Ridderprint

Paradoxes in global talent pipelines 123 2002). With different assignment characteristics (i.e., short duration, development orientation, lesser rewards) and employee profiles (i.e., single, young assignees), the turnover implications of STIA are unknown. Hence, we pose the following alternative hypothesis: Hypothesis 3b: Graduate trainees are more likely to turnover voluntarily as a result of STIA. 6.4 Methods 6.4.1 Sample Descriptions At two multinationals, we tracked new graduate trainees through the HRIS for specific observational timeframes. 6.4.1.1 Consumer Goods The first multinational requested to stay anonymous and will henceforth be referred to as Consumer Goods, as it operates in that sector. Consumer Goods employs over 150.000 employees in over 150 countries and has its headquarters in Western Europe. The observational timeframe lasted from January 1 st , 2011 until March 31 st , 2016 (63 months). The HRIS contained 2577 hiring records for graduate traineeships, 2538 (98.8%) of which represented unique joiners. 8 .Yearly hiring patterns were stable, except for a hiring freeze in 2016. Recruits were mostly female (59.9%) and in the age band of 18 to 29 at hiring (98.4%). Most hires occurred in China (12.7%), India (10.4%), the USA (7.9%), the UK (7.6%), and Turkey (5.1%). All traineeships at Consumer Goods lasted three years. 6.4.1.2 Shell The second multinational, Shell, is a group of companies operating in the oil and gas sector with its headquarters in the Netherlands. Shell operates in over 70 countries and employs around 92.000 employees globally (Shell, 2016). The observational timeframe at Shell lasted from January 1 st , 2010 until May 31 st , 2017 (89 months). The HRIS contained 6630 hiring records for graduate traineeships, 6475 9 (97.4%) of which represented unique joiners. 10 Yearly hiring patterns were stable. Recruits were mostly male (61.4%) and 25 years old on average (SD = 2.735). Most hires occurred in the USA (21.0%), the Netherlands (17.2%), Canada (9.5%), Malaysia (8.7%), and the UK (8.3%). Shell’s 8 Excluded from the Consumer Goods hiring data were seven records representing mergers and acquisitions, nineteen duplicate records due to split-time positions, seven duplicate records due to differing hiring dates (most recent record retained), and six non-graduate records identified by their age band and tenure. 9 The sample used in the Cox proportional hazard models is slightly lower at N = 6437 (97.1%) because seven cases had no gender information and 31 ages were missing. 10 Excluded from the Shell hiring data were fifteen records that had a missing or erroneous date of hire, 56 records representing mergers and acquisitions, 33 duplicate records due to a system error, 24 duplicate records due to differing job titles, fourteen duplicate records due to differing hiring dates (most recent record retained), and nine non-graduate hires identified by a combination of nationality and age.