Affective Mindfulness Measurement Code Questionnaire Component

MA1 It is easy for me to concentrate on what I am doing MPF1R I am preoccupied by the future

MAC1 I can tolerate emotional pain MAC2 I can accept things I cannot change

MAW1 I can usually describe how I feel at the moment in considerable detail MA2R I am easily distracted

MPF2R I am preoccupied by the past

MAW2 It’s easy for me to keep track of my thoughts and feelings MAW3 I try to notice my thoughts without judging them

MAC3 I am able to accept the thoughts and feelings I have MPF3 I am able to focus on the present moment

MA3 I am able to pay close attention to one thing for a long period of time Note. Adapted from “Mindfulness and emotion regulation: The development and initial

validation of the Cognitive and Affective Mindfulness Scale Revised (CAMS-R),” by Feldman, G., Hayes, A., Kumar, S., Greeson, J., & Laurenceau, J. P. (2007). Journal of Psychopathology and Behavioral Assessment, 29(3), 177-190.

Perceived Organizational Support

The eight (8) item scale developed by Eisenberger et al. (1986), was used to measure perceived organizational support. The response format is a 5 Point-Likert type scale ranging from 1=

Strongly Disagree to 5= Strongly Agree. (Coefficient alpha is .90), see Table 3.4.

Table 3.4.

Perceived Organizational Support Measurement Code Questionnaire Component

POS1 The organization values my contribution to its well-being.

POS2R The organization fails to appreciate any extra effort from me.

POS3R The organization ignores any complaints from me.

POS4 The organization really cares about my well-being.

POS5R Even if I did the best job possible, the organization would fail to notice.

(Continued)

26 Table 3.4 (Continued)

Code Questionnaire Component

POS6 The organization cares about my general satisfaction at work.

POS7R The organization shows very little concern for me.

POS8 The organization takes pride in my accomplishments at work.

Note. Adapted from “Perceived organizational support,” by Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Journal of Applied Psychology, 71, 500-507.

Control Variables

Education was controlled in this research due to the positive correlations it had on the variables. Researchers have found that education plays a role on the employee’s decision to act counterproductively (An & Wang, 2016). Employees that possess a higher level of education are perceived as less likely to partake in counterproductive work behavior (NG & Feldman, 2009).

Therefore, the impact of education was controlled in the analysis to examine the valid and comprehensive influence of emotional labor on counterproductive work behavior with mindfulness and perceived organizational support as moderators.

Data analysis

The responses collected from the online questionnaire were analyzed by IBM SPSS Statistics 23. The following analysis techniques were adapted:

Descriptive Statistics Analysis

Descriptive statistics assists readers to have a clearer image of the overall data that was collected from the participants. This study utilized frequency distribution that displayed the percentages and numerical representations of the demographic information, that includes the demographics of participants. The central tendency and standard deviation of all of the variables were identified, including emotional labor, mindfulness, perceived organizational support, and counterproductive work behavior. Descriptive statistics also helps the researchers identify the sample characteristics that may have influenced their conclusions (Thompson, 2009). The frequency and percentage of these demographics are shown in Table 3.5.

The participants for this study are frontline workers from the various organizations in Belize, Central America. A total of 200 respondents were selected from a total of two hundred and 233 respondents who started the study. Also the thirty-three respondents that did not complete the

study were due to the three screeners set in place within the survey to verify that indeed the participant was a Belizean national, employed, and a frontline employee. Included are the screener questions in the questionnaire instrument completed by respondents, see Appendix 1.

Most of the participants were females (61.5%) and approximately eighty-seven percent (87.1%) were employed full-time. As well most of the samples were between the age range of twenty to thirty (20-30) years old (53.5%). Thirty-three percent (33%) of participants had been working in their organization between 1-3 years. The Belize public sector employs 64.5% of them, while the private sector employs 35.5%. In addition, 64.5% of the population considered themselves as “staff” in the hierarchical structure of their organization.

Table 3.5.

