Watson Analytics Assignment Help
The objective of this assignment is purely to develop an insight into IBM’s Watson Analytics program by using explore, predict and assemble commands for large data sets. The sample dataset being used in this assignment is “Retaining Talent and Reducing Turnover” provided on the free trial version of Watson Analytics.
Human resource is undoubtedly the most valuable and important resource to every firm whether doing business at small or large scale (Ciencia, 2010). The success or failure of the businesses depends entirely upon the quality of its employees. Today, one of the most critical issues being faced by many organizations is how to retain the talented employees within the organization. In order to retain the employees, it is important to understand the reasons for employee’s turnover and the ways they can be retained (Hancock, 2013). One of the most important aspects of employee’s retention is job satisfaction and workplace environment.
The current study will evaluate the impact of monthly income (USD), number of years spent in a current role, loyalty with company and hourly rate (USD) on job satisfaction. Apart from exploring the relationships between job satisfaction and other variables, a predictive model is also developed for salary hike (%). The objective is to analyze the motivating components at workplace that increases the job satisfaction level and allows the firms to retain the best and talented workforce. Following research questions are built in order to carry on the research:
Question 1: What are the Values of Monthly Income (USD) for Job Satisfaction?
Question 2: How do the values of Year in Current Role compare by Job Satisfaction?
Question three: What is the trend of monthly income over age by attrition?
In order to answer the above research questions, a dataset of retaining talent and reducing turnover is accessed on Watson Analytics. The tools like bar chart, dial, grid, area, network, table and tree map is used for discussing the links between selected variables. The data set contains the information of about 1470 employees along with their education, gender, marital status, job satisfaction, job level, role, monthly rates, salary hike (%) and performance ratings.
The relationship between job satisfaction and monthly salary is well established by many researchers including Mohanty (2007), Callister (2006) and Parvin (2011). Researchers have posited a causal link between job satisfaction and amount of salary received (hourly, weekly or monthly). A higher amount of salary received might motivate the workers to work more and stay more satisfied with their jobs. Mohanty (2007) identified a positive attitude in the workers with high pay levels and found them to be satisfied with their current jobs. In either case, job satisfaction is driven by increase in salary as per found in existing literature. Both of the variables are found to have general and positive link.
Upon analyzing the chosen data set on Watson Analytics, it was found that the job satisfaction was high or very high for monthly income (USD) that is above $2,864,379 (on average). While, at lower monthly income data points, the job satisfaction was found out to be either low (at $1,896,294) or medium ($1,827,652). This indicates a positive link between job satisfaction and monthly income levels. The higher the monthly salary of a worker is, the more he is satisfied with his job.
The line graph above shows clearly that for both peak points of income above $2.5M (USD), the job satisfaction level is either high or very high. The line chart is flat at $2M (USD) with low or medium job satisfaction levels.
This question explores the relationship between job experience and job satisfaction. According to Meyer et al. (2002) and Iqbal (2010), employees with long experience in current role with the same organization tends to be more liable and are found out to have higher level of job satisfaction. These employees find it difficult to shift jobs due to strong emotional attachment with current organization and high experience in current role. This shows an affective satisfaction level of employee towards its job and high level of loyalty towards the organization (Meyer et al., 2002). It is evident theoretically that employees with higher experience in same job can recognize their job in much better way as compared to the inexperienced ones. Another study by Kardam and Rangnekar (2012) recognized that the more the experience of an employee is, the more he is satisfied with the current job.
The analysis of the data set in Watson Analytics indicated the similar trend as it can be seen in the figure below that the high and very high job satisfaction is associated with higher number of years (total) spent in current role. For example, for more than 1800 years (sum) or for employee who has spent more than 1.2 years (on average i.e. 1800/1470) in the current role, the job satisfaction level is either high or very high. The findings are just in line with the previous literature found in this regard.
Attrition is a very critical issue and is a major challenge facing the industry these days. It is the main problem that is being highlighted in many organizations. Retaining employees and reducing attrition is very critical for the businesses today. According to Trevor (2001), age, attrition and income levels are associated with each other. Level of income and gender influence the employee attrition and retention levels.
