Upload
pallavi412
View
228
Download
0
Embed Size (px)
Citation preview
7/31/2019 BRM Chap 13
1/46
MEASUREMENTAND
SCALING CONCEPTS
GROUP 7
7/31/2019 BRM Chap 13
2/46
WHAT IS TO BE MEASURED???
The CONCEPT
which can beSales Performance, height, Rate of
Absenteeism or anything you want to measure
7/31/2019 BRM Chap 13
3/46
HOW IT WILL BE MEASURED???
True Measurement of a concept requires aprocess of assigning precise scores or numbers
to the attributes of people or objects.
7/31/2019 BRM Chap 13
4/46
CONCEPT
A generalized idea about a class of objects,attributes, occurrences, or processes.
7/31/2019 BRM Chap 13
5/46
CONCEPT cont.. Concrete Concept: Age, Gender etc Abstract Concept: Loyalty, Job Involvement.
Abstract Concept Measurement generally involves anumber of measurements and scales.
7/31/2019 BRM Chap 13
6/46
OPERATIONAL DEFINITIONIt gives meaning to a concept by specifying theactivities/operations necessary to measure it. Ex: Concept: Interest
Operational Definition: A certain increase in
pupil DilationOR
What people say they are interested in.
7/31/2019 BRM Chap 13
7/46
An Example: Media SkepticismConceptual definition
Media skepticism - the degree to whichindividuals are skeptical toward the reality presented in the mass media. Media skepticism varies across individuals, from those who aremildly skeptical and accept most of what they
see and hear in the media to those whocompletely discount and disbelieve the facts, values, and portrayal of reality in the media.
7/31/2019 BRM Chap 13
8/46
Media Skepticism
Operational DefinitionPlease tell me how true each statement is aboutthe media. Is it very true, not very true, or not atall true?1. The program was not very accurate in itsportrayal of the problem.
2. Most of the story was staged forentertainment purposes.3. The presentation was slanted and unfair.
7/31/2019 BRM Chap 13
9/46
Rules of Measurement A Guide that tells someone what to do.
Simply put: Assign the numerals 1 to 7 toindividuals acc to how productive they are.If the individual is extremely productive assign a 7.
If the individual is an unproductive worker assign a1.
7/31/2019 BRM Chap 13
10/46
SCALE Series of items arranged according to value, into which an item can be placed for the purpose of quantification
A continuous spectrum 4 types:
Nominal
Ordinal
Interval
Ratio
7/31/2019 BRM Chap 13
11/46
NOMINAL SCALE The numbers or letters assigned to objects which
serve as labels for identification or classification.
7/31/2019 BRM Chap 13
12/46
ORDINAL SCALE Scale which arranges objects or alternatives
according to their magnitude in an orderedrelationship.
7/31/2019 BRM Chap 13
13/46
ORDINAL SCALE cont.. Ordinal measurements
describe order
But relative size ordegree of difference between the items
CANNOT be measured
7/31/2019 BRM Chap 13
14/46
INTERVAL SCALE An order which not only indicates order, but
also measure order (or distance) in units of equal interval
The location of zero point is arbitrary. Ex: CPI,Sensex
7/31/2019 BRM Chap 13
15/46
INTERVAL SCALE cont.. Ex: The Fahrenheit scale is aninterval scale.
Since each degree is equal butthere is no absolute zeropoint .
This means that although wecan add and subtract degrees(100 is 10 warmer than 90), we cannot multiply values orcreate ratios (100 is not twiceas warm as 50).
7/31/2019 BRM Chap 13
16/46
RATIO SCALE Ratio Scales are the ones which have absolute
rather than relative quantities. Ratio scales have non-arbitrary zero value.
Ex: Money and Weight possess an absolute zeroon the scale where there is an absence of thegiven attribute.
Mass, length, time, plane angle, energy
and electric charge are examples of physicalmeasures that are ratio scales.
7/31/2019 BRM Chap 13
17/46
Key Features
7/31/2019 BRM Chap 13
18/46
Properties: A Comparison
NominalScale
Uniquely classifies Sammy Sosa # 21 Barry Bonds # 27
OrdinalScale
Uniquely classifiesPreserves order
1st , 2nd and 3 rd
7/31/2019 BRM Chap 13
19/46
Properties: A Comparison
IntervalScale
Uniquely classifies Preserves order Equal intervals
Consumer Price Index (Base 100) Fahrenheit temperature
RatioScale
Uniquely classifies Preserves order Equal intervals Natural zero
Weight and distance
7/31/2019 BRM Chap 13
20/46
To Summarize: Typical Examples
NominalScaleMale, Female,Race, Political
Party
(categoricaldata thatcannot beranked)
OrdinalScale
Degree of Satisfaction
at Restaurant(data that
can beranked)
IntervalScale
Temperature,Dates (data
that have anarbitrary zero)
Ratio ScaleHeight,
Weight, Age,Length (data
that has anabsolutezero)
7/31/2019 BRM Chap 13
21/46
What We Can COMPUTE ??
Can Compute.. Nominal Ordinal Interval Ratio
Frequency Distribution. Yes Yes Yes Yes
Median and Percentiles. No Yes Yes Yes
Add or Subtract. No No Yes Yes
Mean, StandardDeviation, Standard Errorof The Mean.
No No Yes Yes
Ratio or Coefficient of Variation. No
No No Yes
7/31/2019 BRM Chap 13
22/46
Index Measures
Attribute A single characteristic or fundamentalfeature of an object, person, situation orissue.
