For practitioners and researchers in the United States, any discussion of future research on reference reports may well be premature. The poor performance of reference reports that has been documented in the research literature, coupled with the current litigious climate prevailing in the United States, would appear to be a strong disincentive for referees to provide
reference checks. Employers will be reluctant to evaluate previous employees
if their judgements can be contested in a court of law with substantial
payments in damages at stake. Ryan and Lasek ( 1 99 1) have examined two
areas of employer liability in relation to pre-employment inquiries in the United States, namely negligent hiring and defamation. They point out that
.,..
these two legal doctrines pose a quandary for employers and may work at
cross-purposes to subvert the selection process. On the one hand, the law relating to negligent hiring emphasises the importance of securing relevant
background information prior to hiring, particularly for jobs in which the employer may be held to have a "special duty of care". On the other hand, many employers are unwilling to provide such information for fear of breaching privacy regulations and/ or incurring defamation suits. Recent privacy legislation introduced by the New Zealand government (Privacy Act,
1993) raises a similar spectre here. The special nature of reference checks (and other types of pre-employment inquiries) compounds this dilemma because, according to Ryan and Lasek, they function largely as screen-out procedures during selection. Screen-out procedures focus on negative
selection, that is, identifying reasons to disqualify an individual from
consideration for employment. Because screen-out procedures seek negative information (such as a lack of requisite knowledge, skills or abilities, or the presence of "character flaws" that may make misconduct more likely) they are clearly vulnerable to accusations of slander, libel, a,dverse impact and
breaches of privacy. Ryan and Lasek note that the situation at present is one where employers are in the position of wanting information from others who are unwilling to provide that information.
Before an organisation introduces any selection procedure a thorough job analysis is recommended, if not essential. In the United States, job analysis represents the standard used to judge the content validity of the assessment procedure and provides the basis from which criteria are developed for the purposes of assessing criterion-related validity (U.S. Equal Employment Opportunity Commission et al. , 1978, 1979). In New Zealand there are no legal requirements for organisations to conduct job analyses, but they do
constitute "best practice." If reference reports are based on a job analysis,
and there is evidence of improved validity, then there is probably no reason
why there should be any special legal impediments preventing employers from
using reference checks. Similar sentiments are expressed by Ryan and Lasek
( 199 1) in their review of pre-employment inquiries. They stress the
importance of comprehensive job analysis procedures and the linkage of knowledge, skills and abilities to sele�tion methods. Employers who adopt
such procedures should be able to reduce the likelihood of negligent hiring
suits. Job analysis is also central to the development of sound, behaviourally based performance appraisal systems. Such systems can provide concrete
evidence to substantiate negative statements made by referees and,
consequently, may be of benefit in avoiding potential claims of defamation.
Altemative methods for collecting infonnation from referees warrants
additional study. The telephone check in particular has been under- researched, yet would appear to have considerable promise. Behaviourally
oriented methods, such as situational interviewing (Latham et al. , 1 980) and
the accomplishment record (Hough, 1984) offer appropriate models on which
to base the telephone check. Using job analysis information, employers can
ask applicants to nominate referees who are able to supply examples of the applicants' achievements in critical areas. The referees can then be contacted
and the information elicited. The referee is no longer responsible for providing ratings, thus minimizing leniency errors, and at the same time guaranteeing
then be able to capitalise on the advantages of appropriate training and
selection.
The incorporation of job analyses and the utilization of behaviourally based
items, or scales, would be expected to enhance reliability and validity of
reference reports. Whether or not this is the case is an empirical question to be addressed by future research. Another avenue researchers may wish to
explore is methods for developing accountability on the part of referees. Mero
and Motowildo ( 1 995) found that participants who were made to feel
accountable by having to justify their ratings to the experimenter provided
more accurate ratings of a simulated subordinate. Requiring referees to
justify their evaluations might result in similar improvements. Simply providing feedback on the accuracy of their judgements, or even the
knowledge that such feedback will be provided, could also result in
improvements in the quality of the information supplied by referees. Clear
specifications by employers as to who are acceptable referees, and the kind of
information they will be asked to supply, may also help ensure greater
consistency across applicants.
The significance of information sources has been noted by Murphy and
Cleveland ( 1 995) . They identified a number of potential sources who could
supply performance appraisal information. Murphy and Cleveland f
categorised these sources according to their hierarchical position within the organisation, and their likely access to information about behaviours and the
relations at work. Table 2 1 is adapted from Murphy and Cleveland. It uses
the same categorisation scheme but. in this case, it is applied to likely
sources of information for reference reports. The material in Table 2 1
suggests that employers must be careful to target appropriate referees if they
wish to obtain information germane to task accomplishment and
interpersonal relations in the work environment. For example, relatives and
friends have very little access to information about an individual's work-
related task and interpersonal behaviours and results. On the other hand, subordinates, peers, and the immediate supervisor all have much greater
access to such information. The question of which source, or combination of sources, that can provide the most valid ratings merits further attention.
Table 2 1
Referees' access to information about task and interpersonal behaviours and results (Adaptedfrom Murphy & Cleveland, 1 995}
Source Task Interpersonal
Behaviours Results Behaviours Results
Relatives Rare Rare Rare Rare
Friends Rare Rare Rare Rare
Subordinates Rare Occasional Frequent Frequent
Peers Frequent Frequent Frequent Frequent
Immediate Supervisor Occasional Frequent Occasional Occasional
Upper Management Rare Occasional Rare Rare
Researchers should also consider how extraneous factors related to the
referee influence the employer's interpretation of evaluations contained in
referee have any bearing on the decisions reached by employers? In a similar vein, investigation of the criteria used by applicants when they select referees would also be worthwhile. What are the characteristics of referees that
applicants value, and why do they choose some people to be referees, and not others?
The nature of the information contributed by reference reports is also worthy of consideration. Is this information unique, or do other selection procedures with superior predictive validity (such as structured interviews, cognitive ability tests, biodata) render the reference report redundant? Multiple assessments will be required to answer this, and to establish if reference reports do have any incremental validity. Although there are many directions in which future research on reference reports can head. studies that explicitly address their deficiencies must be a priority. Only when the deficiencies are remedied can their full potential be realised.