4
&f SMU Publication: Human Resources Date: April, 2005 Headline: Can you trust pyschometric testing? The debate over the valw and validity of psychometric testing rages on, with companies such as Dryden Weatth Management vouching that the tests provide more holistic and accurate assessments of emQloyees. By David Ragdale By the time most people leave school, they've given up on the idea of standardbed testing. As one embarks on a professional w e e r . be or she bids farewell to A-levels. 0-Levels. medicat board and bar exams... or so they think. Whereas in the past. CVs. recommendations, and one-on-one interviews were standard for job applicants, h e y are incressln@y belng augmented by job and company specifUc sssessments during the application process. In the last decade, psychometrCc rest- ing has emerged as a vital tool for human resources spedatists fn staU recruitment, IraMng, and retenttw. WMle lhe tenn conjures images of mad scientists fiddhg wllk enor- The keys to successhrlty iWDkIneming testing urow F usychometric r I selection of tests, extensive training of HR managers, and correct 'norming' of assessments mow metallic machines, if properly inculcated into an organisation's culture. psychometric testing can cut costs and provide greater feed- back to employees and employers. The keys to successRrlly irnplement- ing psychometric testing as an essential component of an HR strat- ~gy are threefold: the proper selec- ion of tests. the extensive training of HR managers, and the correct 'norming' of assessments. - -. - Psychometric tests (PTs) are assess- ment tools developed to measure mental ability and personality traits. They are especially invaluable for HR departments in personnel selec- tion. according to Dr Gary Gregurss,

&f SMU Date: Headline: Can you trust pyschometric testing? · vide immediate ROI because people fail to consider the costs. for example. of recruiting new employees because of turnover

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: &f SMU Date: Headline: Can you trust pyschometric testing? · vide immediate ROI because people fail to consider the costs. for example. of recruiting new employees because of turnover

&f SMU Publication: Human Resources Date: April, 2005 Headline: Can you trust pyschometric testing?

The debate over the valw and validity of psychometric testing rages on, with companies such as Dryden Weatth Management vouching that the tests provide more holistic and accurate assessments of emQloyees. By David Ragdale

By the time most people leave school, they've given up on the idea of standardbed testing. As one embarks on a professional weer . be or she bids farewell to A-levels. 0-Levels. medicat board and bar exams... or so they think. Whereas in the past. CVs. recommendations, and one-on-one interviews were standard for job applicants, h e y are incressln@y belng augmented by job and company specifUc sssessments during the application process. In the last decade, psychometrCc rest- ing has emerged as a vital tool for human resources spedatists fn staU recruitment, IraMng, and retenttw. WMle lhe tenn conjures images of mad scientists fiddhg wllk enor-

The keys to successhrlty iWDkIneming

testing u r o w

F usychometric r

I selection of tests, extensive training of HR managers, and correct 'norming' of assessments

mow metallic machines, if properly inculcated into an organisation's culture. psychometric testing can cut costs and provide greater feed- back to employees and employers. The keys to successRrlly irnplement- ing psychometric testing as an essential component of an HR strat- ~ g y are threefold: the proper selec- ion of tests. the extensive training

of HR managers, and the correct 'norming' of assessments.

- -. - Psychometric tests (PTs) are assess- ment tools developed to measure mental ability and personality traits. They are especially invaluable for HR departments in personnel selec- tion. according to Dr Gary Gregurss,

Page 2: &f SMU Date: Headline: Can you trust pyschometric testing? · vide immediate ROI because people fail to consider the costs. for example. of recruiting new employees because of turnover

SMU Publication: Human Resources Date: April, 2005 Headline: Can you trust pyschometric testing?

A rsvchometrlc assessment "allows yw to compare the nersonallty traits highlighted in the test with the behaulours mu have obsewed in the Intenriews."

associate professor of organisa- tional behaviour at the Singa- pore Management Unlverslty, as they can "e!Eciently and effectively help organisations predict employee behaviour sueh as job performance, theft and turnover." These tests do not stitad h e , however, says Mike Dembo, director of human resources for Astrazeneea inc. a pharmaceutical company hsed in the UK with 64.00 employees worldwide, in that "mostly. the results confirm what we already know from experience or int8Nfews."

