Measurement Validity and
Reliability for Research in
Audiology
 Validity- How to show that the test truly measures
what it claims.
 (Does it measure what it is suppose to measure)
 Reliability- Produces stable, consistent
measurements.
 (How representative is the measurement)
Validity
 On a test with high validity the items will be closely
linked to the test’s intended focus.
 Example: One common technique for demonstrating validity is
to use correlation.
 If test measures intelligence than it should related to IQ,
performance on learning tasks, and problem-solving ability. If
all test scores correlate then new test is valid
Several ways to estimate validity
 Content validity- related to objectives and their
sampling
 Construct validity- theory underlying the target
 Criterion related validity- concrete criteria in the real
world
 Concurrent- correlating with another measure already
validated
 Predictive- Capable of anticipating some later measure.
 Face validity- related to test overall appearance
Validity vs. Reliability
 Measures of validity are similar to measures of
reliability. With reliability, you compare one
measurement of a variable on a group of subjects
with another measurement of the same variable on
the same subjects. With validity, you also compare
two measurements on the same subjects.
Reliability
 Reliability is the extent to which an experiment or
test shows the same result on repeated trails when
the same individuals are measure under the same
conditions.
Tests that Validity vs Reliability Matter
 Acoustic Reflexes
 MRIs
 OAEs
 Vemp
 VNGs
 ABR
 ASSR
 EcochG
Reliability Example
 If your IQ was measured at 113 last week, Expect
nearly the same score if measured again this week.
Reliability and Error in Measurement
 Measured score = true score + error
 Example
 For an IQ test
 True score= actual level of intelligence
 Error = current mood, fatigue, general health, etc.
Reliability
 Equivalency: related to the co-occurrence of two
items
 Stability: related to time consistency
 Internal: related to the intruments
Practice Quiz
 Reliability refers to whether we are truly measuring
the concept of interest in our study.
• True
• False
Practice Quiz
 Reliability refers to whether we are truly measuring
the concept of interest in our study.
• True
• False
Practice Quiz
 Researchers always try to maximize reliability and
validity in their research.
• True
• False
Practice Quiz
 Researchers always try to maximize reliability and
validity in their research.
• True
• False
Practice Quiz
 The validity of a measure refers to the:
a. comprehensiveness with which it
measures the construct.
b. particular type of construct
specification.
c. consistency of the measurement.
d. accuracy with which it measures the
construct.
Practice Quiz
 The validity of a measure refers to the:
a. comprehensiveness with which it
measures the construct.
b. particular type of construct
specification.
c. consistency of the measurement.
d. accuracy with which it measures the
construct.
References
 Cardozo, R., Magdalena, J., Validity, Reliability and Practicality Retrieved from
http://www.slideshare.net/songoten77/presentation-validity-reliability on
November 30, 2009
 Gravetter, F., Wallnau, L. B., Essentials of Statistics for the Behavioral Sciences,
5th Edition, Thomson and Wadsworth, 2005
 Writing Guides Reliabiltiy and Validity, Colorado State University, Retrieved
from http://writing.colostate.edu/guides/research/relval/index.cfm on Nov.
30, 2009

Measurement validity and reliability

  • 1.
    Measurement Validity and Reliabilityfor Research in Audiology
  • 2.
     Validity- Howto show that the test truly measures what it claims.  (Does it measure what it is suppose to measure)  Reliability- Produces stable, consistent measurements.  (How representative is the measurement)
  • 3.
    Validity  On atest with high validity the items will be closely linked to the test’s intended focus.  Example: One common technique for demonstrating validity is to use correlation.  If test measures intelligence than it should related to IQ, performance on learning tasks, and problem-solving ability. If all test scores correlate then new test is valid
  • 4.
    Several ways toestimate validity  Content validity- related to objectives and their sampling  Construct validity- theory underlying the target  Criterion related validity- concrete criteria in the real world  Concurrent- correlating with another measure already validated  Predictive- Capable of anticipating some later measure.  Face validity- related to test overall appearance
  • 5.
    Validity vs. Reliability Measures of validity are similar to measures of reliability. With reliability, you compare one measurement of a variable on a group of subjects with another measurement of the same variable on the same subjects. With validity, you also compare two measurements on the same subjects.
  • 6.
    Reliability  Reliability isthe extent to which an experiment or test shows the same result on repeated trails when the same individuals are measure under the same conditions.
  • 7.
    Tests that Validityvs Reliability Matter  Acoustic Reflexes  MRIs  OAEs  Vemp  VNGs  ABR  ASSR  EcochG
  • 8.
    Reliability Example  Ifyour IQ was measured at 113 last week, Expect nearly the same score if measured again this week.
  • 9.
    Reliability and Errorin Measurement  Measured score = true score + error  Example  For an IQ test  True score= actual level of intelligence  Error = current mood, fatigue, general health, etc.
  • 10.
    Reliability  Equivalency: relatedto the co-occurrence of two items  Stability: related to time consistency  Internal: related to the intruments
  • 11.
    Practice Quiz  Reliabilityrefers to whether we are truly measuring the concept of interest in our study. • True • False
  • 12.
    Practice Quiz  Reliabilityrefers to whether we are truly measuring the concept of interest in our study. • True • False
  • 13.
    Practice Quiz  Researchersalways try to maximize reliability and validity in their research. • True • False
  • 14.
    Practice Quiz  Researchersalways try to maximize reliability and validity in their research. • True • False
  • 15.
    Practice Quiz  Thevalidity of a measure refers to the: a. comprehensiveness with which it measures the construct. b. particular type of construct specification. c. consistency of the measurement. d. accuracy with which it measures the construct.
  • 16.
    Practice Quiz  Thevalidity of a measure refers to the: a. comprehensiveness with which it measures the construct. b. particular type of construct specification. c. consistency of the measurement. d. accuracy with which it measures the construct.
  • 17.
    References  Cardozo, R.,Magdalena, J., Validity, Reliability and Practicality Retrieved from http://www.slideshare.net/songoten77/presentation-validity-reliability on November 30, 2009  Gravetter, F., Wallnau, L. B., Essentials of Statistics for the Behavioral Sciences, 5th Edition, Thomson and Wadsworth, 2005  Writing Guides Reliabiltiy and Validity, Colorado State University, Retrieved from http://writing.colostate.edu/guides/research/relval/index.cfm on Nov. 30, 2009