Rating scales
By-
DHEERAJ KUMAR
M.A. 3rd semester
RATING SCALES
A rating scale is a measuring instrument that
requires the rater or observer to assign a
rated object to categories or continua that
have numerals assigned to them.
Rating asks the respondent to estimate the
magnitude or the extent to which some
characteristic exists.
Is a measurement task that requires
respondents to estimate the magnitude of a
characteristic or quality that a brand, store,
or object processes.
TYPES OF RATING SCALES
Category rating scale
Graphic rating scale
Graphic scale with picture
response categories
 Method of Summated Ratings:
Likert rating scale
TYPES OF RATING SCALES
Composite rating scale
Semantic differential scale
Numerical rating scale
Stepel scale
TYPES OF RATING SCALES
Check list
Forced choice rating scale
Percentage rating scale
Balanced or Unbalanced Rating Scale
Category Rating Scales
The category rating scale presents the observer or
judge with several categories from which he picks
the one that best categorizes the behavior or
characteristics of the object being rated.
The simplest rating scale contains only two
response categories: agree/disagree.
CATEGORY RATING SCALES
If response category is expanded—
It provides the respondent with more flexibility in
the rating task.
Even more information will be provided if the
categories are ordered according to a particular
descriptive or evaluative dimension.
Example of Category Scales
 If the researcher tries to represent something that is
truly bipolar (yes/no, female/male, member/nonmember,
and so on) with more than two categories, error may be
introduced.
Question wording is an extremely important factor in
the usefulness of these scales . Following Table shows
some common wordings used in category scales.
 In buying food for my family, price is no object.
Strongly Disagree Uncertain Agree Strongly
Disagree Agree
(1) (2) (3) (4) (5)
Selected Category Scales
Quality
Excellent Good Fair Poor
Very good Fairly good Neither good nor bad Not very good Not good at all
Well above average Above average Average Below average Well below average
Importance
Very important Fairly important Neutral Not so important Not at all important
Interest
Very interested Somewhat interested Not very interested
Satisfaction
Completely Somewhat Neither satisfied Somewhat Completely
satisfied satisfied nor dissatisfid dissatisfied dissatisfied
Very satisfied Quite satisfied Somewhat satisfied Not at all satisfied
Frequency
All of the time Very often Often Sometimes Hardly ever
Very often Often Sometimes Rarely Never
All of the time Most of the time Some of the time Just now and then
Truth
Very true Somewhat true Not very true Not at all true
Definitely yes Probably yes Probably no Definitely no
Uniqueness
Very different Somewhat Slightly different Not at all different
different
Extremely unique Very unique Somewhat unique Slightly unique Not at all unique
The Likert Scale
A measure of attitudes designed to allow respondents to
rate how strongly they agree or disagree with carefully
constructed statements, ranging from very positive to
very negative attitudes toward some object.
A Likert scale consist of a series of statements each
followed by a series of response alternative for the
respondent to express himself or herself about the
statement.
In a Likert scale the most common number of
alternative is five because it often respondents a sufficient
range of choice without requiring unnecessarily minute
distinctions in attitudes
The Likert Scale……
strongly agree, agree, uncertain, disagree, and strongly disagree
The number of alternatives may range from
three to nine.
 In the following example, from a study of food-shopping
behavior, there are five alternatives :
 In buying food for my family, price is no object.
Strongly Disagree Uncertain Agree Strongly
Disagree Agree
(1) (2) (3) (4) (5)
.
The Likert Scale……….
Concept/Construct
Operational definition
Dimensions
Purpose/objective
Population
Constructing a Likert scale
The Likert Scale……
Begins by developing a series of statements relating
to the variable being measured by using general
criteria for statement .
How much item should developed?
Three times more the number of scale desired for
the final scale.
Generation of item pool
The Likert Scale……
General criteria for generating item
 Should be -Simple ,clear, and direct.
Should be unambiguous.
Must strike a balance between generality and specificity in
writing items.
Should avoid constructing leading items.
Should not be factual.
Avoid items which require respondents to recall information
from the distant past.
Avoid double-barreled items, which is one item asking two
questions. (height &weight)
Should not use word like often, always, never, some time etc.
Should avoid double negative items……………………..
The Likert Scale……
Items screening
On the basis of expert advice-
On the basis of statistical analysis—
Discrimination power index/score - essentially tell us
The degree to which each item differentiates between
Respondents with high score and respondents with
low score on the overall scale.
Correlational technique/internal consistency item
discrimination index(r 0.15≤) & Method of testing `the
significance of difference between two proportions or
percentage (1≤)
Researchers assign scores, or weights, to
each possible response. In this example, numerical scores of 1, 2,
3, 4, and 5 are assigned to each level of agreement, respectively.