Index or Composite MeasureMulti-item instrument constructed tomeasure a single concept with severalattributes.
7/31/2019 BRM Chap 13
23/46
Three Criteria For Good Measurement
Criteria
Reliability Validity Sensitivity
7/31/2019 BRM Chap 13
24/46
RELIABILITY Degree to which measures are free from error and
therefore yield consistent results
Low reliability can be due toimperfections in measuringprocess that effect indifferent ways eachmeasurement.
7/31/2019 BRM Chap 13
25/46
Dimensions of Reliability
Reliability
InternalConsistency
Split-half method
Equivalent-form method
Repeatability
Test-Retestmethod
7/31/2019 BRM Chap 13
26/46
Two main dimensions are:REPEATABILITY Used to assess the consistency of a measure from
one time to another.
Measure of Stability.
INTERNAL CONSISTENCY Used to assess the consistency of results across
items within a test. Measure of Homogeneity.
7/31/2019 BRM Chap 13
27/46
Test-Retest Method It is administering of same scale or measure to
the same respondents at two separate points intime under same conditions to test for stability.
7/31/2019 BRM Chap 13
28/46
Problems in Test-Retest Reliability
PREMEASURE The first measure may sensitize respondents to their participation inresearch project and thus effect results of secondmeasure.
TIME PERIOD BETWEEN MEASUREMENTS -The shorter the time gap, the higher thecorrelation and the longer the time gap, thelower the correlation.
7/31/2019 BRM Chap 13
29/46
Split Half Method
Is used to measure the degree of internalconsistency by checking one half of the results of a set of scaled items against the other half.
Used generally when measure contains a largenumber of items.
7/31/2019 BRM Chap 13
30/46
Equivalent/Parallel Form Method Method of measuring correlation between
alternative instruments, designed to be asequivalent as possible and administered to thesame group of subjects.
There can be problem if low correspondence between two instruments
7/31/2019 BRM Chap 13
31/46
Measurement needs to have not only
reliability but also VALIDITY.
The purpose of measurement is tomeasure what we intend to measure
7/31/2019 BRM Chap 13
32/46
VALIDITY Ability of a scale or measuring instrument to
measure what is intended to measure.
Validity
Face validity Criterion Validity
Concurrent Validity
Predictive Validity
Construct Validity
Convergent Validity
Discriminant Validity
7/31/2019 BRM Chap 13
33/46
Face / Content Validity Professional agreement that a scale logically
appears to accurately measure, what its isintended to measure.
Ex: What is your age?, What is your name?
In research its never sufficient to rely on face judgments alone and more quantifiable methodsof validity are necessary.
7/31/2019 BRM Chap 13
34/46
Criterion Validity
Ability of some measure to correlate with othermeasure of the same construct
7/31/2019 BRM Chap 13
35/46
Concurrent Validity new measure taken at
same time as criterion measure and is valid.
Predictive Validity new measure predicts afuture event or correlates with criterion measure
administered at a later time.
Criterion Validity (contd)
7/31/2019 BRM Chap 13
36/46
Construct ValidityIs the extent to which a measure fits into thetheoretical scheme and research already established on the attribute or construct beingmeasured.
Convergent Validity - is the degree to which anew measure is similar to (converges on) othermeasure that it theoretically should also be similarto.Discriminant Validity is ability of measure tohave low correlation with measures of dissimilarconcepts
7/31/2019 BRM Chap 13
37/46
Example
7/31/2019 BRM Chap 13
38/46
QUESTION
Can survey be
Valid but not reliable ? Reliable but not valid ?
7/31/2019 BRM Chap 13
39/46
Is instruments ability to accurately measure variability in stimuli or responses.
It can be increased by adding more questions.
Index measures are more sensitive than singleitem scales.
SENSITIVITY
7/31/2019 BRM Chap 13
40/46
SUMMARY Four types of measuring scales:
Nominal Scales Ordinal Scales Interval Scales Ratio Scales
Index (composite) Measures Several attributes
7/31/2019 BRM Chap 13
41/46
SUMMARY Contd.
Measurement Criteria for MeasuringInstruments:
Reliability: Provide consistent results
Validity: Measure what we intend to measure
Sensitivity: Measure variability
7/31/2019 BRM Chap 13
42/46
A Small Quiz:
Type of scale used:
Fahrenheit Scale
Coding for Males and Females
Percentile Ranking
Kelvin Scale
7/31/2019 BRM Chap 13
43/46
Quiz Contd.. Consider the centigrade scale for measuring
temperature. Which of the following measurementproperties is satisfied by the centigrade scale?
I. Magnitude.II. Equal intervals.III. Absolute zero.
(A) I only (B) II only (C) III only (D) I and II(E) II and III
7/31/2019 BRM Chap 13
44/46
A researcher developed a measure of shyness and is now asking whether this measure does in fact measure a person'strue state of shyness. This is a question of:
a.) Sensitivity b.) Construct validity c.) Reliability.
Scores on a final exam are related to student grade pointaverage, the amount of time spent studying for the exam, andclass attendance. What type of validity is demonstrated in this
case?a.) Convergent validity b.) Discriminant validity c.) Criterion validity.
Quiz Contd..
7/31/2019 BRM Chap 13
45/46
Which of the following statements accurately portrays the relationship of reliability to validity?
a.) Inferences must be valid before the scores can bereliable.
b.) Scores must be reliable before inferences can be valid.
c.) The more valid the inference is, the higher thereliability of the score must be.
d.) Score reliability is not related to inference validity.
Quiz Contd..
7/31/2019 BRM Chap 13
46/46