[IPAC-List] Candidate Reduction Strategies

[Karen Coffee]

In California many agencies have a long history of only accepting a
predetermined number of applications.  What Lance proposes would also work
in Calif.  A note of caution, in this state an informal Attorney General's
opinion many years ago suggested that random selection before any screening
device would violate the state constitution which requires competition based
on competitive exam.  Random selection absent the preceding use of any
selection process has been deemed competition solely based on chance.  
Karen Coffee

-----Original Message-----
From: ipac-list-bounces at ipacweb.org [mailto:ipac-list-bounces at ipacweb.org]
On Behalf Of Lance Seberhagen
Sent: Monday, March 12, 2012 12:43 PM
To: Demooy, Adele
Cc: 'ipac-list at ipacweb.org'
Subject: Re: [IPAC-List] Candidate Reduction Strategies

Hi Adele:

The best way to reduce the number of applicants to be tested is to use
random selection.  For example, assume 3,000 people apply for the job, and
you want to administer the written test to a maximum of only 1,000
applicants.  Here are the steps that I would recommend:

1.  Screen the 3,000 applications for minimum qualifications to screen out
the obviously unfit applicants.  (Most employers do this as part of the
online application process.)  Let's say that 2,500 of the original
3,000 applicants pass this step.

2.  Assign a unique code number in consecutive order (1-2,500) to each of
the remaining 2,500 applicants.

3.  Use a random numbers generator (e.g., RANDBETWEEN in Excel) to list the
2,500 numbers in random order.

4.  Select the first 1,000 random numbers (applicants) to invite to the
written test.  Place the remaining 1,500 random numbers (applicants) on a
waiting list, or discard them.

By definition, random selection gives each applicant the same probability of
being selected.  Therefore, the method is objective, fair, and
nondiscriminatory, provided that you always assign code numbers to
applicants BEFORE you run the random numbers generator each time.  However,
by definition, random selection also has zero validity (or job-relatedness).
Therefore, there could be an issue if Civil Service Rules or an agency's
Charter require all selection procedures to be merit-based, but most
agencies should be able justify random selection for part of the selection
process as long as the overall process is based on merit.

Approximations of random selection (e.g., date of application, birth date,
Social Security Number, alphabetical order of last name) should not be used
because these methods have zero validity AND they may have adverse impact
due to hidden biases.  Therefore, they could be discriminatory.  There is
also the possibility of cronyism if applicants are selected by date of
application and "insiders" can give advance notice of job opportunities to
their friends and relatives.

Lance Seberhagen, Ph.D.
Seberhagen & Associates
9021 Trailridge Ct
Vienna, VA 22182
Tel 703-790-0796
www.seberhagen.com **




On 3/12/2012 11:16 AM, Demooy, Adele wrote:

> For economic reasons we are looking into ways to reduce the number of

candidates who participate in exams.  One option we are considering is
accepting a predetermined number of applications based on statistics from
previous exam administrations.  I am interested in your feedback on this
approach or learning about any other methods you are using.

>

>

>

>

> _______________________________________________________

> IPAC-List

> IPAC-List at ipacweb.org

> http://www.ipacweb.org/mailman/listinfo/ipac-list

>

_______________________________________________________
IPAC-List
IPAC-List at ipacweb.org
http://www.ipacweb.org/mailman/listinfo/ipac-list