Chapter 3 CHAPTER 3 RESEARCH METHODOLOGY 3.1 Introduction The general intent of this exploratory and descriptive study was two-fold. First was the identification and exploration of the causal agents of attrition in the BPO sector of India. This involved an exhaustive study of the personal characteristics of the respondents; exploring the factors behind attrition, and then comparing these factors across the various personal characteristics. Second was the identification and exploration of the factors of retention in the BPO sector of India and testing the same. This involved detecting the factors through field survey; comparing them across the personal characteristics; and developing a regression model. It was followed by model validation i.e. studying the impact of the recommendations on employees willingness to stay and analysis of the perceived attitudes of the employees. This included assessing the levels of employee motivation, employee satisfaction, employee involvement, and life interest and work compatibility among the employees before and after the recommendations; and the relationship between the perceived attitudes and employees willingness to stay. The research methodology has to be robust in order to minimize errors in data collection and analysis. Owing to this, survey method has been chosen for data collection, as they are the most extensively used technique for data collection, especially in behavioral sciences [155]. This chapter describes the pilot study, participants of the study, instrumentation done for the study, data collection, and data analysis procedures of the entire study. 3.2 Pilot Study A pilot study was done on 100 BPO employees to find out whether the reasons for quitting and reasons for staying in a BPO were same or different. A questionnaire which consisted of 20 items was personally administered to 100 respondents. These items were sourced from the employees themselves. Interactions were held with employees at initial level to jot down their reasons for leaving a BPO organization and what exactly would motivate them to stay. Based on these interactions, a list of factors was prepared that were related to their work and work environment. Ph. D. Thesis 62
Research Methodology Respondents were asked to categorize them as A or B. They were required to mark A for those factors, the absence of which in the required form and volume would make them feel dissatisfied and unhappy about their job and hence increase their probability to quit. Correspondingly, they were asked to mark B for all those factors, the presence of which would satisfy them and motivate them to stay in the organization. The results showed that many of the respondents differentiated between factors that make them quit and factors that make them stay. The A factors were the attrition factors, i.e. the absence of these factors in the desired form and volume would prompt them to leave the job; and the B factors were the retention factors, i.e. the presence of these factors would make them feel satisfied and happy about their job, thus increasing their willingness to stay. Reducing the impact of negative characteristics of a typical BPO job may reduce their probability of quitting but will not ensure longevity of tenure. Similarly the presence of intrinsic characteristics of the work as clubbed in retention factors will increase the willingness of the employees to sustain in an organization; but their absence may not really prompt them to leave their jobs. Table 3.1 shows the number of respondents who rated the 20 factors as A and B respectively. Table 3.1: Number of respondents indicating attrition and retention factors Job Characteristics A B Attrition factors Retention factors Nature of job 85 15 Organizational Support 83 17 Organizational culture 76 24 Recognition 73 27 Salary 72 28 Skill Variety 69 31 Working hours 68 32 Basic Amenities 57 43 Relationship with colleagues 56 44 Job Security 55 45 Working Conditions 54 46 Relationship with boss 48 52 Growth Prospects 46 54 Feeling of accomplishment 19 81 Participation 15 85 Prestige associated with the work 15 85 Accountability 12 88 Authority 12 88 Involvement 10 90 Self-Esteem 8 92 63 Ph. D. Thesis
Chapter 3 Table 3.1 shows the number of respondents indicating the factors as Type A i.e. attrition factor and Type B i.e. retention factor. Seven factors viz. nature of job, organizational support, organizational culture, recognition, salary, skill variety, and working hours are highlighted under Type A factors i.e. attrition factors. It is because out of 100 respondents more than half of them have ticked them as Type A as they feel the absence of these factors in the desired form and volume would cause dissatisfaction and hence prompt them to leave the organization. These characteristics were fairly related to the work itself and the work environment. For example, 85 out of 100 respondents indicated nature of job as an attrition factor i.e. if the nature of job is not conducive to their personal and social aspirations, it would cause dissatisfaction and hence their intention of quitting would grow high. Similarly, 68 respondents out of 100 denoted working hours as Type A, because they feel unusual and extremely long working hours exhaust them and drain them out completely. Since their biological clock is not able to catch rhythm with the work clock, the