Chapter 5: Measurement Concepts

1
Chapter 5Measurement Concepts
2
Reliability of Measures

• Reliability refers to the consistency or
  stability of a measure of behavior.

If you weighed yourself now and then at the end
of class and you weighed the same both times, you
would say the scale is reliable.
?
3
Reliability of Measures (cont)

• Any measure True score Measurement error
• True score real score on the variable
• Reliable scores have considerably less
  measurement error

4
Reliability of Measures (cont)
5
Reliability of Measures (cont)

• How can we assess reliability?
• Correlation coefficients called regression
  coefficients
• - Pearson Product-Moment Correlation coefficient
• Symbolized as r

6
Reliability of Measures (cont)

• Coefficients range from 0.00 to - 1.00 and
  0.00 to 1.00
• Sign of the coefficient indicates direction
• Value of the coefficient indicates the strength

7
Reliability of Measures (cont)
8
Reliability of Measures (cont)

• To assess the reliability of a measure, need at
  least two
• scores on the measure from many
  individuals
• Methods of accessing reliability
• Test-retest reliability
• Internal consistency reliability
• 3. Interrater reliability

9
Reliability of Measures (cont)

• Test-Retest Reliability
• Assessed by measuring the same individuals
• two points in time
• Reliability established when the two scores are
  very similar
• Highly similar scores are close to 1.00

10
Reliability of Measures (cont)

• Internal consistency reliability
• Assessment one point in time
• All items should yield consistent results
• 1. Split-half reliability correlation of
  individuals
• total score on one half of the test with the
  total
• score on the other half of the test
• 2. Cronbachs alpha correlation of each item
• with every other item

11
Reliability of Measures (cont)

• Interrater Reliability (a/k/a INTEROBSERVER
  AGREEMENT)
• Correlation between the observations of raters
• - A reliable measure must show a high agreement
  between raters or judges

12
Construct Validity of Measures

• Construct validity
• refers to the adequacy of the operational
  definition of
• variables
• The degree to which the measurement or
  manipulation of the variable accurately reflects
  the underlying theoretical construct
• - Is the measure that is used actually
  measuring the construct it is intended to measure?

13
Construct Validity of Measures (cont)

• Indicators of construct validity

14
CRITERION VALIDITY

• The degree to which a measurement device
  accurately predicts behavior on a criterion
  measure
• A paper-and-pencil measure of leadership ability
  predicts actual leadership behavior in a group

15
CONVERGENT VALIDITY

• Do scores on the measure relate to other measures
  in expected ways?
• Example do people with high self-efficacy
  predict that they will perform better on a task?
  If so, this would be evidence for the construct
  validity of the measure.

16
DISCRIMINANT VALDITY

• The measure of the variable is NOT related to
  other variables that it theoretically should not
  be related to.
• E.g., scores on the self-efficacy measure are not
  related to reaction time

17
Reactivity of Measures

• Reactivity is a potential problem
• A measure is reactive if awareness of being
  measured changes an individuals behavior
• - Fails to provide an accurate measurement of
  the behavior under natural circumstances
• (Reactivity ADDS ERROR TO SCORES!)

18
Reactivity of Measures (cont)

• Minimize reactivity by
• Allowing time for individuals to become used to
  the presence of an observer or the recording
  equipment
• Use nonreactive or unobtrusive measures

19
Variables and Measurement Scales

• Remember that one of the goals of science is to
  identify relationships that exist among variables
  in our world
• To discover these relationships, we need to
  identify certain characteristics or quantities
  for the variables we are studying (This process
  is called MEASUREMENT)
• More

20
Variables and Measurement Scales

• So whats a scale?

21
Variables and Measurement Scales

• Scale refers to a set of criteria
  (guidelines) for how something is being
  measured.
• Psychologists use 4 basic scales of measurement
  (see next slide)

22
Variables and Measurement Scales

• 1. Nominal Scales
• People are measured according to whether they
  belong in certain categories
• Nominalbased on a name
• No numerical or quantitative properties
• Different levels of the variable consists of
  the categories youve defined
• more

23
Variables and Measurement Scales

• Nominal Scales
• Independent variables are often nominal
  variables
• Effects of gender on math abilities
• Effects of ethnicity on IQ scores
• Effects of religious affiliation on attitudes
  about the death penalty

24
Variables and Measurement Scales (cont)

• 2. Ordinal Scales
• Rank orders the levels of the variable
• Persons or things being measured are ordered
  first to last, best to worst, highest to lowest,
  biggest to smallest, etc.
• Often used to assess peoples needs, preferences,
  attitudes
• More

25
Variables and Measurement Scales (cont)

• Ordinal Scales
• This scale is Quantitative (that is, you can
  measure an amount of something) but no real
  number values are used to compare the
  people/things being measured
• (think about beauty scale people use or movie
  ratings using number of stars)
• More

26
Variables and Measurement Scales (cont)

• Examples of Ordinal Scales
• School teacher Line up in size order!
• Doctor in emergency room Ive listed the order
  for patients to be treated by severity of
  illness!
• E channel Tune in Friday to see the countdown
  of your favorite Hollywood couples!
• Dr. Reeve For the following, please rank in
  order of importance to you as a Psy major more
  varied courses, more course sections, more
  research opportunities, more library resources,
  more guest speakers (etc.)

27
Variables and Measurement Scales (cont)

• 3. Interval Scales
• Differences between the numbers used to measure
  are meaningful
• Intervals are equal in size (IQ differences
  between 90 and 100 is the same amount as
  difference between 130 and 140---both are 10
  points difference)
• Quantitative, but no meaningful zero reference
  point (the test used to measure someone has no
  0 value----think IQ tests and personality tests)

28
Variables and Measurement Scales (cont)

• 4. Ratio Scales
• Most detailed and informative scale of
  measurement!
• Generally has to do with any measurement system
  in which you are counting up a quantity
  (cigarettes smoked, requests made by a child,
  number of students who participated in
  classhint, hint!)
• Lets you determine RATIO differences
• Unlike interval scale, this has an absolute zero
  reference point

29
Identify the measurement scale for each of the
following variables

• Circle your marital status Married Single
  Divorced Engaged
• Are you employed? Yes No
• If you work, how many hours a week do you work?
  _______
• Rate the extent you enjoy college on the scale
  below.
• 1 2 3 4 5
• Not Very Much
  Very Much
• What is your class standing?
• Freshman Sophomore Junior
  Senior
• 6. What is your IQ score (pretend you know it)

30
The End