Psychometric testing can be broken down into three broad categories: ability, aptitude, and personality. Ability tests examine the ability of a person to learn and apply new con- cepts and skills. This b a for- ward-looking test and so differs *om tests of aualmenx, whleb measure pre-exist'mg skilb and knowledge sets. Most school exams are retrospective in that they assess knowledge we already have as opposed to skflls we had or could develop. In an ability test, it is not what you know but what you are able to learn. process and act on that is important. AbWy testing can include a variety of t ~ 8 I that measuw intelhgence on multiple l e e k such as ver- bal and numerical. Aptitude tests are often combined with ability tests, but the former tend to be more Job-spdc, whereas ability tests assess more general skills.

Personality assessments are the mart difllult to d e h e and norm as they are specific to the company's culture, says Carl Pollock. director of human

resources at Dryden Wealth Management Ltd. In Asla, the inlernatiobal private client business of Prudential Financial, lac. of the United Scam*, believes that 'person- alty profUas are not necesmi- ly oriented to directly identify if someone is quaiifled or no3 qualsaed to perform job duties. Aptitude or ability tests can often be more useful in thb area. That safd, psychometric

adjust its norms for cultural differences? The short answer is. p e m t y is not the same a$ culture in that personality is not 'what" we relate to the our- side world (valuedpreeancep- tiom) bur more the *wayo In which we relate to others.

The very name 'psychomet- ric testing" elicits concern and anxiety. Indeed, many HR pro- fessionals, such as Pollock, pre- fer ta avoid rbe "t" word alm-

profiling helps you understand personality traits that are we- Pul in determining if the Indi- vidual will flt with the otgani- sation's cultural team and how they might cope with the rou- tines and pressures of a given role." Personalily tern are an indicator of how a pemn relates to others In varfous social and business situations. The tricky question behind per- sonality teas k, how does HR

gether. 1 don't like the term 'test*, there is no right or wrong when it comes t per- sonality.- Concern and anxiety are expressed by some when taking these assessments - especially by applicants unfa- miliar with the process. *It's imporant to explain ro the caa- didate why and how you use Pfs as well as the imtructions you want them to follow. We always provide feedback,

wMch again Is one-to-one as it provides a better environment to discuss the results and can- didam' reactions."

Organisations are inereas- in& turning to psychometric testlng, espedally in candidate recruitment. as they furnish a more complete profile of appl. cants. Pollock finds they pro- vide 'additional information to twist in the selection and deci. sion making process." A psy- chometric assessment 'allows you to compare the personality traits hightighted in lhe test with the behaviours you have observed in the interviews ... correlations and ctkconnects." PTs, if administered before an interview, can also help to foeus on a prospective candidate's strengths and weaknesses. This idea ofthe 'fuller picture" is an important coneept espe(ially In a competltlve Job market. An applicant may have great tee- ommendations and quamca- ~~QIS. and interview well. but do these make him or her an ideal fit for yaw orgedsution? The more information an organbation has about a Job applicant, the more likely it is to make a correct him deei- sW. Greguras says, "It is much more diHlcuIt to generate ream sons why organtsstions would not use, rather than we, psy- chometric testing for personnel seleethn. If the system Is well designed and all other factors held constant, those that use PI'S will outperPorm those that select employees randomly."

Psychometric testlog is also beneficial because of its anlrlpl- ical aspect. Greguras is of the opinion that PTs, when usgd correctly, can predict an exam-

"If the system Is well designed and all other factors held constant, those that use PTs will omerfmn lhoae lhal select employees randomlv."

Page 3: &f SMU Date: Headline: Can you trust pyschometric testing? · vide immediate ROI because people fail to consider the costs. for example. of recruiting new employees because of turnover

SMU Publication: Human Resources Date: April, 2005 Headline: Can you trust pyschometrii testing?

The more detallad lwyenomeaic took can take around mte8 hours of an II prefasslonal's time to almlnlstet review, score and feedbaclr. this Is c d j . "

inee's job and performance or training potential rar more accurately than any other selection method. "PTs ate used to predict who win and will not behave in a eerldin manner," he says. 'Generally the crfterton of interest is job performance, that is. we are interested in h h g those who will perform at ihe highest level on the job in W n n ~ of sales YOI- ume and number of unib pro- duced. Psychometric testing can be dewkped and used to predict oirfually anjWng i h a the orgnnbatbn is interested iu prediclhg.^ In addition, PTs are used in awessing a can&- date's comfort wlth the prevail- ing culture within an organisa- tion. This is a v i m imporlant &tor for HR directors like Pollock. 'psychometric profil- ing helps you Imdersund per- sonality traits that in turn are useful in determining if the individual will Bt with the organfsatlon's culture and how lhey might cope with rantfnes and pressures of a @en role."