incompatibility results in their quitting of jobs. There were 7 factors which were indicated as Type B i.e. retention factors by the respondents. Respondents stated that these factors if present make them feel happy and satisfied about their job and thus increase their willingness to stay. These factors were quite intrinsic to the work and the individual. The factors included feeling of accomplishment, participation, prestige associated with the work, accountability, authority, involvement, and self-esteem. More than 80% of the respondents agreed in unison that these factors contributed to their satisfaction from work, motivation and hence long term sustainability for e.g. feeling of accomplishment was indicated as Type B by 81 out of 100 respondents i.e. it was considered a retention factor; the presence of this factor would satisfy, motivate and make the employee happy. There were six factors which were almost evenly divided between A and B type. That means approximately one half of the respondents considered them responsible for attrition and other half considered them responsible for retention. For example, relationship with colleagues was indicated as Type A by 56 respondents and as Type B by 44 respondents. This shows that there is no clear delineation of this factor as an attrition or a retention factor. Other such factors included job security, relationship with boss, working conditions, growth prospects, and basic amenities. It was not clear from the responses whether to categorize them as attrition factors or retention factors. Thus, while designing the survey instrument, these factors were a part of both attrition and retention. Ph. D. Thesis 64
Research Methodology The graph in figure 3.1 clearly shows that self esteem, feeling of accomplishment, participation, prestige associated with the work, accountability, authority, and involvement have high indication of responses in the retention category. Similarly, nature of job, organizational support, organizational culture, recognition, salary, skill variety, and working hours have high indication of responses in the attrition category. Self-Esteem Involvement Authority Accountability Prestige associated with the work Participation Feeling of accomplishment Growth Prospects Relationship with boss Factors Working Conditions Job Security Relationship with colleagues Basic Amenities Working hours Skill Variety Salary Recognition Organizational culture Organizational Support Nature of job A Attrition factors 0 10 20 30 40 50 60 70 80 90 100 B Retention factors Number of respondents Figure 3.1: Graphical representation of the respondents indication 65 Ph. D. Thesis
Chapter 3 Figure 3.1 is the graphical representation of the data produced in table 3.1. The X axis shows the factors considered in the study and the Y axis indicates the number of respondents who ticked A and B. The graph clearly depicts factors like self esteem, involvement, authority, accountability, prestige, participation, and feeling of accomplishment have more respondents indicating as Type B. Factors like working hours, skill variety, salary, recognition, organizational culture, organizational support and nature of job have more respondents indicating as type A. The pilot study formed the pedestal for the research. It brought out the fact that factors that are responsible for attrition are different from factors that cause retention. Minimizing attrition will not necessarily enhance retention. Thus, managing attrition and retention are two different challenges for a BPO organization. Based on the findings of this pilot study, the survey instrument was designed. The detailed description of the instrument is discussed in section 3.3. 3.3 Instrumentation A questionnaire was intricately designed to tap the factors responsible for attrition, the factors that are expected to be present in a specific job for retention. The instrument was divided into 4 parts. Part I gathered information about the personal profile of the respondents which included their age, gender, education, marital status, and tenure. Part II consisted of questions about their reasons for change or probable change in their jobs. It also enquired if their current organization fulfilled their expectations. This was gathered to know about the factors behind attrition. Part III aimed at knowing what according to the respondents is important for their sustenance in an organization. The basic idea was to measure the degree to which certain job characteristics were expected to be present that would help them to sustain in their current job. And last of all, Part IV was about their overall perception of the work which included their level of satisfaction, level of motivation, level of involvement and level of life interest and work compatibility. All the aforementioned questions from Part II, III and IV were put on a 5 point Likert scale. Part II listed 21 items responsible for their change or probable change of job. These items were raken from the pilot study discussed in section 3.2. Respondents were asked to rate the items on a 5-point scale; 5 being most important and 1 being least important. Similarly Part III consisted of 20 items denoting various aspects of jobs responsible for their prolonged stay in the organization. Ph. D. Thesis 66
