hr professionals' beliefs about effective human resource practices

Correspondence between Research and Practice • 149

HR PROFESSIONALS’ BELIEFS ABOUTEFFECTIVE HUMAN RESOURCEPRACTICES: CORRESPONDENCEBETWEEN RESEARCHAND PRACTICE

Human Resource Management, Summer 2002, Vol. 41, No. 2, Pp. 149–174© 2002 Wiley Periodicals, Inc. Published online in Wiley InterScience (www.interscience.wiley.com).DOI: 10.1002/hrm.10029

Sara L. Rynes, Amy E. Colbert, and Kenneth G. Brown1

Five thousand human resource (HR) professionals were surveyed regarding the extent towhich they agreed with various HR research findings. Responses from 959 participants sug-gest that there are large discrepancies between research findings and practitioners’ beliefs insome content areas, especially selection. In particular, practitioners place far less faith inintelligence and personality tests as predictors of employee performance than HR researchwould recommend. Practitioners are somewhat more likely to agree with research findingswhen they are at higher organizational levels, have SPHR certification, and read the aca-demic literature. Suggestions are made for more effective dissemination of HR research find-ings. © 2002 Wiley Periodicals, Inc.

Accumulating evidence suggests that certainhuman resource (HR) practices are consis-tently related to organizational productivityand firm financial performance (e.g., Arthur,1994; Becker & Gerhart, 1996; Huselid, 1995;U.S. Department of Labor, 1993). For ex-ample, Welbourne and Andrews (1996) foundthat new companies that placed a high valueon HR (as assessed by content of their pro-spectuses) and that included high levels oforganizationally based pay-for-performancehad a five-year survival rate of 92% as com-pared with 34% for companies that were lowon both dimensions. Similarly, Huselid (1995)found that a one-standard-deviation increasein scores on a “high-performance HR prac-tices” scale (which included such practices asregular attitude surveying, paying for perfor-mance, formal communication programs, anduse of employment tests) was associated witha 23% increase in accounting profits and an8% increase in economic value. In addition,

Terpstra and Rozell (1997) found that com-panies whose HR professionals read the aca-demic research literature have higher finan-cial performance than those that do not.

Nevertheless, it is a well-known fact thatorganizations often fail to adopt practices thatresearch has shown to be effective (e.g., Johns,1993; Rogers, 1995). One potential reason forthis may be a lack of practitioner awarenessof research findings (Gannon, 1983). For ex-ample, a variety of business and personal fac-tors (such as increased competition, newlegislative requirements, and dual-career fami-lies) may leave HR professionals with littletime for reading. In addition, research jour-nals have become so technically complex thatthey are nearly inaccessible to individuals with-out a doctorate degree. Moreover, previousresearch suggests that many of the questionsthat academics find interesting are viewed asrather unimportant by practitioners(Campbell, Daft, & Hulin, 1982). For all these

150 • HUMAN RESOURCE MANAGEMENT, Summer 2002

reasons, managers may be largely unaware ofrecent advances in HR research.

An alternative possibility, however, is thatmanagers and professionals are actually awareof research findings, but for one reason oranother fail to implement them. For example,Pfeffer and Sutton (2000) have argued thatthe research-practice gap is primarily a “know-ing-doing” gap rather than a “knowing” gap.As with knowledge gaps, failures of implemen-tation may occur for a variety of reasons, suchas overwork, risk aversion, political consider-ations, or organizational inertia (Johns, 1993;Rogers, 1995).

Determining the extent to which gaps be-tween research and practice are the result of lackof knowing versus lack of doing requires, as a firststep, explicit investigation of what practitionersdo know. Surprisingly, there does not appear tobe much evidence concerning this question. Assuch, existing research does not allow us to pin-point where the biggest gaps currently exist be-tween research findings and practitioner beliefs.

The present research was conducted toremedy this gap. Specifically, we surveyed alarge sample of HR practitioners with respectto their beliefs regarding various research find-ings. In this way, we were able to identify thosepractices for which there is the greatest (andleast) consistency between research findingsand practitioner beliefs. In addition, we ex-amined the various ways in which HR profes-sionals obtain information about HR practices.The purpose of this latter step was to deter-mine how research findings might be moreeffectively disseminated to HR professionals,and whether some sources of information pro-vide greater research accuracy than others.Taken together, we hope these findings willprovide a base for future efforts to reduce gapsin knowledge and beliefs through improvedinformation dissemination, more useful re-search, or both.

The Study

Members of the Society for Human ResourceManagement (SHRM) comprised the samplefor this study. SHRM’s database manager se-lected a stratified random sample of five thou-sand members with the titles of HR manageror above (e.g., director, assistant or associate

director, vice president or senior vice presi-dent). Surveys were sent to 2,600 HR manag-ers; 1,200 directors, assistant directors or as-sociate directors; and 1,200 vice-presidents orassociate vice-presidents. This sampling strat-egy was designed to ensure that respondentswould be generalists rather than specialists,and that they would have significant respon-sibilities for HR policy and implementation.

To examine the extent to which the be-liefs of HR professionals are consistent withestablished research findings, a thirty-nine-item questionnaire was constructed. Surveycontent was based on five of the seven dimen-sions covered by the Human Resource Certi-fication Institute’s “Professional in HumanResources” (PHR) exam. The included dimen-sions were: Management Practices (motivation,leadership, performance management, em-ployee involvement, and HR roles); GeneralEmployment Practices (legal issues, perfor-mance appraisal and employee attitudes);Staffing (recruitment, selection, and careerplanning); HR Development (training and de-velopment, evaluation of training effective-ness), and Compensation and Benefits (jobpricing, pay structures, compensation strate-gies and effectiveness). The dimensions ofSafety and Labor Relations were not includedbecause many HR departments do not haveresponsibility for these particular functions.

The initial questions were developed byhaving each of the authors construct content-relevant research items for areas in which heor she had particular research expertise. Wewere able to generate items for most contentareas on the basis of our own familiarity withthese research areas. For the few topics onwhich we had little expertise, we examinedresearch-oriented textbooks and researchhandbooks for leads to relevant research cita-tions. In this way, we created thirty-nine ini-tial items, sampled in roughly the sameproportions as their coverage on the PHRexam. By linking our item sampling strategyto this well-established prototype of the HRbody of knowledge, we attempted to create thebest possible opportunity for HR practitionersto demonstrate their awareness of the relevantresearch literature.

Although the general content categorieswere modeled around the PHR exam, the ac-

Determining theextent to whichgaps betweenresearch andpractice are theresult of lack ofknowing versuslack of doingrequires, as afirst step, explicitinvestigation ofwhatpractitioners doknow.


tual nature of the questions was quite differ-ent. Specifically, in contrast to the certifica-tion exam (which tends to focus heavily onlegal, definitional, and procedural issues), thepresent survey focused on research findingsregarding the effectiveness of particular prac-tices. Items were constructed to be either trueor false, based on previous research results.Respondents were asked to indicate whetherthey agreed, disagreed, or were uncertainabout each item, allowing us to determinecontent areas where practitioner beliefs di-verge most sharply from research findings.

The original questionnaire was pretestedon a sample of fifty-nine highly prolific re-searchers in HR and industrial/organizationalpsychology. On the basis of these researchers’responses and feedback, problematic itemswere either reworded or replaced.

In addition, we collected informationabout what types of reading HR professionalsdo, where they go to get help with HR prob-lems or issues, and attitudes toward varioussources of HR information. Lists of commoninformational sources were provided, and re-spondents indicated the frequency with whichthese sources were used. Both archival (e.g.,Web sites, journals) and social (e.g., consult-ants, academics, other HR practitioners) in-formation sources were assessed.

The purpose of collecting this informationwas twofold. First, knowing where HR pro-fessionals go for information is useful for fu-ture attempts to disseminate research findingsand other types of information. Second, by

comparing methods of information searchwith practitioners’ knowledge of the researchliterature, we can assess whether some sourcesof information appear to be more effective atdisseminating research findings than others.

Results: What HR Professionals Believe

Responses were received from 959 partici-pants, for a response rate of at least 19.2%(we do not know how many surveys were un-deliverable). Nearly half the respondents(48.5%) were HR managers, while the restwere either directors (26.1%), vice presidents(18.0%), or other titles (7.4%). Perhaps notsurprisingly (given our sampling strategy), theaverage respondent also had considerable ex-perience in HR (13.8 years, SD = 7.9). Test-retest reliability for the questionnaire was as-sessed using an independent sample of forty-eight mid-level general and HR managers en-rolled in an Executive MBA OrganizationalBehavior class. Over a six-week time span, test-retest reliability was found to be .70.

Between the time of survey administra-tion and preparation of this article, four of theoriginal thirty-nine items were eliminated fromthe survey.2 Of the remaining thirty-five items,the average respondent answered twenty(57%) of the items correctly (Figure 1). How-ever, there was great variability in the extentof agreement, with one subject agreeing ononly nine of the items (26%) and two agree-ing on thirty items (86%). In addition, therewas enormous variation in the extent to which

Figure 1. Histogram of knowledge scores.


respondents (as a group) agreed with particu-lar items, ranging from a low of 16% for oneitem to a high of 96% for another (Table I).

The first column of Table I shows thethirty-five retained items. Column 2 indicateswhether each item is true or false, the per-centage of respondents who agreed with theitem, and the percentage of respondents who

were uncertain (the percentage who disagreedcan be calculated by subtracting the sum ofthese two numbers from one hundred). Thethird column summarizes the nature of theresearch findings that support each item, alongwith sample research citations. (Additionalcitation information for each item can befound in the Appendix). Finally, the fourth

TABLE I Items, Responses, and Supporting Evidence

Management Practices1. Leadership training isineffective because goodleaders are born, notmade.

2. The most importantrequirement for aneffective leader is tohave an outgoing,enthusiastic personality.

3. Once employees havemastered a task, theyperform better whenthey are told to “do theirbest” than when they aregiven specific, difficultperformance goals.4. Companies withvision statementsperform better thanthose without them.

Items Answer–% Correct(% uncertain) Research Evidence Possible Contingencies

False96%(2%)

False82%

(4.5%)

False82%(6%)

True62%

(15%)

Field study evidence that leadershipbehaviors and effectiveness increasefollowing training (Barling et al., 1996).Evidence that leadership behaviors areonly weakly predicted by dispositionalcharacteristics (Judge & Bono, 2000) thatare heritable (Loehlin et al., 1998;Reimann et al., 1997).This kind of personality is, on average, anasset for leadership. A recent meta-analysisestimates a corrected validity coefficient of.31 between extraversion and leader effective-ness (Judge et al., in press). However,intelligence has an even higher correlation(.52; Lord et al., 1986). Also, some highlyeffective leaders are distinctly introverted(Bennis & Nanus, 1997; Collins, 2001).

Employees reach higher levels of perfor-mance when they are given difficult-yet-attainable goals rather than told to do theirbest. This is one of the most robustfindings in all of industrial/organizationalpsychology (Locke & Latham, 1990).

