MGTO 231 Human Resources Management

1
MGTO 231Human Resources Management

• Personnel selection I
• Dr. Kin Fai Ellick WONG

2
Questions for you

• The objective of selection is very simple
• Choosing the correct person
• However, we do not have magic, there are often
  many incorrect choices in our decisions
• A more realistic objective for selection seems to
  be
• Minimizing the probability of making an incorrect
  selection
• How?
• I hope you will get some insights at the end of
  this class

3
Outline

• In Personnel selection I
• The basic concepts
• Measurement errors
• Reliability
• In Personnel selection II
• Validity
• Selection tools
• Letter of recommendation, ability tests,
  personality tests, interviews, handwriting
  analysis, etc.

4
Outline
5
Selection as prediction

• I want to have employees with good performance
• I cannot know their performance before hiring
• I can just predict from applicants various
  characteristics
• Some individuals are more likely to have better
  performance than others (remember the idea of
  individual differences)

6
What are the predictors

• Any characteristics associated with individual
  differences
• Age, gender, IQ, EQ, interpersonal skills,
  education level, openness, extroversion, academic
  performance, etc.
• Some may be better predictors than others (more
  reliable and valid)

7

• All selection methods are to measure these
  characteristics and the associations between
  these characteristics with different aspects of
  performance
• Selection ? individual characteristics ?
  performance
• Tools ? predictors ? performance

8

• Two major issues in selection
• How consistent is the prediction (reliability)
• How well is the prediction (validity)
• Is it really measuring what it is supposed to
  measure?

9
An example of a good and a bad predictor
Job Performance
Job Performance
Y
Y
Academic Results
Interviews rating
Which one has positive relationship with job
performance?
10
(No Transcript)
11
Outline
12
Measurement errors

• Please answer questions 1 and 2

13
Measurement errors

• Error is not equal to mistake
• Error is a kind of inaccuracy and variability
• Error can be reduced if the measuring device is
  precise, but it cannot be completely eliminated

14

• Observed score
• True score systematic errors random errors
• Systematic errors
• A constant is added to every measure
• Observed score True score constant
• Random errors
• A random value is added to every measure
• The random values are normally distributed, with
  mean 0
• Observed score True score random value

15
Sources of errors

• Systematic errors
• Basic characteristics of the measuring device
• Random errors
• Physical
• Temperature, humanity, loudness, brightness, etc.
• Psychological
• Fatigue, motivation, adaptation to new
  environments
• Sampling error
• domain sampling model

16
Outline
17
Reliability (???)

• Definition
• (Textbook) The consistency of scores obtained by
  the same person when retested with the identical
  or equivalent tests

18
Measurement errors and reliability

• Which kind of errors, systematic or random, may
  affect the reliability of a measurement?

19

• The purpose of testing
• Assessing individual difference
• The influence of systematic errors on this
  purpose is, everyones score is biased by a
  constant. The relative performance is unchanged
• Low T always gets low X, and high T always gets
  high X
• The influence of random errors
• Low T may obtain high X, and high T may obtain
  low X
• The probability of Low T getting high X increases
  as random error increases

20
What is a reliable measurement?

• A test that can truly reflect the True score is
  claimed as a reliable test
• For a perfectly reliable test
• The correlation between the true score (T) and
  the observed score (X) is 1
• X can be fully accounted by T
• For a perfectly unreliable test
• The correlation between T and X is 0
• T accounts for nothing about X

21

• Tests that are relatively free of measurement
  (random) errors are deemed to be reliable
• There is much less concern with systematic errors
• The term errors usually refers to random errors

22
Which one is more reliable?

• Tests that are relatively free of measurement
  random errors are deemed to be reliable
• Reliability refers to consistency of measurement
  or repeatability

Trial Thermometer A Thermometer B Thermometer C
1 101 103 108
2 99 97 108
3 100 102 107
4 101 103 108
5 99 95 108
23
Methods measuring reliability

• Test-retest
• Parallel/alternate forms
• Split half
• Alpha (internal consistency)

24

• How do you interpret this picture?
• Please tell me a story about the picture
• Is this test reliable?
• Why?

25
Conclusion

• Ive asked
• A more realistic objective for selection seems to
  be
• Minimizing the probability of making an incorrect
  selection
• How?
• Step 1Using test with high reliability
• Checking the reliability before and after using
• Ill discuss the other step about validity in
  next lesson