Research Methodology These items were also taken from the pilot study as discussed in section 3.2. These items were to be rated on a scale of 5 according to the importance of that particular job characteristic for the sustenance as perceived by the participant. Part IV was designed to assess the overall satisfaction from job; the overall motivation of the employee in the job; the degree to which the participant felt involved in the job; and the extent to which his life interest and his job were compatible; and if given an alternative with the same pay, would the participant continue in the same organization, leave the organization but remain in the same industry or shift to another industry altogether. To test the validity of the instrument, a study was done on 100 participants in the month of January 2008. Based on their responses, validity tests were done to check for the validity and usability of the instrument. Cronbach alpha, KMO measure of adequacy and Bartlett s test of sphericity were conducted. Cronbach alpha was calculated to measure the internal consistency reliability of the instrument. The cronbach alpha came as 0.898 for Part II and 0.963 for Part III thus the instrument was considered reliable for the study. Kaiser-Meyer-Olkin test was done to measure the homogeneity of variables and Bartlett's test of sphericity was done to test for the correlation among the variables used. The KMO value for part I of the instrument was 0.745, and for part II was 0.767, both of which are acceptable as a middling value. The Bartlett s test showed significant results for both the parts and hence the instrument was accepted for further study. Table 3.2 summarizes the entire result viz. cronbach alpha, KMO test values, and Bartletts s significance of the instrument. On getting quite meritorious results of the validity, the instrument was floated for data collection. No. of items Table 3.2: Tests of Validity of the questionnaire Cronbach s Alpha Kaiser-Meyer-Olkin Measure of Sampling Adequacy Bartlett's Test of Sphericity Part II 21.898.745 Part III 20.963.767 Approx. Chi- Square 6052.012 df 210.000 Sig..000 Approx. Chi- Square 12554.051 df 190.000 Sig..000 67 Ph. D. Thesis
Chapter 3 3.4 Data Collection The study aimed at employees working in the urban centric BPOs i.e. the BPOs operating in tier-i cities. Bangalore, NCR, and Mumbai are Tier I cities, that are leading ITES cities in India and only NCR was chosen as the region for this study. A list of 119 BPOs operating in the National capital region was compiled and it was decided to take 15% of the population for the sample. The names of the BPOs were pooled out from www.fundoodata.com and simple random sampling technique was used to select eighteen BPOs. Barclays Shared Services, Convergys India Services, EXL Service Holdings, Genpact, HCL Technologies BPO Services, IBM Daksh, Hewitt Associates, Patni, CSC BPO services, Allied Infoline Pvt. Ltd., ATS services Pvt. Ltd., Cybiz, ienergizer India Pvt. Ltd., Mynd Solutions, Quatrro BPO Solutions, Serco, Tech Mahindra Ltd. and Hughes Netfusion were the eighteen BPOs chosen for the study. Since it was not possible to build one-toone rapport with all the employees, a sample size of 500 was chosen for the study. Simple random sampling technique was used to gather data from the low and middle level employees, where attrition is highest. In a simple random sample of a given size, all such subsets of the frame are given an equal probability. This minimizes bias and simplifies analysis of results. In particular, the variance between individual results within the sample is a good indicator of variance in the overall population, which makes it relatively easy to estimate the accuracy of results. Information was gathered through questionnaire and telephonic interviews were also held in order to substantiate the data gathered. Thus, with the unconditional assistance of various associates, data was collected comfortably from the respondents working in the selected BPOs. Main data collection began in the month of July 2008 and went on till December 2008. The questionnaire was printed and 70 questionnaires were posted to each of the ten selected BPOs. Thus, the questionnaire was sent to 700 respondents, of which 521 responded, thus making the response rate to be 74.4%. 500 fully completed instruments were considered for the study. 3.5 Data Analysis Procedures Statistical Package for the Social Sciences (SPSS) version 17.0 was religiously used for the statistical analyses. Coding of variables in a quantitative research is very critical for better interpretation of results. Age, gender, education, tenure, and marital status were all coded and were Ph. D. Thesis 68