 REVERSE RECODING
If a statement is framed negatively the numerical scores would
need to be reversed.
For example-
“I carefully budget my food expenditures”
The Likert Scale…..
A method of making sure all the items forming a
composite scale are scored in the same direction.
Negative items can be recoded into the equivalent
responses for a non-reverse coded item.
For a typical 1–5 scale, the formula of REVERSE RECODING
REVERSE RECODING
Xnew value = 6-Xold value
Reliability of Likert scale
Cronbach`s/Coefficient alpha(α)
α = (n/n-1)×{ΣDi-Σ(SDi
2) }/SDi
2
It is the most popular multiple-item formats
because of the many advantages it processes.
The Likert Scale……
ADVANTAGES
Data produced by Likert scale at least ordinal level
and many consider them Interval level, which enables
us to use more powerful statistical procedure than
with nominal data.
Likert measure are fairly straightforward to
construct.
The Likert Scale……
DISADVANTEGES
One must be careful in interpreting a single score based on
a Likert Scale because it is a summary of so much information
(separate response to a number of items ).
To generate unambiguous and unbiased item.
COMPOSITE SCALES
A way of representing a latent construct by summing
or averaging respondents’ reactions to multiple items
each assumed to indicate the latent construct.
A Likert scale may include several scale items to form
a composite scale.
Each statement is assumed to represent an aspect of
a common attitudinal domain.
Likert Scale Items for Measuring Attitudes toward Patients’
Interaction with a Physician’s Service Staff
COMPOSITE SCALES…..
1. My doctor’s office staff takes a warm and personal interest in me.
2. My doctor’s office staff is friendly and courteous.
3. My doctor’s office staff is more interested in serving the doctor’s needs than in serving my
needs.
4. My doctor’s office staff always acts in a professional manner.
Item 3 is negatively worded and therefore it is reverse
coded , prior to being used to create the composite scale.
Example-
Semantic Differential
A measure of attitudes that consists of a series
of seven point rating scales that use bipolar
adjectives to anchor the beginning and end of each
scale.
Bipolar adjectives—“good” and “bad,” “modern”
and “old fashioned,” or “clean” and “dirty” etc.
Respondents are instructed to check the place
that indicates the nearest appropriate adjective.
 Scoring of the semantic differential
From left to right, the scale intervals are interpreted as:
“extremely modern,” “very modern,” “slightly modern,”
“both modern and old-fashioned,” “slightly old-fashioned,”
“very old-fashioned,” and “extremely old-fashioned”:
7 6 5 4 3 2 1
Modern - - - - - - - Old fashioned”
Semantic differential scale…..
The semantic differential technique originally was
developed as a method for measuring the meanings
of objects or the “semantic space” of interpersonal
experience.
Scoring of the Semantic Differential….
To reduce response set (i.e., a subject responding
in a consistent manner that does not represent his
or her “real” sentiments) several score weights are
reversed.
For example:
(a) Eager (7)……………….. Indifferent (1)
(b) Useless (1)………………Useful (7)
Scoring of Semantic Differential……..
Which type of data it will produce?
This item format produces interval level data
and parametric statistics may be applied.
All of the Measures of central tendency and
MCV`s are appropriate.
The number and percentage of respondents
endorsing each response option is also reported.
Advantages and disadvantages of
semantic differential scale
SDs are much easier and less time consuming to construct
than other scales.
Adjective pairs are easier to develop than are unambiguous
and unbiased statement.
Adjective pairs from prior studies are more readily adaptable
to new studies because of their general and nonspecific
nature.
Identifying the abstract dimensions tapped by the adjective
pairs is somewhat subjective and judgmental.
The validity of the conclusion drawn is only as good as the
judgment of those who identify the dimension.
Numerical Scales
An attitude rating scale similar to a semantic
differential except that it uses numbers, instead of
verbal descriptions, as response options to identify
response positions.
Now that you’ve had your automobile for about one
year, please tell us how satisfied you are with your Ford
Taurus.
Extremely Dissatisfied 1 2 3 4 5 6 7 Extremely Satisfied
Stapel Scale
 A simplified version of the semantic differential scale in
which a single adjective or descriptive phrase is used
instead of bipolar adjectives.
Characteristics
1. The scale measures both the direction and intensity of the
attribute simultaneously.
2. It has properties similar to the semantic differential.
Example:
Model A
-3 -2 -1 Durable Car 1 2 3
-3 -2 -1 Good Fuel Conaumption 1 2 3
Response Bias
Source of response bias-
Response set
Response pattern anxiety
Social desirability effect
Hard to attach a verbal
explanation to response
Visual impact, easy for poor
readers
Choose a visual picture8. Graphic scale-picture
response
No standard answersVisual impact, unlimited
scale points
Choose a point on a
continuum
7. Graphic scale
Endpoints are numerical,
not verbal.