Baum et al. (1998) found that growth-oriented visions and strong vision commu-nication produced significantly higher firmgrowth rates. Hoch et al. (1999) foundthat 93% of the most successful softwarefirms had clear and ambitious visions, ascompared with 25% of the least successful.The importance of having a product visionfor successful product development hasalso been reviewed by Brown & Eisenhardt(1995).

Although it is unlikely that anyfuture variables will producehigher validities than intelli-gence, work on leaderbehaviors (such as vision-setting and communicationskills) is just beginning, solittle is known about theiraverage effect sizes. We doknow (e.g., Barling et al.,1996; Baum et al., 1998) thatleadership success can beaffected by leaders’ behaviorsand not just their traits, sofuture research would beuseful.

(continued on next page.)


5. Companies with verylow rates of professionalturnover are lessprofitable than thosewith moderate turnoverrates

6. If a company feels itmust downsize employ-ees, the most profitableway to do it is throughtargeted cuts rather thanattrition.

7. In order to be evalu-ated favorably by linemanagers, the mostimportant competencyfor HR managers is theability to manage change.

8. On average, encourag-ing employees toparticipate in decisionmaking is more effectivefor improving organiza-tional performance thansetting performancegoals.

General EmploymentPractices9. Most managers giveemployees lowerperformance appraisalsthan they objectivelydeserve.

False62%

(23%)

True54%

(17%)

True50%

(12%)

False18%(9%)

False94%(3%)

Bain & Company (Reichheld, 1996) hasanalyzed the economics of professionalturnover in several industries. For example,in brokerage firms, they calculated that anincrease in broker retention rates from 80%to 90% results in an increase of 155% inthe average net present value of a newbroker. A general model of the “economicsof employee loyalty” is also presented.

Morris, Cascio, & Young (1999) trackedchanges in employment of the S&P 500companies over a 12-year period. Theyfound that companies whose downsizingwas associated with sale of assets (i.e.,strategic or targeted downsizing) hadimproved their return on assets by thesecond year after downsizing. In contrast,pure “employment downsizers” still hadlower ROAs two years later than they hadbefore downsizing. See also Cameron et al.(1993).This was found by Ulrich et al. (1995),based on more than 12,000 peer andsupervisory assessments of the performanceof nearly 2,000 HR professionals. Ability tomanage change explained 41.2% of thevariance in supervisors’ and peers’ evalua-tions of HR professionals, as comparedwith 23.3% for HR knowledge and deliveryand 18.8% for knowledge of the business.Meta-analytic evidence that the effects ofparticipation are weaker (< 1%) than theeffects of goal setting (20%; Locke et al.,1980). These effects also hold at multiplelevels of analysis. Further evidence showsthat the effects of goal setting are robust(Locke & Latham, 1990), while the effectsof participation are highly variable (e.g.,Wagner, 1994).

Appraisal leniency is much more commonthan stringency (Jawahar & Williams,1997; Longenecker et al., 1987)

This finding might not holdtrue in certain types ofinternal labor market contexts—e.g., organizations withstrong employment “guaran-tees” (e.g., governmentemployment), or organizationswith very strong contingencyreward systems or “up-or-out”promotion & partnershipsystems. More contingency-based research would beuseful.

TABLE I (continued.)




10. Poor performers aregenerally more realisticabout their performancethan good performersare.11. Teams with membersfrom different functionalareas are likely to reachbetter solutions tocomplex problems thanteams from a single area.

12. Despite the popular-ity of drug testing, thereis no clear evidence thatapplicants who scorepositive on drug tests areany less reliable orproductive employees

13. Most people over-evaluate how well theyperform on the job.

14. Most errors inperformance appraisalscan be eliminated byproviding training thatdescribes the kinds oferrors managers tend tomake and suggestingways to avoid them.

Training & EmployeeDevelopment15. Lecture-basedtraining is generallysuperior to other formsof training delivery.

False88%(3%)

True88%(5%)

False57%

(21%)

True54%(4%)

False25%(5%)

False96%(2%)

Primary study evidence that poor perform-ers are less accurate about their relativeperformance than are good performers(Kruger & Dunning, 1999).

Multiple studies on cross-functional teamshave shown positive outcomes for productand project quality (e.g., Keller, 2001;Lutz, 1994; Northcraft et al., 1995; Pelledet al., 1999). At the top management level,Hambrick et al. (1996) found that moreheterogeneous teams (with respect tofunctional area, education, and tenure)made bolder (although slower) competitivemoves, causing an overall net positiveeffect on firm market share and profits.Norman et al. (1990) drug tested more than4,000 applicants and then followed themthrough more than 1 year of employment.Those who tested positive had a 59% higherabsenteeism rate and a 47% higher involun-tary turnover rate. Parish (1989) foundsignificant results for disciplinary actions,and Winkler & Sheridan(1989) for vehicularaccidents, absenteeism, and medical costs.See also McDaniel et al. (1988).Meta-analytic and primary study evidencethat self-ratings have higher means thanpeer and supervisor ratings (e.g., Brown,1986; Harris & Schaubroeck, 1988; Mabe& West, 1982; Thornton, 1980).Most performance appraisal errors areextremely resistant to change (Latham &Wexley, 1994). In addition, training toeliminate one kind of error often intro-duces other types of errors (e.g., Bernardin& Pence, 1980). Many managers are wellaware that they commit errors, butcontinue to do so for personal and socialreasons (Longenecker et al., 1987).

Meta-analytic evidence that computer-based instruction is slightly more effectivethan traditional instruction (Kulik & Kulik,1991). Indirect evidence is provided byeducational theory suggesting that themedium for instruction is less critical thanthe events of instruction in determininglearning outcomes (e.g., Clark, 1983). Inthis regard, lecture-based training is apassive technique, while active techniquesare typically more effective (Gagne &Medsker, 1996).



Although product outcomesare generally positive, psycho-social outcomes such asmember satisfaction or stressare frequently negative. Also,it is possible that at very highlevels of multiple types ofdiversity, interpersonal conflictmight overtake the perfor-mance benefits of diversity.



16. Older adults learnmore from training thanyounger adults.

17. The most importantdeterminant of howmuch training employeesactually use on their jobsis how much theylearned during training.

18. Training for simpleskills will be moreeffective if it is presentedin one concentratedsession than if it ispresented in severalsessions over time.

Staffing19. The most validemployment interviewsare designed aroundeach candidate’s uniquebackground.20. Although people usemany different terms todescribe personalities,there are really only fourbasic dimensions ofpersonality, as capturedby the Myers-BriggsType Indicator (MBTI).21. On average, appli-cants who answer jobadvertisements are likelyto have higher turnoverthan those referred byother employees.

22. Being very intelligentis actually a disadvantagefor performing well on alow-skilled job.

Meta-analytic evidence that age is nega-tively associated with learning outcomes(Colquitt et al., 2000). Large sampleevidence that age is associated withdecreases in cognitive capacities associatedwith learning (Horn & Cattell, 1967).Primary study evidence of negativerelationship with learning in organizationalsetting (Warr & Bunce, 1995).Meta-analytic evidence that post-trainingknowledge has a smaller relationship withtransfer than some individual differencesand contextual variables (Colquitt et al.,2000). Primary study evidence thattransfer of training climate has a strongerrelationship with transfer than learningdoes, in an organizational setting (Traceyet al., 1995).Meta-analytic evidence that training forsimple skills is more effective when spacedover time than massed in one session(Donovan & Radosevich, 1999; Lee &Genovese, 1988).

Meta-analytic evidence that structuredinterviews (where all candidates receive thesame questions) have higher validities thanunstructured ones (Schmidt & Hunter,1998; Wiesner & Cronshaw, 1988).There are five basic dimensions ofpersonality—the “Big Five” of Conscien-tiousness, Extraversion, Openness toExperience, Agreeableness, and EmotionalStability/Neuroticism (Digman, 1990).Except for Extraversion, these are not thetraits assessed by MBTI.

Meta-analytic (e.g., Conard & Ashworth,1986) and primary study evidence (e.g.,Decker & Cornelius, 1979; see Rynes,1991, for a review).

The validity coefficient for intelligence isalways positive. Hunter (1986) and Schmidt& Hunter (1998) estimate the correctedvalidity coefficient for unskilled jobs to be .23.(Comparable figures are .40 for semi-skilled,and .58 for professional-managerial jobs.)



False68%

(18%)

False60%

(11%)

False59%

(11%)

False70%(6%)

False49%

(23%)

True49%

(13%)

False42%

(12%)

Recent evidence on the effectsof recruitment sources hasbeen less consistent thanearlier evidence. The emer-gence of the Web as a majorrecruitment source may alsochange future recruitingsource dynamics.



The validity of personality measures aspredictors of performance will depend to alarge extent on the degree to which theytap important “Big Five” personalitydimensions, particularly Conscientious-ness. So, for example, Big Five personalitymeasures are relatively good predictors ofperformance, while the MBTI doesn't evenclaim to predict performance (Gardner &Martinko, 1996).Even if applicants or employees do distorttheir answers, validity of these instrumentsis still substantial. A large meta-analysissuggested that the overall corrected validitycoefficient for integrity tests (across alltypes of performance measures) is .41.Counterproductive behaviors such as theftor absenteeism are somewhat betterpredicted (.47) than overall job perfor-mance (.34).A recent study based on four large-sampledatabases showed “trivial” differences (lessthan .15 standard deviations in all cases)across Caucasians, Asians, Native Ameri-cans, and African Americans (Ones &Viswesvaran, 1998).There is considerable meta-analyticevidence that intelligence is a betterpredictor of job performance than consci-entiousness (Schmidt & Hunter, 1998;average validity coefficient for intelligence= .51 vs. .31 for conscientiousness).Meta-analytic evidence that intelligence isthe best established predictor of jobperformance (Schmidt & Hunter, 1998).Primary study evidence that the effect ofvalues on job performance is dependent onmany factors (Adkins et al., 1996); valuesinfluence perceptual organization andguide decision making, but their link to jobperformance is unclear (Ravlin & Meglino,1987).

Both laboratory and organizational fieldresearch shows that providing procedurallyjust explanations of pay cuts can dramati-cally reduce the negative side effects(Greenberg, 1990, 1993).



23. There is very littledifference amongpersonality inventoriesin terms of how wellthey predict anapplicant’s likely jobperformance.

24. Although there are“integrity tests” that tryto predict whethersomeone will steal, beabsent, or otherwisetake advantage of anemployer, they don’twork well in practicebecause so many peoplelie on them.25. One problem withusing integrity tests isthat they have highdegrees of adverseimpact on racialminorities.26. On average,conscientiousness is abetter predictor of jobperformance than isintelligence.

