Ridderprint

Chapter 6 124 technical traineeships (60.9%) lasted two to three years, commercial-oriented traineeships (18.5%) lasted between 2.5 and 4.5 years, and traineeships for corporate functions (18.0%) lasted three years. 6.4.2 Measures 6.4.2.1 Voluntary Turnover We measured voluntary turnover using the duration of employment during the observational timeframe. Turnover was recorded on a daily level in Shell’s HRIS, but on a monthly basis at Consumer Goods. During the observational timeframe at Consumer Goods, 496 trainees left voluntarily (19.6%) after a median 731 days. At Shell, 672 employees left (10.4%) after a median 850 days. We right-censored cases of involuntary turnover, including redundancies, divestments, and health-related contract endings (5.7% at Consumer Goods, 9.5% at Shell), and employees who were still employed at the end of our observational timeframes (74.6% at Consumer Goods, 80.1% at Shell), as is the convention in survival analysis (e.g., Nyberg, 2010). 6.4.2.2 Performance At both organizations, supervisors recommended performance ratings for the end- of-year ranking panels, composed of the supervisors, HR business partners, and senior line managers of a peer population. This panel calibrated the final relative performance ranking of a population, with ratings following a forced distribution. Only employees who had joined before October were considered in the year’s end calibration sessions (Consumer Goods, 2017; Shell, 2017). At Consumer Goods, performance was evaluated on a five-point scale with a forced average of three. Unfortunately, data were sparse in 2011 and 2012. Graduates hired after 2012 received their first rating after a median 214 days and these matched the forced distribution (M = 3.043, SD = 0.507) with relatively underpopulated tails (0.7%, 7.9%, 78.9%, 11.7%, and 0.9% for ratings 1 through 5). In the full dataset, employees spent a median 426 days without rating and, for these 3987 records, the rating was replaced by the modal rating of 3. At Shell, performance was evaluated on a decimal scale from 0.0 to 1.5 with a forced average of 1.03. Employees’ first performance ratings matched the forced distribution (M = 0.978, SD = 0.108), with the majority of ratings either 0.9 (49.9%) or 1.0 (25.8%). Employees spent a median of 169 days without rating and, for these 7903 records, the rating was replaced by the modal rating of 0.9. 6.4.2.3 Leadership Potential At both organizations, supervisors recommended potential ratings for their direct reports to calibration panels – composed of relevant managers, HR business partners, and senior line managers – who discussed employees alongside their peer population. The subsequent ratings were used to differentiate HRM processes, such as succession planning and learning and development (Consumer Goods, 2017; Shell, 2017).