The dataset “employee satisfaction” comes from one of Visao user. It has been anonymized and provides 15 variables (on satisfaction levels, socio-demographics, organization structure) on 209 employees.
By uploading the dataset on Visao platform, the dataset is processed and a graph is created displaying the relations between the values of the 15 variables. The graph is built from the correlation between the variables. The thickness of the link between the values is proportional to the correlation.
We can 2 big clusters (low level of employee satisfaction on the top right, highlighted in red, and high level on the bottom left).
By focusing on employees with low global satisfaction, we can see it is very much correlated with all the other measures of dissatisfaction: work conditions, work relations, work content, values at work, career development,equal opportunity.
Low global satisfaction is also correlated to, but at a lesser degree (links are thinner), medium to high seniority (11 - 20, > 20), old age (50-59), working at a branch, low classification and the absence of management role.
By symmetry, employees with high global satisfaction tend also to rate favorably all the other measures of satisfaction: work conditions, work relations, work content, values at work, career development, equal opportunity.
And to a lesser degree it correlates to high classification (C3), a management role, medium seniority (5-10), and a HR function.
Let’s focus on people who want to leave the company. The graph shows that there are 3 reasons for this:
- low satisfaction, globally and on each measure of satisfaction
- young age, although this group tends to rate average their global satisfaction
- classification C2 (= middle management), which tends to be sensitive to equal opportunity and career development.
We observed that low global satisfaction was correlated with each other measures of dissatisfaction, on the basis of 3 discretisation levels (low, average, high) for each variable. This means that employees who express global insatisfaction rate each measure indifferently: work conditions, work relations, work content, values at work, career development,equal opportunity.
We may want to investigate whether by increasing the granularity of the measures we see some differentiation emerging.
By increasing from 3 to 5 levels, there is no change.
By increasing from 5 to 10, no real change neither.
We could thus continue to provide new perspectives of analysis but it would be more useful to do it with those who are involved in human ressources management. Do not hesitate to contact us if you have a comparable dataset, we will explain the various possible options for the valorization of the dataset, all the indices that we calculate with their theoretical foundations.
