4. Scale
For Measurement ,understanding,comparision .
Series of items arranged according to value for the purpose
of quantification,numbers,symbols etc
A continuous spectrum
5. Scale characteristics
Description/classification/label (attendendence register)
Order or rank or > or < or = (100mtr Dash ranking)
Distance (Age difference btw 1st brother and 2nd brother is 2 years,2nd
brother and 3rd brother is 2 years )
Origin has a fixed beginning or true zero not assumed zero.(Income)
12. Ordinal Scale Properties
Uniquely classifies
Preserves order
Win, place, & show
IPL(Indian Premier League) standings 2017
13. Interval Scale Properties
Uniquely classifies
Preserves order
Equal intervals
Consumer Price Index (Base 100)
Fahrenheit temperature
FIFA qualifiers 3 groups =6+2(Best 2 loosing team with less
goal difference among three groups)
14. Ratio Scale Properties
Uniquely classifies
Preserves order
Equal intervals
Natural zero
Weight and distance, batsman score,
Zero base
15. Index Measures
ATTRIBUTES A single characteristic or fundamental feature
that pertains to an object, person, or issue
Measure Social class index: occupation, education and area
of residence.
COMPOSITE MEASURE A composite measure of several
variables to measure a single concept; a multi-item
instrument: Job satisfaction of sales person
18. Reliability
The degree to which measures are
free from random error and
therefore yield consistent results
19. Old Rifle New Rifle New Rifle
Sun glare
Low Reliability High Reliability Reliable but Not
Valid
(Target A) (Target B) (Target C)
Reliability and Validity on
Target
22. Scale evaluation
Validity( Scale used should measure what is intended to be measured)
Reliability (Internal consistency)(Student’s MCQ result and teacher
prediction of their performance in essay)
Eg: Money counting machine and weighing machine can be used to
measure the value of money which is valid? Reliable?)
Reliability is a necessary, but not sufficient condition for validity.
In other words, you could be inaccurately measuring something
consistently.
Validity and then Reliability make sense
24. Face validity/ Content validity : Experts acknowledgement on the measure
Measuring customer loyalty:
1.I prefer to purchase my groceries at Kannan departmental stores
2.I am very satisfied with my purchase from Kannan store
3.Kanna store offers good value for the product
The degree that measure covers the breath of the domain of interest but not
beyond.
Eg: Content validity of a Exam question paper
Syllabus: ch1 to ch 5
Questions in the question paper :40% from ch1 and 40 %from ch2 ,10% from ch3 and
and 10 % from ch 6.
Eg: Assessing experience of spa customer
1.Asking about service delivery(utilitarian value) doesn’t give you the exact
experience but also the pleasure and comfort (Hedonic value)
25. Criterion validity/pragmatic: The ability of a measure to correlate with
other standard measure of similar constructs or established criteria.
Judging of criterion validity :Relevence,Freedom from bias ,reliability
and availability.
Classification: Concurrent and predictive validity
Practical test on day 1 and theory test of same subject on day 2 ,if the
scores are same it has concurrent validity
Team lead’s score of the employee and job performance score test of
the employee are same then team lead’s score has concurrent validity
Capability of predicting some latter measure
This institution only admits those who are capable of scoring 80% CGPA
Eg: TNSET scorer 45/60 and CGPA in PG 80 %
Here the TANCET score has a predictive validity
26. Construct Validity:A type of validity that addresses the question of what
construct or characteristic the scale is measuring.
Classification:Convergent,Discriminent,Nomological
If we believe customer loyalty and consistent feedback by customer are
related,then both should be moderately correlated then customers
providing feedback has convergent validity towards customer loyalty
measure.
If the correlation is very high then it has no discriminant validity and
vise versa.
Nomological Validity is a statistical term that refers to a "nonlogical"
necessity or a law of nature.
For example, the tendency to purchase prestige brands
should show a high correlation with a person’s need for status and
materialism and a negative correlation with price sensitivity.
27. Reliability:It’s a measure of internal consistency (measure
several similar but non identical questions about the object
or characteristics).
Test methods: Internal consistency = Split-half and
Coefficient a (Alpha),
Test-retest, alternative forms.
Split half: Dividing items to two half's and they should
highly correlate,how to divide two halfs?perform coefficient
alpha .
Coefficient alpha computes all possible splithalf relaiability
for multiple item scale,and they should converge.Aplha(a)
=0 To 1if its below 0.6 poor relaiability.
Test-retest: same test same respondents same condition at
different point of time.
Alternative forms:Two equivalent forms of scale,same
respondents at two different time