Interpreting a Data Base of Railway Workers using Optimal Scaling Techniques

Resumen

Optimal scaling techniques, in particular Categorical Principal Components, are usedin order to interpret the information contained in a database of railway workers. Thedata set consists of eight variables, three quantitative and five qualitative, measuredon 527 workers who joined the Spanish railway company MZA during the period1882 to 1885. The analysis revealed that workers whose place of birth was not Spaintended to be employed in more senior jobs and were paid higher salaries than workerswhose place of birth was Spain. It also revealed that most workers who left thecompany in an abnormal way (redundancy, or disciplinary dismissal) did so not longafter they had joined. It was also found that the reason for leaving was unrelated toboth first salary and seniority at the time of joining.