Psychometric testing is not used merely tor recruitment purposes. Greguras belleves they can perform a vaiupbie serpide in training and evaluat- ing staff. 'After people are hired, you may use testhrg to decide who gets promoted, who needs training in what compe- te& and to assess leader- ship potential or develop lead- ers." Obviously for HR execu- Uves such as Pollock, dilbrent groups need Werent types of PTs. *In recruitment and s e k - tion. I prefer to use detailed PTs because they provide more information. For development and team bul ld i , I tend to use the took that gtve the broad headlines." says Pollock.

Ashazeneca's Dembo 'al- wags uses verbal. numertc. and

abstract reasoning (PTs) for top-level or middle-tellel man- agers using local benchmark data." A sound psychometric assessment strategy presup- poses the use of these tests throughout the organisation, s t sP1 levels. This ensures that employees are fmllier with PT9 and accept their utUlty for not only the organisation. but also for their own development and growth.

PTs are usually purchased ftom an outside vendor. Dembo prefers online rssessment tools vendm such as SHL Pollock belleves this is cost eRedive because 'developing such pro- filing tooh is complex and time consuming if done correctly. Most organisations do not have the h e or Lhe resources to develop these mIs in house. The carts d purchasing the tests vary, according to P O W , 'Some proBUng toals are very detailed and others look only a t high level pemnrllty traits, the latter tend to be less expensive. easier to admInlster end pro- vide feedback." A major expense for an organisation is the time and resources dedicat- ed to training M staff and administering the tests. Becau4e of the investment involved, Pollock must decide

when a paychometrfc mess- ment is appropriate. T h e more detailed psychometric took can take around three hours of an HR professional's time to administer, review, score and feedback. this is costly. Hence we use them selectively. Some psychometrics offer soRware which helps with admiasstra- Uon and feedback, bowever. the liceuces can be cmtly.

y.rry-- Greguras belleves that Ihe return on investment (ROO for psychomsule @sting ls Imme- diate and far-reaehtng. 'Be- cause valid assessments pre- dict employee behaulours. the R01 b e e s as soon as the Indi- vidual enters the organbation. Ask any manager who has hired an employee who is chronically late, careleas. or unmotivated and he or she will tell you that the differences between high perfonhug and tow performing employees are inunedlstely evklent and kmme- Giaiely impact on an organh- tion." Creguras thinks there is a mlsconceptlon that vald selection sgstems do dot pm- vide immediate ROI because people fail to consider the costs. for example. of recruiting new employees because of

turnover. the costs associated with employee theft, or the costs associated with poor job perfomance - dl of which influence an organbation's bot- tom h e . 'We often do not have a spreadsheet for tbse costs, yet, they d e c t an organ&- a n ' s profitability. How much does bad advice given to a cus- tomer cost an organisation? Again. we do not have a spreadsheet f o r t h . but n mat- ters," he says.

QuantIfLing ROI b dimcult but Greguras points to the Ueawrement Error in Re- search on Human Resources and Flrns Pqfommcd study. publfshed h 2000, which sug- gested that '6rms that use higb performance HR practices have accounting profits or cash flow that are approximately 20% higher on average than eompa- nies utiLising fewer hi& perb tormanee HR practices." Psy- chometrlc testing can also erea ate efBcien& as they anow managers to standardbe sssessments throughout their qanisations. This positive ROI for psychometric testing L not always acknowledged, as PTs are c d e r e d by game to be costly, *cold" and burdensome. Greguras references a n ~ t h e r study. known as The fib!&# of Utility Analyls. wMch lndl- cates chat 'ulfllrp estimates indicating a substantial net gain for ~ ~ t f o n s If they implemented a selecttoo sys- tem actuaIIy decrease man- agers' support for the imple- mentatson of s u e a system." This raises an hportant ques- tion; do nm-HR manegers realbe the ROI benefits of p y . chometric testing?

Ulttmately, the positive ROI for psychometric testing is obvious. Organbatlons spend bitlions of dollars wry year on recruiting. Studles show thal people with highend skills and

Page 4: &f SMU Date: Headline: Can you trust pyschometric testing? · vide immediate ROI because people fail to consider the costs. for example. of recruiting new employees because of turnover

SMU 1

Publication: Human Resources Date: April, 2005 Headline: Can you trust pyschometric testing?