In order to understand the dynamics of attrition, age and monthly income, Watson Analytics was used. The observed model showed a trend of monthly income over age by attrition for the given data set. In the line graph below, it can be identified that at the age of 40, the worker has the highest income and the lowest attrition rate. On the other hand, the employees between 18 to 26 and 56 to 60 have the highest attrition rate. The reason can be that the employees at these ages are either not satisfied by this jobs or have low pay levels. This line chart provides a very clear view of why the attrition is highest for the specific age of workers.
The current dataset allowed for developing a decision tree with discrete target variable which I chose to be “Salary Hike (%)”. The aim was to get a predictive model for salary hike within the organization on the decision tree model. Upon asking the question “What is the predictive model for Attrition?” Watson Analytics developed a decision tree along with rules.
The predictive model on Watson Analytics showed that the strongest predictor variables for salary hike were performance rating and years at company with 60% predictive strength. The single best predicter is performance rating because it has the highest percentage of salary hike i.e. 22.23%. The high salary hike % means that employees get more increase in their monthly income whose performance rating is outstanding and who have spent more than 7 years at the company. This group of employee is most likely to get salary hike than any other groups.
The decision rules for this predictive model are (see the picture as well):
- The group of people with a performance rating of outstanding and years at company of more than 7 are likely to get 22.23% salary hike.
- The group of people with a performance rating of outstanding and years at company of less than or equal to 7 are likely to get 21.61% salary hike.
- The group of people with a performance rating of excellent and years at company of less than or equal to 2 are likely to get 14.43% salary hike.
- The group of people with a performance rating of excellent and years at company of more than 2 are likely to get 13.87% salary hike.
These rules can be useful for an organization to understand the link between performance ratings and salary hikes. HR can use this information for setting a salary hike in future based on ratings of performance. Furthermore, employees can use this predictive model to understand the standard of salary hike in an organization. Employees can be motivated to improve their performance in order to get high salaries in future. The decision rule can predict the top two salary hikes rates for each group of employee in an organization. This can be used to take action in order to increase the salaries in future.
The decision tree shows that for outstanding performance, average salary hike (%) is 21.85% and with excellent performance rating, the average salary hike is 14%. For outstanding performance ratings, the salary hike is unusually high at 22.23% for workers who have spent more than or equal to 7 years in the company. For the excellent performance ratings, the lowest average salary hike was 13.87% and the highest average salary hike was 14.43% (see the decision tree below).
Callister, R.R., 2006. The impact of gender and department climate on job satisfaction and intentions to quit for faculty in science and engineering fields. Journal of Technology Transfer, 31(2), pp.367-75.
Ciencia, G., 2010. Reducing employee turnover: A Retention strategy. Applied HRM Research , 8(2), pp.63-72.
Hancock, J., 2013. Meta-analytic review of employee turnover as a predictor of firm performance. Journal of Management , 39(3), pp.573-603.
Iqbal, A., 2010. Empirical Assessment of Demographic Factors, Organizational Ranks and Organizational Commitment. International Journal of Business and Management, 5(2), pp.16-27.
Kardam, L.B. & Rangnekar, S., 2012. Job Satisfaction: Investigating the role of experience & education. Journal of Arts, Science & Commerce, 4(1), pp.16-23.
Meyer, J.P., Stanley, D.J., Herscovitch, L. & Topolnytsky, L., 2002. Affective, continuance, and normative commitment to the organization: A meta-analysis of antecedents, correlates, and consequences. Journal of Vocational Behaviour, 61(2), pp.20-52.
Mohanty, M.S., 2009. Effects of positive attitude on earnings: evidence from the US longitudinal data. The Journal of Socio-Economics, 38(2), pp.357-71.
Parvin, M.M., 2011. Factors affecting employee job satisfaction of pharmaceutical sector. Australian Journal of Business and Management Research, 1(9), pp.113-23.
Trevor, C.O., 2001. Interactions among actual ease of movement determinants and job satisfaction in the prediction of voluntary turnover. Academy of Management Journal, 44(2), pp.621-38.