27.02.2018 Views

HRM textbook

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

CHAPTER 6 EMPLOYEE TESTING AND SELECTION 177<br />

Here s how it works. 9 When a prospective employee applies for a job, his or her<br />

information (such as school and previous employers) goes into Google s applicanttracking<br />

system (ATS). The ATS then matches the applicant s information with that<br />

of current Google employees. When it finds a match, it asks those Google employees<br />

to comment on the applicant s suitability for the position. This helps give<br />

Google recruiters a valuable insight into how the Google employees actually doing<br />

the work think the applicant will do at Google. And it supports Google s strategy,<br />

by fostering a sense of community and interaction among Google employees,<br />

who see themselves working together to select new Googlers.<br />

1 Explain what is meant<br />

by reliability and validity.<br />

BASIC TESTING CONCEPTS<br />

A test is one popular selection tool. A test is basically a sample of a person s behavior.<br />

Using a test (or any selection tool) assumes the tool is both reliable and valid.<br />

Few things illustrate evidence-based HR the deliberate use of the best-available<br />

evidence in making decisions about the human resource management practices you<br />

are focusing on as do checking for reliability and validity.<br />

Reliability<br />

Reliability is a test s first requirement and refers to its consistency: A reliable test is<br />

one that yields consistent scores when a person takes two alternate forms of the test or<br />

when he or she takes the same test on two or more different occasions. 10<br />

Reliability is very important. If a person scores 90 on an intelligence test on a<br />

Monday and 130 when retested on Tuesday, you probably wouldn t have much faith<br />

in the test.<br />

You can measure reliability in several ways. One is to administer a test to a group<br />

of people one day, readminister the same test several days later to the same group, and<br />

then correlate the first set of scores with the second (test-retest reliability estimates.) 11<br />

Or you could administer a test and then administer what experts believe to be<br />

an equivalent test later; this would be an equivalent or alternate form estimate.<br />

The Scholastic Assessment Test (SAT) is an example. Or, compare the test taker s<br />

answers to certain questions on the test with his or her answers to a separate set<br />

of questions on the same test aimed at measuring the same thing. For example, a<br />

psychologist includes 10 items on a test believing that they all measure interest in<br />

working outdoors. You administer the test and then statistically analyze the degree<br />

to which responses to these 10 items vary together. This is an internal comparison<br />

estimate. (Internal comparison is one reason that you find apparently repetitive<br />

questions on some test questionnaires.)<br />

Many things cause a test to be unreliable. These include physical conditions<br />

(quiet tests conditions one day, noisy the next), differences in the test-taker (healthy<br />

one day, sick the next), and differences in the person administering the test (courteous<br />

one day, curt the next). Or the questions may do a poor job of sampling the material;<br />

for example, test one focuses more on Chapters 1, 3, and 7, while test two focuses<br />

more on Chapters 2, 4, and 8.<br />

Because measuring reliability generally involves comparing two measures that<br />

assess the same thing, it is typical to judge a test s reliability in terms of a reliability<br />

coefficient. This basically shows the degree to which the two measures (say, test score<br />

one day and test score the next day) are correlated.<br />

negligent hiring<br />

Hiring workers with questionable backgrounds<br />

without proper safeguards.<br />

reliability<br />

The consistency of scores obtained by the<br />

same person when retested with the identical<br />

tests or with alternate forms of the same test.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!