27. Companies thatscreen job applicantsfor values have higherperformance than thosethat screen for intelli-gence.

Compensation &Benefits28. When pay must bereduced or frozen, thereis little a company cando or say to reduceemployee dissatisfac-tion and dysfunctionalbehaviors.

False42%

(30%)

False32%

(34%)

False31%

(50%)

False18%

(10%)

False16%

(27%)

False72%

(13%)

Overall findings unlikely tochange. However, additionalstudies are needed at theorganizational level of analysis.In addition, more “values”studies are needed thatmeasure performance as anoutcome (most values researchhas examined affectiveoutcomes such as satisfaction,tenure, and commitment).



Multiple studies have demonstrated thisresult (e.g., BNA, 1988; Cable & Judge,1994). For example, Bretz & Judge (1994)found that of seven organizational charac-teristics, the one that best predictedsimulated organizational choice was pay forindividual (versus team-based) productivity.Positive relationships have been shownbetween merit systems and organization-level performance by Kopelman &Reinharth (1982) and Kopelman, Rovenpor,& Cayer (1991). Heneman (1992) reviewedfive studies establishing a positive meritsystem-performance link. Even the majorempirical study to raise “cautions” aboutmerit pay (Pearce et al., 1985) foundincreases in performance after merit payimplementation; the increases simply failedto reach statistical significance (with a verysmall sample size).Gerhart & Milkovich (1990) found thatcompanies with 80% managerial eligibilityfor stock options had 25% higher return onassets than companies where only 20% ofmanagers were eligible. (See alsoWelbourne & Andrews, 1996).New companies that placed a high valueon their employees (as coded fromprospectuses) and that included high levelsof organizational-performance-based payhad dramatically higher five-year survivalrates (92%) than those who were low onboth dimensions (34%: Welbourne &Andrews, 1996).In a field study of nine different sites,Prince & Lawler (1986) found that salarydiscussions had positive rather thannegative effects on employee attitudes andsubsequent performance improvement. Inaddition, the positive effects were strongestfor those with lower initial performanceand where initial perceptions of perfor-mance were most discrepant betweensupervisors and employees. For similarresults based on employee surveys atGeneral Electric, see Welch (2001).



29. Most employeesprefer to be paid on thebasis of individualperformance rather thanon team or organizationalperformance.

30. Merit pay systemscause so many problemsthat companies withoutthem tend to have higherperformance thancompanies with them.

31. There is a positiverelationship between theproportion of managersreceiving organizationallybased pay incentives andcompany profitability.32. New companies havea better chance ofsurviving if all employeesreceive incentives basedon organization-wideperformance.

33. Talking about salaryissues during perfor-mance appraisals tends tohurt morale and futureperformance.

True81%(8%)

False66%(7%)

True62%

(23%)

True59%

(17%)

False51%

(10%)

Could use more research onthis, but the limited researchavailable suggests that neithermorale nor performancesuffers from linking paydiscussions with performancediscussions. Even if short-termmorale were to be affected,one could not infer that eitherlonger-term morale orperformance would benegatively affected. Forexample, equity theorypredicts the effect can easilygo the other way—e.g., peopleworking harder to reduce theirdissatisfaction with theprevious appraisal.



column suggests limitations of the currentresearch in some areas, as well as possiblecontingency variables that might be revealedin future research.

With respect to items in the ManagementPractices domain, a clear majority of practi-tioners tended to agree with research findingsthat leadership abilities can be improvedthrough training (96%); that having an outgo-ing personality is not the most important char-acteristic of a leader (82%, although it certainlyhelps; see Judge, Bono, Ilies, & Werner, inpress), and that performance is generallyhigher when people are given specific, diffi-cult goals rather than told merely to do theirbest (82%). Another group of items garneredonly slim majority support: the organizational

benefits of having a leadership vision (62%),the benefits of very low professional turnover(62%), and the general superiority of targetedrather than attrition-based downsizing (54%).

Interestingly, only 50% of respondentsagreed with Ulrich, Brockbank, Yeung, andLake’s (1995) finding that the most importantcompetency for HR managers (in terms ofsupervisory and peer evaluations) is the abil-ity to manage change. This would seem to bea particularly important finding, in that it sug-gests that modern HR managers need to knowfar more than the “traditional” HR knowledgecovered in HR textbooks and certification ex-ams. Rather, they also need both cognitive andpractical knowledge regarding how to getthings done in complex social systems. Finally,



34. Most employeesprefer variable paysystems (e.g., incentiveschemes, gain sharing,stock options) to fixedpay systems.

35. Surveys that directlyask employees howimportant pay is to themare likely to overesti-mate pay’s true impor-tance in actual deci-sions.

A national survey showed that 63% ofworkers surveyed prefer straight salary,followed by individual incentives (22%) andcompany-wide incentives (12%; BNA,1988). Cable & Judge (1994) found asimilar preference for fixed pay among job-seeking college students. Also, theories ofrisk and agency theory have as a coreassumption that employees require acompensating risk differential in order tomake variable pay acceptable to them (e.g.,Jensen & Meckling, 1976).Probably due to social desirability and/orlack of self-insight, people tend to say payis less important to them than the weightsthey actually place on pay in making choicedecisions (Feldman & Arnold, 1978; Ryneset al., 1983). People also think that otherswho are “just like them” place a higherimportance on pay than they themselves do(Jurgensen, 1978)—further evidence ofpossible lack of self-insight about motiva-tions. These results are also consistent withbroader findings from the decision sciencesthat people tend to underestimate theimportance of the most important factorsin their decisions (Slovic & Lichtenstein,1971).

False40%

(12%)

False35%

(10%)

Note: Full citations for sources not listed in the references section of this article can be found in the Appendix, where they are listed by item number.


only 18% believed that goal setting was moreeffective than employee participation for im-proving organizational performance, althoughthis finding has been supported in multiplestudies over the past twenty years.

With respect to items falling under thedomain of General Employment Practices,there was very high agreement with three ofthe items. Specifically, most respondents(94%) agreed that managers do not give overlystringent performance appraisals, that poorperformers are generally less realistic abouttheir performance than higher performers(88%), and that multifunctional teams tendto reach better solutions to complex problemsthan single-function teams (88%). However,there was considerably less awareness thatdrug testing has been empirically shown todifferentiate applicants in terms of post-hireperformance (57%), and that most peopleoverestimate how well they perform on thejob (54%).

Finally, respondents were much more op-timistic about the possibility of eliminatingerrors from the performance appraisal processthan research findings warrant. Specifically,only 25% disagreed with the questionnaireitem, “Most errors in performance appraisalscan be eliminated by providing training thatdescribes the kinds of errors managers tendto make and suggesting ways to avoid them.”In reality, however, a long line of researchshows that attaining accuracy in performanceappraisals is extremely difficult, and cannotbe achieved by these simple training methods(Latham & Latham, 2000; Latham & Wexley,1980). Moreover, many managers are fullyaware that they are committing certain kindsof errors (especially leniency), but they do soanyway for social, political, or motivationalreasons (Longenecker, Gioia, & Sims, 1987).

With respect to Training and EmployeeDevelopment, nearly all respondents (96%)recognized that pure lecturing is generally notthe best way to deliver training. In addition,68% accurately believed that older adults donot learn more from training than youngeradults, while 60% realized that how much islearned during training is not the most impor-tant determinant of how much training is ac-tually used on the job. Finally, 59%(accurately) disagreed with the statement that

training for simple skills is more effective whenpresented in a single concentrated session.

On the other hand, in the Staffing area,practitioner beliefs were notably inconsistentwith research findings. In fact, on only oneitem was there substantial agreement betweenresearch findings and practitioner beliefs. Spe-cifically, 70% (accurately) disagreed that themost valid employment interviews are de-signed around each candidate’s unique back-ground. Rather, considerable evidencesuggests that the most valid interviews are notunique, but rather structured to be the sameacross all candidates.

On all other selection-related items, fewerthan 50% of respondents agreed with prevail-ing research findings. One particularly notabledifference between practitioners’ beliefs andresearch findings concerned the perceivedusefulness of intelligence (or general mentalability, GMA) as a predictor of performance.Generally speaking, research findings suggestthat GMA is a much better predictor of sub-sequent job performance than mostpractitioners believe. In the words of Schmidtand Hunter (1998):

Research evidence for the validity of GMAmeasures for predicting job performance isstronger than that for any othermethod…literally thousands of studies havebeen conducted over the last ninedecades….Because of its special status,GMA can be considered the primary per-sonnel measure for hiring decisions, andone can consider the remaining personnelmeasures as supplements to GMA. (pp.264, 266)

In contrast to this strong research-basedendorsement of GMA as a selection device,our respondents revealed considerable skep-ticism about the value of intelligence via theirresponses to at least three items. Specifically,57% falsely believed that companies thatscreen job applicants for values have higherperformance than those that screen for intel-ligence (and 27% were uncertain), while 72%falsely believed that conscientiousness is abetter predictor than intelligence (10% uncer-tain). In addition, 46% incorrectly believedthat intelligence is actually a disadvantage for

One particularlynotabledifferencebetweenpractitioners’beliefs andresearch findingsconcerned theperceivedusefulness ofintelligence (orgeneral mentalability, GMA) asa predictor ofperformance.Generallyspeaking,research findingssuggest that GMAis a much betterpredictor ofsubsequent jobperformancethan mostpractitionersbelieve.


performing well at low-skilled work (12% wereuncertain).

Contrary to practitioners’ beliefs, researchfindings suggest that GMA has an average va-lidity coefficient of .51, as compared with .31for conscientiousness (Schmidt & Hunter,1998). Similarly, although research connect-ing applicant or employee values to job per-formance is far less plentiful than researchon conscientiousness, existing evidence sug-gests that relationships between values andperformance are in part determined by fit withother employees, and that they can even beinconsistent in direction (Adkins, Ravlin, &Meglino, 1996; Lauver & Kristof-Brown, inpress). Intelligence, on the other hand, hasbeen shown to be a positive predictor of per-formance at all levels of job skill, although thesize of the relationship is smaller for lower-level jobs than for jobs of greater complexity(e.g., � = .23 for unskilled; � = .40 for semi-skilled, and � = .58 for managerial and pro-fessional jobs; see Schmidt & Hunter, 1998,p. 264).

Our results also suggest considerable dis-crepancies between research and practitionerbeliefs with respect to both the nature andusefulness of personality traits as a basis forselection. These discrepancies are evident infour items. Specifically, only 49% of respon-dents (correctly) disagreed with the statementthat the four dimensions represented by theMyers–Briggs Type Indicator (MBTI) repre-sent the basic dimensions of personality; only42% correctly disagreed with the statementthat there is very little difference among per-sonality inventories with respect to how wellthey predict job performance; only 32% (cor-rectly) disagreed with the statement that in-tegrity tests are not useful for selection, andonly 31% correctly disagreed with the state-ment that integrity tests result in high degreesof adverse impact against minorities.