Participants Demographic Statistics (n=200)

Item Frequency Percentage Item Frequency Percentage

Gender Working Hours

28 Table 3.5. (Continued)

Item Frequency Percentage Item Frequency Percentage

Doctoral or > 7 3.5 Information... 7 3.5

Financial… 14 7.0

Employment Real Estate 3 1.5

Full-time 175 87.1 Professional, 10 5.0

Part-time 26 12.9 Support Service 19 9.5

Public Ad… 14 7.0

Job Position Education 55 27.5

Staff 129 64.5 Human Heal … 15 7.5

Supervisor 26 13.0 Art, Enter… 9 4.5

Mid-level Super… 22 11.0 Technology 9 4.5

Senior Executives 23 11.5

Service Years

Daily Customers < 1 Year 30 15.0

< 20 81 40.5 1-3 years 66 33.0

20-25 29 14.5 4-7 years 41 20.5

26-30 14 7.0 8-10 years 24 12.0

36-40 9 4.5

41-45 2 1.0 Job Tenure

46-50 3 1.5 <1 Year 45 22.5

> 50 53 26.5 1-3 Years 75 37.5

4-7 Years 38 19.0

Work Area 8-10 Years 14 7.0

Marketing 21 10.5 > 10 Years 28 14.0

Finance 36 18.0

Operations Man… 75 37.5 Previous Years

Human Resource 49 24.5 <1 Year 20 19.8

(Continued)

29 Table 3.5. (Continued)

Item Frequency Percentage Item Frequency Percentage

Information Tech… 19 9.5 1-3 Years 41 40.6

4-7 Years 21 20.8

8-10 Years 10 9.9

> 10 Years 9 8.9

Table 3.6.

CR and AVE values of Emotional Labor Deep Acting, Emotional Labor Surface Acting, Mindfulness, Perceived Organizational Support and Counterproductive Work Behavior

Variable CR AVE

Emotional Labor Deep Acting .856 0.6

Emotional Labor Surface Acting .866 0.6

Mindfulness .839 0.4

Perceived Organizational Support .859 0.5

Counterproductive Work Behavior .904 0.5

Construct Validity and Reliability

Construct reliability and validity was executed using AMOS combined with Excel stat tools from Gaskin (Paul & Gaskin, 2014). Threshold measures were used to establish the validity and reliability of the model as shown in Table 3.7. (Hair, Black , Babin, & Anderson, 2010). To assess construct validity, the researcher examined: Convergent validity. In order to measure each validity, the following measurements were used: Composite Reliability (CR) and Average Variance Extracted (AVE).

In order to calculate the validity and reliability of the research Model in AMOS, the researcher focused on the estimates: regression weights, shown in Table 3.6. The researcher utilized an Excel stat tools from Gaskin(2013) to calculate the validity and reliability table shown in Table 3.6. The result obtained illustrates all Composite Reliability (CR) values greater than the required threshold of 0.7 suggesting adequate reliability of the constructs.

To determine the Convergent validity researchers state that loading estimates that are statistically significant and provide a useful start in assessing the convergent validity of the measurement Model (Hair et al., 2010). Two criteria were used to determine the validity: (1) CR

has to be greater than the Average Variance Extracted (AVE) and (2) AVE has to be greater than 0.5. Based on the results illustrated in Table 3.7, 4 out of 5 variables met the criteria with the exception of mindfulness achieving a score of .371.

Table 3.7.

Threshold Guidelines for Construct Validity and Reliability

Measure Threshold Research Model

Reliability CR > 0.7 5/5

Convergent Validity CR > (AVE) 5/5

AVE > 0.5 4/5

Common Method Variance

Common method bias is caused by the measurement that is used for data collection within a study and not necessarily caused but the causes and effects within the Model being studied (Kock, 2015). With studies related to admitting to acting defiant in forms of counterproductive work behavior, the possibility that the implicit social desirability associated with answering the questions within a questionnaire in such a way that appeases society would likely cause the indicators to share a particular amount of common variance. Another factor would be the cross sectional design of the survey, notably the survey was administered using an online platform, this might have an influence on the results from the participants who may be unfamiliar with completing electronic survey platforms compared to those that are familiar.

