12 Background high skilled jobs,employers invested much more resources in selection.While they might use an application form for the initial short-listing of applicants,they would use psychometric testing or aptitude tests to confirm the conclusions of interviews.The consequence of this greater investment was that higher skill jobs attracted higher wages,lasted longer and led to more satisfaction with the person recruited.It is worth noting that the analysis by Pelizzari appears to contradict the information given by around two-thirds (69 per cent)of respondents,who had said their establishment used a standardised recruitment procedure for all vacancies. Drawing their evidence from the employee element of the 1998 Workplace Employee Relations Survey,Noon and Hoque (2001)identified less favourable treatment of ethnic minorities in the workplace in several key respects.The analysis was based on almost 24,000 employee responses4,and covered four topics that employees might have discussed in the past year with a supervisor:how they were getting on with their job,their chances of promotion,their training needs and their pay.Summarising their findings,the analysis showed: ethnic minority men were less likely than white men to have held discussions with a supervisor on their job and pay; ethnic minority women received poorer treatment than white women in all respects; workplaces with positive action measures and monitoring of equal opportunities policies had equal treatment of both men and women; unionised workplaces showed greater inequality in treatment,while there was more equal treatment in non-unionised workplaces (although ethnic minorities may benefit from collective bargaining on pay in unionised workplaces). 1.5 Field experiment approaches:actors and correspondence tests The method of 'field experiments'has been developed to estimate the prevalence of discrimination in recruitment.Initially,in both the UK and the US,this was based on actors who presented themselves as applicants for housing or job vacancies. This approach has been extensively applied in a range of settings,notably in testing for discrimination in access to housing,and similarly in access to financial arrangements and in retailing of cars.Where the negotiation of access occurs face to face,it is difficult to see an alternative.A recent discussion of this approach is provided in National Research Council (edited by Blank et al.,2008),where this is presented as the usual approach to field experiments. Riach and Rich(2002)examined the existing studies on ethnic discrimination and came to the conclusion that experiments involving personal approaches have The sample included 328 ethnic minority men and 364 ethnic minority women.12 high skilled jobs, employers invested much more resources in selection. While they might use an application form for the initial short-listing of applicants, they would use psychometric testing or aptitude tests to confirm the conclusions of interviews. The consequence of this greater investment was that higher skill jobs attracted higher wages, lasted longer and led to more satisfaction with the person recruited. It is worth noting that the analysis by Pelizzari appears to contradict the information given by around two-thirds (69 per cent) of respondents, who had said their establishment used a standardised recruitment procedure for all vacancies. Drawing their evidence from the employee element of the 1998 Workplace Employee Relations Survey, Noon and Hoque (2001) identified less favourable treatment of ethnic minorities in the workplace in several key respects. The analysis was based on almost 24,000 employee responses4 , and covered four topics that employees might have discussed in the past year with a supervisor: how they were getting on with their job, their chances of promotion, their training needs and their pay. Summarising their findings, the analysis showed: • ethnic minority men were less likely than white men to have held discussions with a supervisor on their job and pay; • ethnic minority women received poorer treatment than white women in all respects; • workplaces with positive action measures and monitoring of equal opportunities policies had equal treatment of both men and women; • unionised workplaces showed greater inequality in treatment, while there was more equal treatment in non-unionised workplaces (although ethnic minorities may benefit from collective bargaining on pay in unionised workplaces). 1.5 Field experiment approaches: actors and correspondence tests The method of ‘field experiments’ has been developed to estimate the prevalence of discrimination in recruitment. Initially, in both the UK and the US, this was based on actors who presented themselves as applicants for housing or job vacancies. This approach has been extensively applied in a range of settings, notably in testing for discrimination in access to housing, and similarly in access to financial arrangements and in retailing of cars. Where the negotiation of access occurs face to face, it is difficult to see an alternative. A recent discussion of this approach is provided in National Research Council (edited by Blank et aI., 2008), where this is presented as the usual approach to field experiments. Riach and Rich (2002) examined the existing studies on ethnic discrimination and came to the conclusion that experiments involving personal approaches have 4 The sample included 328 ethnic minority men and 364 ethnic minority women. Background