In contrast to the typical HR manager’sbeliefs, research suggests that there are five(rather than four) basic dimensions of person-ality, and that they are not the ones assessedby the MBTI (see Table I). In addition, differ-ent personality measures can be expected tohave very different levels of usefulness for pre-dicting performance because different mea-sures contain varying proportions of the five

basic dimensions, some of which are muchmore predictive than others (e.g., Barrick &Mount, 1991). Third, integrity tests exhibitrelatively high validity (� = .41) as selectiondevices, even if some applicants do tend todistort their responses in a favorable direction.3

Finally, research based on several very largesamples suggests that differences among ra-cial or ethnic groups on integrity tests are“trivial” and, as such, unlikely to lead to ad-verse impact (Ones & Viswesvaran, 1998).

It should be noted about personality that,in contrast to most other areas of the ques-tionnaire, relatively high proportions of re-spondents were uncertain about all of thepersonality items. For example, 23% were un-certain about whether the four dimensions ofthe MBTI represent the basic dimensions ofpersonality; 30% were uncertain aboutwhether there are sizable differences in thevalidity of different personality inventories;34% were uncertain about whether integritytests have selection validity; and a very con-siderable 50% were uncertain about whetherintegrity tests have adverse impact againstminorities. To some extent, the large degreeof uncertainty regarding personality tests isunderstandable, given that supportive resultshave emerged only in the past decade afterseveral decades of discouraging (pre-meta-analytic and pre-Big Five) results. Althoughevidence concerning the usefulness of severalof the Big Five personality traits has begun tobe translated to practitioner audiences, mostsuch translations are still quite recent (e.g.,Barrick & Mount, 2000; Behling, 1998).

Although discrepancies between researchand practitioner beliefs are quite large in se-lection, they appear to be somewhat smallerwith respect to compensation and benefits. Ingeneral, respondents tended to agree with re-search findings: that there is indeed somethingmanagers can do to soften the blow of pay cuts(72% agreement); that most people wish to bepaid on the basis of individual rather thangroup or team performance (81% agreement);that merit pay systems are not detrimental toorganizational performance (66% agreement);and that having higher proportions of manag-ers on organizationally based pay incentivesis associated with higher organizational per-formance (62% agreement). In addition, a slim

Our results alsosuggestconsiderablediscrepanciesbetween researchand practitionerbeliefs withrespect to boththe nature andusefulness ofpersonality traitsas a basis forselection.


majority (59%) agreed that new companieshave a better chance of surviving if all em-ployees receive incentives based on organiza-tion-wide performance, and that talking aboutsalary issues during performance appraisalsdoes not destroy employee morale and perfor-mance (51%).

On the other hand, in contrast with exist-ing research (Cable & Judge, 1994), only 40%of respondents disagreed with the (false) state-ment that most employees prefer variable paysystems (e.g., incentives, gain sharing, stockoptions) over fixed pay systems. This resultmay have been obtained because this surveywas conducted just before the stock marketcrash in the spring of 2000. As such, respon-dents may have still been reading popular pressarticles about how top-tier business studentswere eschewing “corporate” jobs (or even com-pleting their MBAs) in favor of cashing in onthe dot.com and IPO bonanzas.

Finally, only 35% agreed with the find-ing that people are likely to report that pay isless important to them than it really is. Thisappears to occur for one of two reasons: it issomewhat “socially undesirable” to admit tostrong preferences for pay (e.g., Feldman &Arnold, 1978; Rynes, Schwab, & Aldag,1983), and the more general tendency to un-derreport the importance of the most signifi-cant decision variables, while overreporting

less significant ones (Slovic & Lichtenstein,1971). Failure to recognize these general bi-ases in employees’ response patterns mightwell cause employers to underestimate theextent to which they need to attend to com-pensation issues.

Where HR Professionals Get TheirInformation

In addition to asking about respondents’ be-liefs, we also inquired as to where respondentsget their information, as well as their attitudestoward different informational sources. TableII shows the extent to which our respondentsread various types of periodicals that mightyield information about the effectiveness ofHR practices.

As the table indicates, the only periodicalthat respondents read more than “sometimes”is HR Magazine, which the typical respondent“usually” reads. Three periodicals were alsoread somewhere between “rarely” and “some-times”: The Wall Street Journal, HR Focus, andHR Executive. Consistent with other research(e.g., Offermann & Spiros, 2001), very fewrespondents (fewer than 1%) reported thatthey “usually” read the three most research-oriented journals on the list—Journal of Ap-plied Psychology, Personnel Psychology, andAcademy of Management Journal.

Periodical

Human Resource MagazineWall Street JournalHR FocusHuman Resource ExecutiveHuman Resource Management JournalWorkforceBusiness WeekFortuneForbesHarvard Business ReviewHuman Resource Planning JournalInc.Fast CompanyPersonnel PsychologyJournal of Applied PsychologyAcademy of Management ExecutiveAcademy of Management Journal

MeanRating

4.282.722.692.622.232.001.941.921.771.771.511.491.461.221.191.111.11

Standard Deviation

.861.191.381.481.341.311.001.00 .931.02 .93 .80 .95 .60 .55 .42 .39

TABLE II Frequency with Which Periodicals Are Read

Note. 1–5 scale, where 1 = “never,” 2 = “rarely,” 3 = “sometimes,” 4 = “usually,” and 5 = “always.”


Table III provides additional informationabout how practitioners seek or receive infor-mation about HR issues. This table shows thatthe most common sources to which practitio-ners turn when faced with HR problems areother HR professionals in the same organiza-tion, the SHRM Web site, and other Internetsites. Interestingly, despite the very low ratesof academic journal reading reported in TableII, practitioners still report going to the re-search literature more often than to consult-ants, academics, or HR professionals in otherorganizations. Although we cannot be sure ofthe reason for this apparent inconsistency, onepossibility is that social desirability consider-ations cause people to inflate their answers togeneral questions about how much they readthe research literature, as opposed to morespecific questions about how often they readparticular journals. Another possibility is thatpractitioners include a wider range of sources

under the category of “research literature”than we did in our survey.

Finally, Table IV shows respondents’ atti-tudes toward academics and academic re-search. This table suggests that practitionersare close to neutral on questions about theusefulness or applicability of research findings,but that they nevertheless wish they had moretime to read about them (3.9 on a five-pointscale). This was particularly true of those withless experience in HR; there was a negativecorrelation of –.21 (p < .001) between experi-ence and desire to learn about academic re-search.

What Characteristics Are Associated withResearch Knowledge?

As indicated earlier, there was very high vari-ability in the extent to which respondents’beliefs were consistent with research findings.

ResourceOther HR professionals in my organizationSHRM Web siteOther Web sitesHR research literatureHR professionals in other organizationsConsultantsAcademics

MeanRating3.152.982.742.622.481.981.42

StandardDeviation1.381.041.101.10 .92 .89 .73

Note. 1–5 scale, where 1 = “rarely or never,” 2 = “a few times per year,” 3 = “about once a month,” 4 = “several times permonth,” and 5 = “almost daily.”

AttitudeI wish I had more time to read aboutacademic HR research findings.Most research findings make sense intheory, but don’t work well in practice.I would like to spend more timetalking with academics about HRproblemsI generally don’t find academic HRresearch to be very useful.I often wish I could call an academicto help me solve HR problems

TABLE III Sources of Help for Solving HR Problems

TABLE IV Practitioner Attitudes toward Academics and HR Research

Note. 1–5 scale, where 1 = “rarely or never,” 2 = “a few times a year,” 3 = “about once a month,” 4 = “several times permonth,” and 5 = “almost daily.”

Mean Rating3.91

3.04

2.86

2.78

2.65

Standard Deviation1.03

.84

1.09

.91

1.07


This raises the question of whether differ-ences in research knowledge are reliably as-sociated with differences in individual char-acteristics (e.g., education) or information-seeking strategies (e.g., academic reading).

To assess this question, we constructeda correlation matrix of major demographiccharacteristics, information-seeking strate-gies, and number of correct responses (TableV). As Table V indicates, two individual dif-ference characteristics—job level (r = .14,p < .01) and SPHR certification (r = .10,p < .01)—were associated with higher lev-els of agreement with research findings. Twoinformation-seeking strategies were also sig-nificantly correlated with research knowl-edge: academic reading (r = .08, p < .05)4

and use of consultants to seek information(r = .09, p < .05).

Finally, we conducted a regression analy-sis of all the independent variables from TableV to determine the extent to which bivariateresults held up after controlling for the ef-fects of all other variables. This analysisshowed that three of the variables remainedsignificant in this multivariate analysis: joblevel (b = .50, t = 3.14, p < .01), SPHR certi-

fication (b = .81; t = 2.56, p = .01), and aca-demic reading (b = .60, t = 1.94, p = .05).However, the overall predictability of themodel was quite modest (R = .20).

Discussion

In a book called The Knowing-Doing Gap, twoprominent academics recently argued that themajor differences in organizational competitive-ness were no longer created by differences inknowledge, but rather by differences in theability to implement what is known (Pfeffer &Sutton, 2000). Their basic assertion is that in-formation dissemination has become so effi-cient that everyone is likely to know about bestpractices; thus, best practices are no longerlikely to be sources of competitive advantage:

We now live in a world where knowledgetransfer and information exchange are tre-mendously efficient, and where there are nu-merous organizations in the business of col-lecting and transferring best practices. So,there are fewer and smaller differences inwhat firms know than in their ability to acton that knowledge. (p. 243)

Notes: N = 827. * p < .05 (two-tailed test). ** p < .01 (two-tailed test). Job level is coded 1 = Other, 2 = HR Manager, 3 =Director, 4 = VP; Tenure = number of years in HR; Education is coded 1 = H.S., 2 = Bachelors, 3 = Masters, 4 = Ph.D.; HRmajor = 1, else = 0; PHR certification = 1, else = 0; SHPR certification = 1, else = 0. Knowledge = # correct of 35 possible.All other variables on 5-point scales. Reliability for Knowledge is test-retest; other reliabilities are coefficient alphas.

TABLE V Correlations among Major Study Variables


In clear contrast to this assertion, our re-sults suggest that there are in fact very largedifferences across companies in what theirHR leaders know about best practices in HRand, furthermore, that the average level ofknowledge does not appear to be very impres-sive. Moreover, it is quite likely that our re-sults represent a “best case” scenario, giventhat we sampled the highest-level HR practi-tioners and that those who knew even lesswere probably disproportionately likely to benonrespondents. In addition, the weakestknowledge areas were not limited to obscure,unimportant, or little-researched issues.Rather, the biggest gaps between researchfindings and practitioner beliefs concernsome of the most central issues in HR: first,how to choose the best employees and, sec-ond, how to effectively motivate themthrough appropriate goal-setting and effec-tive performance management.