"OtmRlSatlWI~ that use PTs must haw a dear stratm and pesition an W and 'LOW they use than, dadatatlens and boundaries".

abllilies require less invest- ment and training. and provide more *bang lor the buck" for their employers. It makes per- fect sense to have the ability to accurately IdentllS. these Mgh- er-level p e d m e r s prior to any job offer. Psychometric testing provides organisations with the tooh to make this assessment.

Organlsatioas must avoid the temptation of purchasing PTs and using them in a haphazard manner. 'Being unclear about why and how you uss PTs is a major pitfall,' says Pollock. 'Orgaisations that use PTs must have a clear strategy and position on 'why' sad 'how* they use them, declarations and boundaries - I recommend that they have a clear policy statement on (be subjecr." Greguras believes that before adopiing the use of PTS. an orgabation must ask itself the question, "What are we trying to aceompUsh by using psycho. metric tasting?" Training is essential for HR sta& Edlaeat- h g wn-HR employees on the benefits and procedures assmi- ated mith psychometric testlng L also a smart idea. While HR staft will have the mast in- depth knowledge and inthnacy with PTs. demysttfying this tool for thu entire organisation is a key element ip suc~essful implementation. A job candi- date will fed more at ease when taking a PT if the entire organhalion fmk comfortsble with and appreciates the process. On the other hand. U staff autside HR are unfamiliar and diStmtfufUI of PPs, then thfs confusion can spread to ap* cants and po&bly skew the results. For Gregunrs. the devil is in the details, *Once an orgknisatlon bas identified the purpose for PT, the eext step is ta identi@ approprtate assess- ments or profession& to devel- op appropriate assessments."

Another potential pielall whep using psychometric test- ing is h reviewing the assess- ments. Test scores are deter- mined by norms, and these norms differ from job to job. It is importan1 lor an o r g a s h - Uon to w~ider this when scor- Ing the tests. For Pollock, ?tit's important Lhat the norm groups are carehlly considered and selected. Respected and rep- utable PT specialist organism- tioas give you access to a range of norm tables b a t can be very specific combining factors such as sex, nationality, age and industry." If an organlsation is huge enough. it is sensible to develop its w n norm tables. However. these individualised norms are not for everyone, adds Pollock. 'Some PT s p cialbt firms will work with client companies to build spe- d& company norm groups, bur you'll typically need the data of a large number of company empbyees to provide robust and representative norm groups. This leads many corn. panies to simply select PT from specialist firms which have norm groups representative of the Indusiry in which they operate." Most organisations b a d in Singapore that have both foreign and iocal employ- ees must tailor the norms used for scoring PTs to refiwl the diversity of its workR~m.

PT5 are a40 time-sersitive. Faillng to provide timely feed- back not only u o & d m its utility but also perpelttales con- fusion and &(rust about them. In his own experience, Pollock IIDds that these tests are "prob- ably only reliable far an 18- month perlod." For stsfl train- in$ and management s u m - sioa these asseslvments are not a 'one shot deal." Employees grow and change as do organi- sstions. and so PTs must con- stantly updated, re-adminis-

tered. and recalibrated to get the most accurate results.

OanrtJmtMameHR Psychometric testfng requkes s stgnificant investment born organbations that lmplemenl then. This cost cannat. howev- er, k borne by HR alone. In order to achieve positive ROI. the entire s t a r must under- stand and accept psychometric teslhg as an integral pert or a company's growth strategy. Before adopting psychometric testing, an orgdsation must answer "why". "what", and "how', 'why are we using PTs?" 'What tests are we going to use and what do rbey look like?" and anally. 'how are we going to [mplement psychamt- FIC testing in our organisa- EIMI?" When psychometric test-

ing becomes part and parcel of an organisallon's culture. truly and deeply ingrained, confu. sion and diirust substdes. A healthy part of this process is the Inclusion of non-HR s U in discussions surrounding psy- chometrie testing. AfTording ckear and upfkont inPormatbon to all partie6 involved is a h vital. Psychometric testing is a tool char should be embraced for providing a p~sitive ROI for personnel decisions. or as Pollock puts it, "an informed decision is typically a good decision".

* W e n t i 0 1 Finandat, IRC. oJ the United Sates is not a&Wat- Bd with PncdenLid plc, which is Beadparteted In bke United KiRgdOlR