Research Methodology entered in to the computer. Gender, education and marital status had only two categories, thus they were taken as dummy variables. Gender had two categories viz. male and female and they were coded as 0 and 1 respectively. Marital status also had two categories viz. unmarried and married and they were also coded as 0 and 1 respectively. The two categories in education were graduates and post graduates. Graduates were coded as 0 and post graduates as 1. Taking such variables as dummy variables eases the statistical operations to be done on them and helps in better assessment. The questions and responses were coded and entered in the computer using Microsoft Excel Software. Required analysis was done with the aid of Statistical Package for Social Sciences 17.0 Version. Certain statistical methods were applied on the data to get the results which were analyzed. The procedures used for the analysis of the research questions of this study are described below. PART 1: Exploring employee attrition in BPOs Research Question 1. What are the factors responsible for attrition in the BPO industry? The analysis of this question was done in two parts. First was on the basis of secondary data and the second was on the basis of primary data collected. Identifying factors through secondary data was the beginning of the research study. In order to validate the results and assess them in light of primary data, the second part was carried out. For the secondary data analysis, simple weighted average method was used. The factors extracted from various studies were compared and weights were assigned to the studies to get ranks for the factors. Based on these rankings, most important factors for attrition were found out. For the primary data analysis, factor analysis was done to extract the most important factors out of 21 factors which are essentially responsible for attrition in BPO. Principal component analysis was used as the method of extraction and varimax was used as the rotation method. Eight dimensions of attrition were extracted through factor analysis. After factor analysis, means and standard deviations were calculated for each of the dimensions. On the basis of their means, attrition factors were ranked. 69 Ph. D. Thesis
Chapter 3 Research Question 2. How much do the dimensions of attrition vary across the personal characteristics of the respondents? In order to study how these dimensions of attrition vary across the personal characteristics of the respondents, t-test and Duncan s mean test were used. These tests revealed how these dimensions diverge in case of males and females; married and unmarried; graduates and undergraduates and at different age groups and tenure periods. T-test was used for gender, marital status and education, as there were only two categories in each of these and Duncan s mean test was used for age and tenure as there were three categories. Research Question 3. How much do the attrition factors contribute to employee attrition in the BPOs? Stepwise regression was carried out to study the influence of the dimensions of attrition on an employees probability of quitting the organization. Probability of quitting the organization was kept as the dependent variable and the dimensions of attrition and personal characteristics were kept as the independent factors. Before this procedure, correlations of the dimensions and the personal characteristics with the probability of quitting were computed. It was followed by regression, which gave quite meritorious results. PART II: Exploring employee retention in BPOs Research Question 4. What are the factors that lead to employee retention in the BPO industry? In order to identify the factors that lead to employee retention, factor analysis was done to extract the most important factors out of 20 factors which lead to retention in BPO. Principal component analysis was the method of extraction and varimax was the rotation method. Four factors were extracted whose eigen values were more than 1. After factor analysis, means and standard deviations were calculated in order to rank each of the dimensions. Ph. D. Thesis 70
Research Methodology Research Question 5. How much do the dimensions of retention vary across the personal characteristics of the respondents? The dimensions of retention were compared across the various personal characteristics of the respondents. Again, t-test and Duncan s mean test were done to see if there was any significant difference among the perceived values across age, gender, marital status, tenure and education. Research Question 6. How much do these dimensions of retention contribute to an employee s willingness to stay in the same organization? Correlations were used to assess the relationship of the retention factors with the perceived attitudes of the employees. The four factors of retention along with personal characteristics were put for regression with the dependent variable i.e. employees willingness to stay in the organization. Research Question 7. What is the impact of the recommendations on employees willingness to stay in the organization? For assessing the impact of recommendations on employees willingness to stay in an organization, a patch test was done. Simple means and standard deviations have been calculated to assess their willingness to stay in the organization. After this, paired sample students t-test was also done to see whether the change in willingness occurred due to the recommendations given. Research Question 8. What is the mean level of the perceived attitudes i.e. employee satisfaction, employee motivation, employee involvement and life interest and work compatibility among the BPO employees before and after the implementation of recommendations? Means, measures of central tendency and dispersion were computed to assess the level of employee satisfaction, employee motivation, employee involvement and life interest and work compatibility. Results were presented in descending order of means to observe aspects reflecting relatively greater and relatively lesser satisfaction levels. 71 Ph. D. Thesis