Research at multiple levels of analysis in-creasingly suggests that differences in knowl-edge and/or application of these practices canhave large impacts on a firm’s bottom line (e.g.,Huselid, 1995). Moreover, results from Ulrichet al.’s (1995) study of how HR practitionersare evaluated by others suggests that lack ofHR knowledge can also be a serious detrimentto the career reputations of individual HRpractitioners. Given the apparent importanceof knowledge to both individual and organiza-tional outcomes, a central issue for future re-searchers, practitioners, and professionalorganizations such as SHRM concerns howto disseminate research findings more accu-rately and effectively.

A second finding from our research is thatpractitioners’ beliefs diverge from researchfindings much more widely in some contentareas than in others. Clearly, the area inwhich the largest gaps were uncovered in thisstudy was selection. Specifically, five of thelargest discrepancies involved selection-re-lated issues: the relative validity of selectingfor intelligence versus values (only 16% agree-ment), the relative validity of intelligenceversus conscientiousness as predictors ofperformance (18% agreement), the validityof integrity tests (31%), the absence of ad-verse impact for integrity tests (32%), and theusefulness of intelligence for performing even

low-skilled jobs (42%). Even more specifically,three of the most problematic items revealedlack of awareness of the usefulness of intel-ligence as a predictor of performance in al-most every type of job.

Despite the extensive research base un-derlying these selection findings, our resultssuggest that HR managers are generally un-aware of this entire body of literature. A num-ber of factors may account for this. First,selection research is more technical than re-search in many other areas of HR, frequentlyinvolving such complex procedures as meta-analysis, corrections for measurement error,and utility analysis. Thus, unless this researchis translated for nonacademic audiences, it haslittle chance of being read and understood bymost practicing managers.

Second, the frame of reference with respectto selection appears to be different for academ-ics than it is for practitioners. For example, se-lection researchers focus on gatheringknowledge about rather abstract characteristicsof people (such as “intelligence” or “conscien-tiousness”), whereas most recruitment and se-lection activities are designed around the job.Thus, for example, interviewers are trained todevelop questions that focus on applicants’ abili-ties to perform specific job tasks or to handleparticular job-related situations, rather than toattempt to infer someone’s personality or intel-ligence from interview responses. The practiceof structuring job descriptions and interviewquestions around job analysis is consistent withearlier recommendations (following passage ofTitle VII of the Civil Rights Act) designed to im-prove selection validity and to reduce the likeli-hood of unintentional bias or discrimination. Infact, these recommendations (to standardizeinterview content around job-related questions)have been notably successful in improving thevalidity of the employment interview (McDanielet al., 1994; Schmidt & Hunter, 1998). How-ever, they may not be very helpful in revealingthe linkages between the interview behaviorsexhibited by applicants and the underlying traits(such as intelligence or conscientiousness) thatcause some applicants to produce more effec-tive answers than others.

A third reason for the failure of selectionresearch to take root in practice probably re-sides in the negative coverage that intelligence

… our resultssuggest that thereare in fact verylarge differencesacross companiesin what their HRleaders knowabout bestpractices in HRand,furthermore, thatthe average levelof knowledgedoes not appearto be veryimpressive.


and ability testing continue to receive in thepopular press (e.g., Goleman, 1998). Despiteoverwhelming evidence of the usefulness ofability tests for predicting performance (seeHunter, 1986), the persistence of negativepress coverage has almost certainly convincedsome practitioners that intelligence is an over-rated construct, and/or that it cannot be use-fully assessed via paper-and-pencil tests.Furthermore, the fact that ability tests pro-duce adverse impact against certain groups hasbrought testing squarely into the politicalarena, where facts are often selectively cho-sen or distorted to support desired ends. Inturn, political and legal uncertainties abouttesting may have caused some practitionersto lose interest in research about intelligence,feeling that attempts to assess it are too risky.5

Whatever the reasons, lack of awarenessof broad selection principles can be verycostly to organizations, since the character-istics of selected employees place inevitablelimits on the extent to which other manage-ment practices (such as training and perfor-mance management) will be effective. Basedon a meta-analysis of twenty-five years’ worthof data, researchers at Gallup came to thesame conclusion: “People don’t change thatmuch….great managers know there is a limitto how much remolding they can do to some-one” (Buckingham & Coffman, 1999, pp. 56–57). Because of this, it is crucial to know howto select the right people in the first place.As further evidence of this proposition,Terpstra and Rozell (1993) showed that theuse of research-validated staffing practices(such as use of general mental ability tests,structured interviews, and weighted applica-tion blanks) was significantly associated withboth annual profit (r = .20) and profit growth(r = .14) across industries, with relationshipsbeing particularly strong in the service andfinancial sectors.

A third contribution of our research wasto uncover some clues about the characteris-tics of HR managers that seem to be associ-ated with higher knowledge. First, managersat higher job levels appear to have somewhatbetter knowledge of HR practices. For ex-ample, respondents at the highest job level(HR vice president) scored 20.86 on average,as opposed to 18.66 for those at the lowest

level (raw difference = 2.20; d = .60, p < .01).Similarly, those who had received SPHR cer-tification scored 20.84 versus 19.99 for thosewith no certification or only PHR certifica-tion (raw difference = .85; d = .23, p < .01).

Finally, our findings suggest that knowl-edge of research findings could be improvedif practitioners were to read academic jour-nals. For example, we found that practitio-ners who “usually” read HR researchjournals scored an average of 3.15 pointshigher than all other respondents—a 15%advantage (d = .86; p < .05). However, ourresults also revealed a fatal flaw with this po-tential “solution”: the vast majority of practi-tioners simply do not read the academicliterature. In our study, only six of our 959respondents (fewer than 1%) typically readthese sources. In stark contrast, fully 75% ofrespondents reported that they never read anyof the three top academic journals. The mes-sage seems clear: alternative means of dissemi-nation must be found if practitioners are toacquire greater research knowledge.

Looking to the Future

Because practitioners do not read academicjournals, the most promising solution is to putmore research content into practitioner jour-nals and other formats commonly used by HRpractitioners (e.g., the SHRM Web site). Atpresent, nearly all the content in practitionerjournals is provided by consultants and prac-ticing HR professionals. Although these canbe useful sources of information, it is impor-tant to have researcher input as well, since(as we have seen) research findings can some-times vary considerably from practitioner be-liefs and consultant recommendations.

As an illustration of these points, we offeran example taken from the most widely readpractitioner journal, HR Magazine, discuss-ing the most commonly misunderstood HR re-search area (selection). In 1999, the Magazinepublished an article entitled “Reading Employ-ees,” a pro-and-con discussion of handwrit-ing analysis (graphology) as a selection tool(Leonard, 1999). The article began by statingthat a growing number of U.S. companies areusing graphology to screen and place job ap-plicants. It then quoted the president of a

… lack ofawareness ofbroad selectionprinciples can bevery costly toorganizations,since thecharacteristics ofselectedemployees placeinevitable limitson the extent towhich othermanagementpractices (suchas training andperformancemanagement)will be effective.


handwriting consulting firm who said, “I defi-nitely believe that this will be the career guid-ance tool of the future. With handwritinganalysis, you get right to the truth about aperson. It can reveal weaknesses as well asstrengths and hidden talents.” It also quoteda Senior VP of HR who said that her com-pany “lives and dies by handwriting analysis.”Although the article does contain warningsfrom two lawyers about possible legal prob-lems with handwriting analysis due to weakvalidity evidence, these warnings are subse-quently countered by a graphologist who says,“The data clearly support the claims that gra-phology works and can be an accurate assess-ment tool.” Finally, the article concluded witha quote by another VP with SPHR status:“We’ve used it for nine years and haven’t hadone complaint. You can be overly cautious andchoose not to do anything, but then you endup being just too afraid to move. I feel thatwhen you have an excellent and accurate as-sessment tool at your disposal, then you shoulduse it.”

In other words, although the article raisessome possible cautions about handwritinganalysis (particularly with respect to its legalstatus), both the introduction and conclusionto the article are very positive about graphol-ogy. However, not a single academic researcherwas consulted for the article; a notable exclu-sion, since the presence of a credible academicopinion would likely have produced an alto-gether different conclusion. This is becausecumulative research on “content-free” hand-writing analysis (i.e., analysis of handwritingonly, as opposed to handwriting-plus-contentanalysis of essays assessing personality orwork-related characteristics) has shown thatgraphology has absolutely no validity as a se-lection tool. According to Schmidt andHunter’s (1998) summary of previous graphol-ogy research,

Whatever limited information about person-ality or job performance there is in the hand-writing samples comes from the content andnot the characteristics of thehandwriting…actually, most of the variationin handwriting is due to differences amongindividuals in fine motor coordination of thefinger muscles. And these differences in fin-

ger muscles and their coordination are prob-ably due mostly to random genetic variationsamong individuals…which do not appear tobe linked to personality. (p. 270)

Thus, a simple phone call to FrankSchmidt or Jack Hunter would likely haveshifted the theme of the article from a “pro-con” debate to a “beware of this invalid prac-tice” theme. Moreover, the article would thenhave had little reason to debate legality, sincethe absence of validity is a much more fun-damental reason for not spending money ona potential selection practice than is uncer-tain legality.

Although practitioner journals currentlydo not offer much in the way of research find-ings, we believe that now may be a particularlygood time to increase research coverage. First,our results suggest that most practitioners,particularly the less experienced ones, wouldlike to know more about research findings(Table IV). Second, two relatively recent pat-terns have emerged in HR research that webelieve have increased the potential useful-ness of research findings to practitioners.

One of these developments concerns theincreased use of meta-analysis to summarizefindings from multiple studies of the same is-sue. By empirically combining data from mul-tiple studies and giving greater weight to thosestudies with larger samples, researchers arenow able to provide much more reliable esti-mates of effect sizes and to give an indicationof the extent to which findings appear to beuniversal, as opposed to dependent on particu-lar circumstances. By using meta-analysis,researchers have been able to obtain muchmore reliable estimates of important relation-ships such as the ones between personalityand job performance (Barrick & Mount, 1991;Ones, Viswesvaran, & Schmidt, 1993) or be-tween employee satisfaction and job perfor-mance (Judge, Thoresen, Bono, & Patton,2001). These are truly “big questions” withclear implications for practitioners.

The other positive development is the ex-tent to which HR researchers have shiftedfrom a nearly exclusive concern with individualbehavior to a broader concern with organiza-tional activities and outcomes. For example,the past decade has seen the emergence of

Althoughpractitionerjournalscurrently do notoffer much in theway of researchfindings, webelieve that nowmay be aparticularly goodtime to increaseresearchcoverage.


important research concerning how organiza-tional performance is related to compensationpractices (Gerhart & Milkovich, 1990), selec-tion practices (Terpstra & Rozell, 1993), anddifferent configurations of overall HR systems(e.g., Arthur, 1994; Delery & Doty, 1996;Huselid, 1995). The result is that, in an in-creasing number of areas, we no longer needto “infer up” from the individual to the orga-nizational level of analysis.