For the purposes of this research, Harman’s single factor test was adapted as a technique to identify common method variance. SPSS was used to analyze the data set collected and received results of 21.5% percent chance of variance, which is less than fifty percent (50%) which is an acceptable amount that confirms reliability of the measurement (Podsakoff & Organ, 1986).

Correlation and Reliability Analysis

The reliability and Pearson correlation analyses were conducted to ensure the reliability of the measurements and to examine the correlations among the variables. Table 3.8 displays the mean, standard deviations and reliability for each variable and Pearson correlation coefficient among each variable. The acceptable Cronbach’s alpha is > 0.7 (Nunnally, 1978). The results show emotional labor deep acting had a Cronbach’s alpha of .87, emotional labor surface acting had

a .85, counterproductive work behavior had .87, mindfulness had .83, and perceived organizational support had a .87. All variables had a Cronbach’s alpha of above .83, which is a considered to be of high value.

The Pearson correlation coefficient in a sample is represented using r. The coefficient can take value from -1 through 0 to +1. If the sign displayed is positive, then a positive correlation exists. However, if the sign is negative, then a negative correlation exists and the opposite results are determined. A correlation of zero indicates that no linear association between the two variables exists (Sedgwick, 2012). As well, coefficient values r < 0.4 were considered low correlations. Also, coefficient vales between r > 0.4 and r < 0.7 were considered medium, while high correlation are r > 0.7. The correlation values are what indicated the strength of the relationship between variables. In this study, correlation analysis was used to examine the relationship between emotional labor, mindfulness, perceived organizational support and counterproductive work behavior.

In accordance with the results achieved by Bechtoldt et al. (2007), Table 3.6 presents counterproductive work behavior with a mean of around two (2), indicating that the participants admitted to partaking in defiant behavior at least once or twice. The results of the Pearson correlation analysis (Table 3.6.) showed that most of the major variables in this study had significant correlation coefficients. Counterproductive work behavior had a negative, low correlation to emotional labor deep acting (r = -.30, p < .001). At the same time, counterproductive work behavior had a medium, positive correlation with emotional labor surface acting (r = .44, p < 0.01). As for mindfulness, this variable had a low, negative correlation with emotional labor surface acting (r = -.22, p < 0.01) and counterproductive work behavior (r

= -.39, p < 0.01). While perceived organizational support had a low negative correlation with emotional labor surface acting (r = -.20, p < 0.01) and counterproductive work behavior (r = -.33, p < 0.01).

32 Table 3.8.

Mean, Standard Deviations, Correlations and Reliability (n=200)

Mean SD 1 2 3 4 5

1. ELDA 4.82 1.55 (.87)

2. ELSA 3.26 1.69 -.059 (.85)

3. CWB 1.62 .661 -.301** .448** (.87)

4. MINDFULNESS 4.36 .728 .300** -.219** -.386** (.83)

5. POS 3.36 .934 .192** -.200** -.329** .337** (.87)

Note. Cronbach’s Alpha are in boldface.

Note, n=200, *p< .05, **p< .01, ***p< .001; ELDA= Emotional Labor Deep Acting; ELSA=

Emotional Labor Surface Acting; CWB= Counterproductive Work Behavior; MA= Mindfulness Attention; MPF= Mindfulness Present Focus; MAC= Mindfulness Acceptance; MAW=

Mindfulness Awareness; POS= Perceived Organizational Support

Confirmatory Factor Analysis

For the confirmatory factor analysis (CFA) a model fit was conducted on the priori model with two independent variables, two moderators, and one dependent variable to verify how well the Model matches the data. The priori model diagram included symbols explained by researchers (Kline, 2010), and was used to create the Model to determine the goodness of fit. This study has a total of 204 respondents prior to running CFA. Before running AMOS the output tab was selected from the analysis properties icon, and standardized estimates, residual moments modification indices, test for normality and outliers were all selected to determine the model fit. There are various thresholds measure that are obtained from the results of the analysis which were used to

在文檔中 The Impact of Emotional Labor on Counterproductive Work Behavior: The Moderating Roles of Mindfulness and Perceived Organizational Support Among Frontline Employees in Belize (頁 26-48)