Download - Scales and Measures in Research
![Page 1: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/1.jpg)
Scales and Measures in Research
By
Dr. Anamika RayAssistant Professor in Mass Communication
Gauhati University, Assam, India
![Page 2: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/2.jpg)
THURSTONE SCALE
• Developed by L.L. Thurston.
• To measure direction and intensity or salience.
• Often used in psychology and education research.
![Page 3: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/3.jpg)
GUTTMAN SCALE
• Developed by Louis Guttmann in 1950.
• To measure intensity or commitment to an opinion, cause,
person, organization or course of action in a hierarchical way.
![Page 4: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/4.jpg)
FEELING THERMOMETERS
• Mainly used in political planning and research.
• Subjects more often than not find it preferable to
communicate to the thermometer and pick a number, but they
avoid very high and low ones.
![Page 5: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/5.jpg)
RANKING SCALES
• A mean for comparing two or more reference on one or more
quality.
• Subjects are presented with a list and told to put items on it.
![Page 6: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/6.jpg)
COMPOSITE MEASURES
• Simplifying data analysis by reducing the number of variables
relevant in a research work.
![Page 7: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/7.jpg)
QUALITY CONTROL
• One of the quickest way of research.
• Involvement of the examination of both questions and
response options to make sure they are unbiased.
![Page 8: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/8.jpg)
RELIABILITY
• Stable and dependable measure.
• Many times use helps to get similar result.
• Accuracy is easy to achieve.
![Page 9: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/9.jpg)
Concurrent Reliability: The reliability of a measure is
most often assessed by checking the strength and statistical
significance of its correlation with some other measure.
Alternate Forms Reliability: It is a variant of the basic
correlation technique.
TYPES OF RELIABILITY
![Page 10: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/10.jpg)
Split Half Reliability: Here the subjects are divided into two
groups then compares the scores on a measure for one group to
those for the other.
Test Retest Reliability: Here, we administer the same test to the
same subjects at two different times.
Crouchback's Alpha (Co-efficient Alpha): It is used
exclusively to check the reliability of a composite measures. It
can be laborious task with large sample.
Intra And Inter Coder Reliability: The easiest way to
determine whether a single observer is reliable to perform an
inter coder reliability check
![Page 11: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/11.jpg)
VALIDITY
Validity depicts the concept of reliability, external validity-the ability to generalize presumes internal validity.
• Content validity: It is a simplest form of research. It is reasonable one. It appears to be measuring what is intended to measure.
• Criterion Validity: using this method, concurrent validity can be established by finding a statistically significant, positive co relation between the measure or test in question for which validity has previously been put.
![Page 12: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/12.jpg)
• Predictive Validity: It can access whether scores on the test
are positively related to success of the college.
• Construct Validity: It is commonly used in academic
research. The common way to establish construct validity is
to test the measure in question against two other measures
known from previous work which are valid.
![Page 13: Scales and Measures in Research](https://reader034.vdocuments.site/reader034/viewer/2022042615/55a87f8c1a28ab37058b46ac/html5/thumbnails/13.jpg)
Thank you