At the same time that we urge practitionerjournals to increase their coverage of researchfindings, we also caution that this may not besufficient to entice researchers to present theirwork in such journals. Issues of academicincentive structures (e.g., promotion andtenure) aside, researchers may be reluctantto have their work published in journals wherescientifically rigorous, peer-reviewed findingsare published alongside articles (such as theone on handwriting analysis) that flatlycontradict research findings. Thus, the editorsof practitioner outlets will not only have tohave a stronger commitment to publishingresearch findings, but also to validating theinformation contained in non-research-basedarticles. The mechanisms to ensure that thiswill occur are not entirely clear at present,although one suggestion might be forprofessional associations such as SHRM, TheSociety for Industrial & OrganizationalPsychology (SIOP), and the Academy ofManagement to maintain a clearinghouse tolink research experts to reporters.

Because such changes will take time, theaddition of special research sections to prac-titioner outlets might also be accompanied by“handbooks” or edited volumes that call onrespected researchers to translate their work(and its implications) for practicing manag-ers. Two notable attempts in this regard haverecently been undertaken by Cooper andLocke (2000) and Locke (2000).

Another recommendation is that journalsthat traditionally cover only “HR” knowledgebegin to expand to cover research and imple-mentation issues related to change manage-ment and, to a lesser extent, general businessissues. As mentioned earlier, Ulrich et al.(1995) found that the Change Managementcompetency, more than any other, determinedhow HR practitioners were viewed in others’

eyes. Given this reality, even considerably bet-ter knowledge of the HR literature might notbe sufficient for the HR function (and indi-vidual HR practitioners) to improve their cred-ibility with others in the organization.

A final recommendation would be for re-searchers to take a closer look at some of theissues revealed by our survey to exhibit widegaps between research findings and practi-tioner beliefs. One such issue would be theenormous gap between researcher and prac-titioner beliefs about intelligence. More de-tailed analysis of practitioner beliefsconcerning intelligence may help to disen-tangle the extent to which the discrepancy isdue to each of the following potential expla-nations: (a) lack of knowledge about researchfindings (as suggested by this research), (b)exposure to conflicting findings by others whoalso position themselves as “research experts”(e.g., Goleman, 1998), (c) concerns aboutlegality, (d) direct disagreement with the find-ings (despite knowledge), or (e) agreementwith the findings, but disagreement with thepractical implications of the findings (e.g.,that adverse impact might be created againstminorities), and hence, potential misrepre-sentation of beliefs.6

The fruits of such a (more detailed) re-search approach may be considerable.Previous research has shown that some of themost important advances in HR science andpractice have emerged from attempts to re-solve differences in beliefs across researcherand practitioner perspectives (Cooper &Locke, 2000). As just one example, for decadespractitioners persisted in using the employ-ment interview as a selection device despitescathing reviews of their validity by academ-ics (e.g., Dunnette & Bass, 1963). However,this continued use by practitioners spurred re-searchers to find ways to improve the validityof the interview through such techniques asjob analysis, structured interviewing, note-tak-ing, and formalized scoring. Eventually,meta-analytic techniques were developed thatnot only demonstrated the superiority of thenewer interviewing methods (McDaniel et al.,1994), but also showed that even unstructuredinterviews have validity, as asserted by practi-tioners all along (see also Kluger &Tikochinsky, 2001). Thus, many decades of

Researchers maybe reluctant tohave their workpublished injournals wherescientificallyrigorous, peer-reviewed findingsare publishedalongside articlesthat flatlycontradictresearchfindings. Thus,the editors ofpractitioneroutlets will notonly have to havea strongercommitment topublishingresearchfindings, but alsoto validating theinformationcontained innon-research-based articles.


research ultimately revealed that the re-searcher and practitioner perspectives eachhad some validity, with the end result beingimproved research and improved practice(Rynes, Barber, & Varma, 2000).

Thus, researchers may find rich rewardsfrom attempting to probe more directly intothe discrepancies revealed by the presentsurvey, such as the differing views of practi-tioners and academics with respect to theroles of intelligence, values, personality, andemployee participation in predicting indi-vidual and organizational performance. Bygetting closer to the root sources of thesediscrepancies, research as well as practiceis likely to benefit.

APPENDIX: SOURCES OFADDITIONAL

INFORMATION ABOUT THEQUESTIONNAIRE ITEMS

IN TABLE I

1. Barling, J.E., Weber, T., & Kelloway, E.K. (1996).Effects of transformational leadership trainingon attitudinal and financial outcomes: A fieldexperiment. Journal of Applied Psychology, 81,827–833. Loehlin, J.C., McCrae, R.R., Costa,P.T., & John, O.P. (1998). Heritabilities of com-mon and measure-specific components of the BigFive personality factors. Journal of Research inPersonality, 32, 431–453. Judge, T.A., & Bono,J.E. (2000). Five-factor model of personality andtransformational leadership. Journal of AppliedPsychology, 85, 751–765. Riemann, R.,Angleitner, A., & Strelau, J. (1997). Genetic andenvironmental influences on personality: A studyof twins reared together using the self- and peer-report NEO-FFI scales. Journal of Personality,65, 449–475.

2. Bass, B.M. (1990). From transactional to trans-formational leadership: Learning to share the vi-sion. Organizational Dynamics, 18 (3), 19–31.Judge, T.A., Bono, J.E., Ilies, R., & Werner, M.(in press). Personality and leadership: A qualita-tive and quantitative review. Journal of AppliedPsychology. Lord, R.G., Devader, C.L., & Alliger,G.M. (1986). A meta-analysis of the relationshipbetween personality traits and leadership percep-tions—An application of validity generalizationprocedures. Journal of Applied Psychology, 71,402–410. Bennis, W., & Nanus, B. (1997). Lead-

ers: Strategies for taking charge. New York:Harper Business. Collins, J. (2001). Good togreat. New York: Harper-Collins.

3. Locke, E.A., & Latham, G.P. (1990). A theory ofgoal-setting and task performance. Upper SaddleRiver, NJ:. Locke, E.A., Schweiger, D.M., &Latham, G.P. (1986). Participation in decisionmaking: When should it be used? OrganizationalDynamics, 14, 65–79. Wagner, J.A., III. (1994).Participation’s effect on performance and satis-faction: A reconsideration of the research evi-dence. Academy of Management Review, 19,312–330.

4. Baum, J.R., Locke, E.A., & Kirkpatrick, S.A.(1998). A longitudinal study of the relation ofvision and vision communication to venturegrowth in entrepreneurial firms. Journal of Ap-plied Psychology, 83, 43–54. Hoch, D.J., Roeding,C., Purkert, G., & Lindner, S.K. (1999). Secretsof software success: Management insights from100 software firms around the world. Boston: HBRPress. Brown, S., & Eisenhardt, K. (1995). Acad-emy of Management Review.

5. Reichheld, F.F. (1996). The loyalty effect. Bos-ton: Harvard Business School Press.

6. Morris, J.R., Cascio, W.F., & Young, C.E. (1999).Downsizing after all these years: Questions andanswers about who did it, how many did it, andwho benefited from it. Organizational Dynam-ics, 27 (3), 78–87. Cameron, K.S., Freeman, S.J.,& Mishra, A.K. (1993). Downsizing and redesign-ing organizations. In G.P. Huber & W.H. Glick(Eds.), Organizational change and redesign (pp.19–65). New York: Oxford University Press.

7. Ulrich, D., Brockbank, W., Yeung, A.K., & Lake,D.G. (1995). Human resource competencies: Anempirical assessment. Human Resource Manage-ment, 34, 473–495.

8. Locke, E.A., Feren, D.B., McCaleb, V.M., Shaw,K.N., & Denny, A.T. (1980). The relative effec-tiveness of four methods of motivating employeeperformance. In K.D. Duncan, M.M. Gruneberg,& D. Wallis (Eds.), Changes in working life (pp.363–388). London: Wiley. Locke & Latham,1990, op. cit., note 3. Locke, E.A., Schweiger,D.M., & Latham, G.P. (1986). Participation indecision making: When should it be used? Orga-nizational Dynamics, 14, 65–79. Wagner, J.A.,III. (1994). Participation’s effect on performanceand satisfaction: A reconsideration of the researchevidence. Academy of Management Review, 19,312–330.


9. Longenecker, C.O., Sims, H.P., & Gioia, D.A.(1987). Behind the mask: The politics of em-ployee appraisal. Academy of Management Ex-ecutive, 1, 183–193. Jawahar, I.M., & Williams,C.R. (1997). Where all the children are aboveaverage: The performance appraisal purpose ef-fect. Personnel Psychology, 50, 905–926.

10. Kruger, J., & Dunning, D. (1999). Unskilled andunaware of it: How difficulties in recognizingone’s own incompetence lead to inflated self-as-sessments. Journal of Personality and Social Psy-chology, 77, 1121–1134.

11. Hambrick, D.C., Cho, T.S., & Chen, M.J. (1996).The influence of top management team hetero-geneity on firms’ competitive moves. Administra-tive Science Quarterly, 41, 659–684. Keller, R.T.(2001). Cross-functional project groups in re-search and new product development: Diversity,communications, job stress and outcomes. Acad-emy of Management Journal, 44, 547–555. Lutz,R.A. (1994). Implementing technological changewith cross-functional teams. Research Technol-ogy Management, 37 (2), 14–18. Northcraft, G.,Polzer, J., Neale, M., & Kramer, R. (1995). Di-versity, social identity, and performance: Emer-gent social dynamics in cross-functional teams.In S. Jackson & M. Ruderman (Eds.), Diversityin work teams (pp. 69–96). Washington, DC: APAPress. Pelled, L.H., Eisenhardt, K.M., & Xin, K.R.(1999). Exploring the black box: An analysis ofwork group diversity, conflict, and performance.Administrative Science Quarterly, 44, 1–28.

12. McDaniel, M.A. (1988). Does pre-employmentdrug use predict on-the-job suitability? Person-nel Psychology, 41, 717–729. Normand, J.,Salyards, S.D., & Mahoney, J.J. (1990). An evalu-ation of pre-employment drug testing. Journal ofApplied Psychology, 75, 629–639. Parish, D.C.(1989). Relation of the pre-employment drugtesting result to employment status: A one-yearfollow-up. Journal of General Internal Medicine,4, 44–47. Winkler, H., & Sheridan, J. (1989,September). An examination of behavior relatedto drug use at Georgia Power Company.

13. Brown, J.D. (1986). Evaluations of self and oth-ers: Self-enhancement biases in social judgment.Social Cognition, 4, 353–376. Kruger & Dun-ning (1999), op. cit., note 10. Harris, M.M., &Schaubroeck, J. (1988). A meta-analysis of self-boss, self-peer, and peer-boss ratings. PersonnelPsychology, 41, 43–62. Larwood, L., & Whittaker,W. (1977). Managerial myopia: Self-serving bi-

ases in organizational planning. Journal of Ap-plied Psychology, 62, 194–198. Mabe, P.A., III,& West, S.G. (1982). Validity of self-evaluationof ability: A review and meta-analysis. Journal ofApplied Psychology, 67, 280–296. Taylor, S.E.,& Brown, J.D. (1988). Illusion and well-being: Asocial psychological perspective on mental health.Psychological Bulletin, 103, 193–210. Thornton,G.C. (1980). Psychometric properties of self-ap-praisals of job performance. Personnel Psychol-ogy, 33, 263–271.

14. Latham, G.P., & Wexley, K.N. (1994). Increas-ing productivity through performance appraisal.Reading, MA: Addison-Wesley. Wexley, K.N.,Sanders, R.E., & Yukl, G.A. (1973). Training in-terviewers to eliminate contrast effects in em-ployment interviews. Journal of AppliedPsychology, 57, 233–236. Bernardin, H.J., &Pence, E.G. (1980). The effects of rater train-ing: Creating new response sets and decreasingaccuracy. Journal of Applied Psychology, 65, 60–66. Longenecker et al. (1987), op. cit., note 9.

15. Clark, R.E. (1983) Reconsidering research onlearning from media. Review of Educational Re-search, 53, 445–459. Kulik, C.C., & Kulik, J.A.(1991). Effectiveness of computer-based instruc-tion: An updated analysis. Computers in humanbehavior, 7, 75–94. Gagne, R.M., & Medsker,K.L. (1996). The conditions of learning: Train-ing applications. Fort Worth, TX: HarcourtBrace. Simonson, M., Schlosser, C., & Hanson,D. (1999). Theory and distance education: A newdiscussion. The American Journal of DistanceEducation, 13, 60–75.

16. Colquitt, J.A., LePine, J.A., & Noe, R.A. (2000).Toward an integrative theory of training motiva-tion: A meta-analytic path analysis of 20 years ofresearch. Journal of Applied Psychology, 85, 678–707. Salas, E., Cannon-Bowers, J.A., Rhodenizer,& Bowers, C.A. (1999). Training in organizations:Myths, misconceptions, and mistaken assump-tions. In G. Ferris (Ed.), Research in personneland human resources management (Vol. 17, pp.123–161). Greenwich, CT: JAI Press. Horn, J.L.(1980). Concepts of intellect in relation to learn-ing and adult development. Intelligence, 4, 285–317. Horn, J.L., & Cattell, R.B. (1967). Agedifferences in fluid and crystallized intelligence.Acta Psychologica, 26, 107–129. Gist, M., Rosen,B., & Schwoerer, C. (1988). The influence oftraining method and trainee age on the acquisi-tion of computer skills. Personnel Psychology, 41,


255–265. Warr, P., & Bunce, D. (1995). Traineecharacteristics and the outcomes of open learn-ing. Personnel Psychology, 48, 347–375.

17. Colquitt et al., op. cit., note 16. Rouiller, J.Z., &Goldstein, I.L. (1993). The relationship betweenorganizational transfer climate and positive trans-fer of training. Human Resource DevelopmentQuarterly, 4, 377–390. Tracey, J.B., Tannenbaum,S.I., & Kavanaugh, M.J. (1995). Applying trainedskills to the job: The importance of the work en-vironment. Journal of Applied Psychology, 80,239–252. Warr, P., Allan, C., & Birdi, K. (1999).Predicting three levels of training outcome. Jour-nal of Occupational and Organizational Psychol-ogy, 72, 351–375.

18. Donovan, J.J., & Radosevich, D.J. (1999). Ameta-analytic review of the distribution of prac-tice effect: Now you see it, now you don’t. Jour-nal of Applied Psychology, 84, 795–805. Lee,T.D., & Genovese, E.D. (1988). Distribution ofpractice in motor skill acquisition: Learning andperformance effects reconsidered. ResearchQuarterly, 59, 277–287.

19. McDaniel, M.A., Whetzel, D.L., Schmidt, F.L.,& Maurer, S.D. (1994). The validity of employ-ment interviews: A comprehensive review. Jour-nal of Applied Psychology, 79, 599–616. Wiesner,W.H., & Cronshaw, S.F. (1988). The moderat-ing impact of interview format and degree ofstructure on the validity of the employment in-terview. Journal of Occupational Psychology, 61,275–290.

20. Digman, J.M. (1990). Personality structure:Emergence of the five-factor model. Annual Re-view of Psychology, 41, 417–440. Gardner, W.,& Martinko, M. 1996. Using the MBTI to studymanagers: A literature review and researchagenda. Journal of Management, 22, 45–83.

21. Conard, M.A., & Ashworth, S.D. (1986). Recruit-ing source effectiveness: A meta-analysis and re-examination of two rival hypotheses. SIOP annualmeeting. Decker, P.J., & Cornelius, E.T., III.(1979). A note on recruiting sources and job sur-vival rates. Journal of Applied Psychology, 64,463–464. Rynes, S. (1991). Recruitment, jobchoice, and post-hire consequences: A call fornew research directions. In M.D. Dunnette & L.M. Hough (Eds.), Handbook of industrial andorganizational psychology (2nd ed., Vol. 2, pp.399–444). Palo Alto, CA: Consulting Psycholo-gists Press.

22. Hunter, J.E. (1986). Cognitive ability, cognitive

aptitudes, job knowledge, and job performance.Journal of Vocational Behavior, 29, 340–362.Schmidt, F.L., & Hunter, J.E. (1998). The valid-ity and utility of selection methods in personnelpsychology: Practical and theoretical implicationsof 85 years of research findings. PsychologicalBulletin, 124, 262–274.

23. Barrick, M.R., & Mount, M.K. (1991). The bigfive personality dimensions and job performance—A meta-analysis. Personnel Psychology, 44, 1–26.Gardner, W.L., & Martinko, M.J. (1996). Usingthe Myers–Briggs type indicator to study manag-ers: A literature review and research agenda. Jour-nal of Management, 22, 45–83.

24. Ones, D.S., Viswesvaran, C., & Schmidt, F.L.(1993). Comprehensive meta-analysis of integ-rity test validities: Findings and implications forpersonnel selection and theories of job perfor-mance. Journal of Applied Psychology, 78, 679–703. Ones, D.S., Viswesvaran, C., & Reiss, A.D.(1996). Role of social desirability in personalitytesting for personnel selection: The red herring.Journal of Applied Psychology, 81, 660–679.

25. Ones, D.S., & Viswesvaran, C. (1998). Gender,age, and race differences on overt integrity tests:Results across four large-scale job applicant datasets. Journal of Applied Psychology, 83, 35–42.

26. Schmidt, F.L. & Hunter, J.E. (1998). The valid-ity and utility of selection methods in personnelpsychology: Practical and theoretical implicationsof 85 years of research findings. PsychologicalBulletin, 124, 262–274.

27. Ibid. Adkins, C.L., Ravlin, E.C., & Meglino, B.M.(1996). Value congruence between co-workersand its relationship to work outcomes. Group andOrganization Management, 21, 439–460.Meglino, B.M., & Ravlin, E.C. (1998). Individualvalues in organizations: Concepts, controversies,and research. Journal of Management, 24, 351–389. Ravlin, E.C., & Meglino, B.M. (1987). Ef-fect of values on perception and decision making:A study of alternative work values measures. Jour-nal of Applied Psychology, 72, 666–673.

28. Greenberg, J. (1990). Employee theft as a reac-tion to under-payment inequity: The hidden costof pay cuts. Journal of Applied Psychology, 75,561–568. Greenberg, J. (1993). Stealing in thename of justice: Informational and interpersonalmoderators of theft reactions to underpaymentinequity. Organizational Behavior and HumanDecision Processes, 54, 81–103.

29. Cable, D.M., & Judge, T.A. (1994). Pay prefer-


ences and job search decisions: A person–orga-nization fit perspective. Personnel Psychology, 47,317–348. Bretz, R.D., Ash, R.A., & Dreher, G.F.(1989). Do people make the place? An examina-tion of the attraction-selection-attrition hypoth-esis. Personnel Psychology, 42, 561–580.

30. Kopelman, R.E., & Reinharth, L. (1982). Researchresults: The effect of merit-pay practices on whitecollar performance. Compensation Review, 14 (4),30–40. Heneman, R.L. (1992). Merit pay: Link-ing pay increases to performance ratings. Read-ing, MA: Addison-Wesley. Kopelman, R.E.,Rovenpor, J.L., & Cayer, M. (1991). Merit pay andorganizational performance: Is there an effect onthe bottom line? National Productivity Review,299–307. Pearce, J.L., Stevenson, W.B., & Perry,J.L. (1985). Managerial compensation based onorganizational performance: A time series analy-sis of the effects of merit pay. Academy of Man-agement Journal, 28, 261–278.

31. Gerhart, B., & Milkovich, G.T. (1990). Organi-zational differences in managerial compensationand firm performance. Academy of ManagementJournal, 33, 663–691. Welbourne, T.M., &Andrews, A.O. (1996). Predicting the perfor-mance of initial public offerings: Should humanresource management be in the equation? Acad-emy of Management Journal, 39, 891–919.

32. Welbourne & Andrews (1996), op. cit., note 31.33. Prince, J.B., & Lawler, E.E., III. (1986). Does

salary discussion hurt the developmental perfor-

mance appraisal? Organizational Behavior andHuman Decision Processes, 37, 357–375. Welch,J. Jack: Straight from the gut. New York: WarnerBooks.

34. Bureau of National Affairs. (1988). Changingpay practices: New developments in employeecompensation. Washington, DC: Author.Cable, D.M., & Judge, T.A. (1994), op. cit.,note 29. Brown, K.A., & Huber, V.L. (1992).Lowering floors and raising ceilings: A longi-tudinal assessment of the effects of an earn-ings-at-risk plan on pay. Jensen, M.C., &Meckling, W.H. (1976). Theory of the firm:Managerial behavior, agency costs, and own-ership structure. Journal of Financial Econom-ics, 3, 305–360.

35. Feldman, D.C., & Arnold, H.J. (1978). Positionchoice: Comparing the importance of organiza-tional and job factors. Journal of Applied Psy-chology, 63, 706–710. Jurgensen, C.E. (1978).Job preferences: What makes a job good or bad?Journal of Applied Psychology, 63, 267–276.Rynes, S.L., Schwab, D.P., & Heneman, H.G.,III. (1983). The role of pay and market pay vari-ability in job application decisions. OrganizationalBehavior and Human Performance, 31, 353–364.Slovic, P., & Lichtenstein, S. (1971). Compari-son of Bayesian and regression approaches to thestudy of information processing in judgment. Or-ganizational Behavior and Human Performance,6, 649–744.


REFERENCES

Adkins, C.L, Ravlin, E. C., & Meglino, B. M. (1996).Value congruence between co-workers and itsrelationship to work outcomes. Group and Or-ganization Management, 21, 439–460.

Arthur, J.B. (1994). Effects of human resource sys-tems on manufacturing performance and turn-over. Academy of Management Journal, 37,670–687.

Barrick, M.R., & Mount, M.K. (1991). The big fivepersonality dimensions and job performance: Ameta-analysis. Personnel Psychology, 44, 1–26.

Barrick, M.R., & Mount, M.K. (2000). Select on con-scientiousness and emotional stability. In E.A.Locke (Ed.), The Blackwell handbook of prin-ciples of organizational behavior. Oxford:Blackwell.

Becker, B., & Gerhart, B. (1996). The impact of hu-man resource management on organizationalperformance. Academy of Management Journal,39, 779–801.

Behling, O. (1998). Employee selection: Will intelli-

gence and conscientiousness do the job? Acad-emy of Management Executive, 12, 77–85.

Buckingham, M., & Coffman, C. (1999). First, breakall the rules: What the world’s greatest managersdo differently. New York: Simon & Schuster.

Cable, D.M., & Judge, T.A. (1994). Pay preferencesand job search decisions: A person–organizationfit perspective. Personnel Psychology, 47, 317–348.

Campbell, J.P., Daft, R.L., & Hulin, C.L. (1982). Whatto study: Generating and developing researchquestions. Beverly Hills, CA: Sage.

Cooper, C.L., & Locke, E.A. (2000). Industrial andorganizational psychology: Linking theory withpractice. Oxford: Blackwell.

Delery, J.E., & Doty, D.H. (1996). Modes of theoriz-ing in strategic human resource management:Tests of universalistic, contingency, and configu-rational performance predictions. Academy ofManagement Journal, 39, 802–835.

Dunnette, M.D., & Bass, B.M. (1963). Behavioralscientists and personnel management. IndustrialRelations, 2, 115–130.

Sara L. Rynes is the John F. Murray Professor and Chair of the Department of Man-agement and Organizations at the University of Iowa. Rynes is currently an AssociateEditor for the Academy of Management Journal and previously served on the editorialboards of the Journal of Applied Psychology, Personnel Psychology, and Quality Man-agement Journal. She is a Fellow of the American Psychological Association and theSociety for Industrial and Organizational Psychology. Her research interests includerecruitment and job choice, compensation, and knowledge transfer between academ-ics and practitioners. She holds a Ph.D. in Industrial Relations from the Universityof Wisconsin and was previously on the faculties of the University of Minnesota andCornell University.

Amy E. Colbert is a doctoral candidate and University of Iowa Fellow in the Depart-ment of Management and Organizations. Her research interests are in the areas ofleadership, personality, and innovation. Prior to joining the doctoral program at Iowa,Ms. Colbert received an M.A. in Decision Sciences from St. Louis University and aB.S. in Accounting and Math from Culver-Stockton College. She is a former internalauditor and winner of The Wall Street Journal Award for Accounting Excellence.

Kenneth G. Brown is an Assistant Professor of Management & Organizations at theUniversity of Iowa. His research focuses on training, individual and organizational learn-ing, computer technology and human performance, group influences in organizations,motivation and self-regulation, and knowledge transfer. He has published articles inJournal of Applied Psychology, Personnel Psychology, Organizational Behavior and Hu-man Decision Processes, and Human Resource Planning. He received his B.S. in Psy-chology from the University of Maryland and his Ph.D. in Industrial & OrganizationalPsychology from Michigan State University.


Feldman, D.C., & Arnold, H.J. (1978). Positionchoice: Comparing the importance of organiza-tional and job factors. Journal of Applied Psy-chology, 63, 706–710.

Gannon, M.J. (1983). Managerial ignorance. Busi-ness Horizons, May–June, 26–32.

Gerhart, B., & Milkovich, G.T. (1990). Organizationaldifferences in managerial compensation and fi-nancial performance. Academy of ManagementJournal, 33, 663–691.

Goleman, D. (1998). Working with emotional intelli-gence. New York: Bantam Books.

Hunter, J.E. (1986). Cognitive ability, cognitive apti-tudes, job knowledge, and job performance. Jour-nal of Vocational Behavior, 29, 340–362.

Huselid, M.A. (1995). The impact of human resourcemanagement practices on turnover, productivity,and corporate financial performance. Academyof Management Journal, 38, 635–672.

Johns, G. (1993). Constraints on the adoption of psy-chology-based personnel practices, Lessons fromorganizational innovation. Personnel Psychology,46, 569–592.

Judge, T.A., Bono, J.E., Ilies, R., & Werner, M. (inpress). Personality and leadership: A qualitativeand quantitative review. Journal of Applied Psy-chology.

Judge, T.A., Thoresen, C.J., Bono, J.E., & Patton, G.K.(2001). The job satisfaction–job performancerelationship: A qualitative and quantitative review.Psychological Bulletin, 127, 376–407.

Kluger, A.N., & Tikochinsky, J. (2001). The error ofaccepting the “theoretical” null hypothesis: Therise, fall, and resurrection of commonsense hy-potheses in psychology. Psychological Bulletin,127, 408–423.

Latham, G.P., & Latham, S.D. (2000). Overlookingtheory and research in performance appraisal atone’s peril: Much done, more to do. In C.L. Coo-per & E.A. Locke (Eds.), Industrial and organi-zational psychology: Linking theory with practice(pp. 199–215). Oxford: Blackwell.

Latham, G.P., & Wexley, K.N. (1980). Increasing pro-ductivity through performance appraisal. Read-ing, MA: Addison-Wesley.

Lauver, K., & Kristof-Brown, A. (in press). Distinguish-ing between employees’ perceptions of person-job and person-organization fit. Journal ofVocational Behavior.

Leonard, B. (1999). Reading employees. HumanResource Magazine, 44 (4), 67–73.

Locke, E.A. (Ed.). (2000). The Blackwell handbook

of principles of organizational behavior. Oxford:Blackwell.

Longenecker, C.O., Gioia, D.A., & Sims, H.P. (1987).Behind the mask: The politics of employee ap-praisal. Academy of Management Executive, 1,183–193.

McDaniel, M.A., Whetzel, D.L., Schmidt, F.L., &Maurer, S.D. (1994). The validity of employmentinterviews: A comprehensive review. Journal ofApplied Psychology, 79, 599–616.

Offermann, L.R., & Spiros, R.K. (2001). The scienceand practice of team development. Academy ofManagement Journal, 44, 376–392.

Ones, D.S., Viswesvaran, C., & Schmidt, F.L. (1993).Comprehensive meta-analysis of integrity testvalidities: Findings and implications for person-nel selection and theories of job performance.Journal of Applied Psychology, 78, 679–703.

Ones, D.S., & Viswesvaran, C. (1998). Gender, age,and race differences on overt integrity tests: Re-sults across four large-scale job applicant datasets. Journal of Applied Psychology, 83, 35–42.

Pfeffer, J., & Sutton, R. (2000). The knowing-doinggap. Boston: Harvard Business School Press.

Rogers, E. (1995). The diffusion of innovations (4thed.). New York: The Free Press.

Rynes, S.L., Barber, A.E., & Varma, G.H. (2000).Research on the employment interview: Useful-ness for practice and recommendations for fu-ture research. In C.L. Cooper & E.A. Locke(Eds.), Industrial and organizational psychology:Linking theory with practice (pp. 250–277). Ox-ford: Blackwell.

Rynes, S.L., Schwab, D.P., & Heneman, H.G. (1983).The role of pay and market pay variability in jobapplication decisions. Organizational Behaviorand Human Performance, 31, 353–364.

Schmidt, F.L., & Hunter, J.E. (1998). The validity andutility of selection methods in personnel psychol-ogy: Practical and theoretical implications of 85years of research findings. Psychological Bulle-tin, 124, 262–274.

Slovic, P., & Lichtenstein, S. (1971). Comparison ofBayesian and regression approaches to the studyof information processing in judgment. Organi-zational Behavior and Human Performance, 6,649–744.

Terpstra, D.E., & Rozell, E.J. (1993). The relation-ship of staffing practices to organizational levelmeasures of performance. Personnel Psychology,46, 27–48.

Terpstra, D.E., & Rozell, E.J. (1997). Sources of hu-


man resource information and the link to orga-nizational profitability. Journal of Applied Behav-ioral Science, 33, 66–83.

Ulrich, D., Brockbank, W., Yeung, A.K., & Lake, D.G.(1995). Human resource competencies: Am em-pirical assessment. Human Resource Manage-ment, 34, 473–495.

U.S. Department of Labor. (1993). High performancework practices. Washington, DC: Author.

Welbourne, T.M., & Andrews, A.O. (1996). Predict-ing the performance of initial public offerings:Should human resource management be in theequation? Academy of Management Journal, 39,891–919.

ENDNOTES

1. The authors gratefully acknowledge funding andadministrative support from the SHRM Founda-tion, as well as helpful comments from DebraCohen, Herb Heneman, Ken Pearlman, LiseSaari, Tom Stone, Ray Weinberg, and GaleVarma. However, the interpretations, conclu-sions, and recommendations are those of theauthor(s) and do not necessarily represent thoseof the Foundation.

2. Two items were eliminated due to ambiguouswording, one due to the discovery of new cri-tiques of existing research, and one to the dis-covery of new findings.

3. There are two kinds of integrity tests: some overtlyask about attitudes and behaviors related to theft,absenteeism, violence, and the like; others are“disguised-purpose” personality tests that tapthree of the “Big Five” personality dimensions(Conscientiousness, Extraversion, and EmotionalStability).

4. As indicated earlier (see Table II), there was ex-

treme restriction of range on this variable, whichalmost certainly reduced its observed correlationwith overall knowledge.

5. We transmitted our findings with respect to theresearch–practice gap concerning intelligence toa number of HR practitioners, and asked for anexplanation. The most common response was thatmany HR practitioners think ability testing islikely to lead to legal difficulties. As such, it isimportant to emphasize that we did not ask re-spondents what they believed about the legalityof ability testing. Rather, we asked (in a varietyof ways) whether they thought that ability was abetter, or weaker, predictor of performance thanother less-predictive characteristics (e.g., valuesor conscientiousness). Clearly, practitioners’negative beliefs about intelligence go far beyondconcerns about legal defensibility.

6. It is possible that these last two reasons—dis-agreement with a known research finding ordisagreement with its practical implications—may account for some of the gap uncoveredin this study. Specifically, because we askedabout practitioners about their beliefs, it ispossible that some practitioners stated nega-tive beliefs about intelligence despite know-ing about intell igence-related researchfindings. However, we believe the numbers ofsuch respondents are probably very small, fortwo reasons: first, fewer than 1% reportedreading HR research journals and, second,only one respondent noted in the margins ofthe survey that he or she “knew” the researchfinding, but disagreed with it. Given morethan 37,000 opportunities for this to occur(959 subjects � 39 items), we suspect thatthe proportion of the gap explained by dis-agreement with known findings is negligible.

hr professionals' beliefs about effective human resource practices

Documents