Easier to construct than
semantic differential
Choose point on scale
with 1 center adjective
6. Stapel scale
Bipolar adjectives must be
found, data may be
ordinal, may be interval
Easy to construct, norms
exist for comparison, e.g.
profile analysis
Choose points between
bipolar adjectives on
relative dimensions
4. Semantic differential
and numerical scales
Hard to judge what a
single score means
Easiest scale to constructEvaluate statements on a
5-point scale
3. Likert scale
Ambiguous items, few
categories, only gross
distinction.
Flexible, easy to respondIndicate a response
category
2.Category scale
1. Simple attitude scaling
DisadvantagesAdvantagesSubject must:Rating Scale
Characteristics Different Types of Rating Scales
Issues Related to Rating Scale
1. Whether to use single or index measure.
2. Whether to use a ranking, sorting, choice, or rating
scale.
3. Whether to use monadic or comparative scale.
• Monadic rating scale is one in which respondents
evaluate an object in isolation
• Comparative scale s one in which the object is evaluated
in relation to other objects
• Construction and labeling is different for monadic and
comparative scales
4. Whether to use category labels or not.
5. If the decision is to use category labels, what labels
to use.
Issues Related to Rating Scale
6. Number of response options (scale categories) to use, i.e
whether to use 2, 3, 4, 5, etc response categories
• In general, the larger the number of categories the more sensitive the
scale is; but also the more difficult it is for respondents to answer
7. Whether to use balanced or unbalanced scale.
• A balanced scale has an equal number of points to the left and right of
a mid-point. An unbalanced scale has more response options on one
side than the other
8. Whether the scale should force choice among the response
categories, i.e should the scale contain a “neutral” or “don’t
know” category.
•Thank You

Rating scale

  • 1.
  • 2.
    RATING SCALES A ratingscale is a measuring instrument that requires the rater or observer to assign a rated object to categories or continua that have numerals assigned to them.
  • 3.
    Rating asks therespondent to estimate the magnitude or the extent to which some characteristic exists. Is a measurement task that requires respondents to estimate the magnitude of a characteristic or quality that a brand, store, or object processes.
  • 4.
    TYPES OF RATINGSCALES Category rating scale Graphic rating scale Graphic scale with picture response categories  Method of Summated Ratings: Likert rating scale
  • 5.
    TYPES OF RATINGSCALES Composite rating scale Semantic differential scale Numerical rating scale Stepel scale
  • 6.
    TYPES OF RATINGSCALES Check list Forced choice rating scale Percentage rating scale Balanced or Unbalanced Rating Scale
  • 7.
    Category Rating Scales Thecategory rating scale presents the observer or judge with several categories from which he picks the one that best categorizes the behavior or characteristics of the object being rated. The simplest rating scale contains only two response categories: agree/disagree.
  • 8.
    CATEGORY RATING SCALES Ifresponse category is expanded— It provides the respondent with more flexibility in the rating task. Even more information will be provided if the categories are ordered according to a particular descriptive or evaluative dimension.
  • 9.
    Example of CategoryScales  If the researcher tries to represent something that is truly bipolar (yes/no, female/male, member/nonmember, and so on) with more than two categories, error may be introduced. Question wording is an extremely important factor in the usefulness of these scales . Following Table shows some common wordings used in category scales.  In buying food for my family, price is no object. Strongly Disagree Uncertain Agree Strongly Disagree Agree (1) (2) (3) (4) (5)
  • 10.
    Selected Category Scales Quality ExcellentGood Fair Poor Very good Fairly good Neither good nor bad Not very good Not good at all Well above average Above average Average Below average Well below average Importance Very important Fairly important Neutral Not so important Not at all important Interest Very interested Somewhat interested Not very interested Satisfaction Completely Somewhat Neither satisfied Somewhat Completely satisfied satisfied nor dissatisfid dissatisfied dissatisfied Very satisfied Quite satisfied Somewhat satisfied Not at all satisfied Frequency All of the time Very often Often Sometimes Hardly ever Very often Often Sometimes Rarely Never All of the time Most of the time Some of the time Just now and then Truth Very true Somewhat true Not very true Not at all true Definitely yes Probably yes Probably no Definitely no Uniqueness Very different Somewhat Slightly different Not at all different different Extremely unique Very unique Somewhat unique Slightly unique Not at all unique
  • 11.
    The Likert Scale Ameasure of attitudes designed to allow respondents to rate how strongly they agree or disagree with carefully constructed statements, ranging from very positive to very negative attitudes toward some object. A Likert scale consist of a series of statements each followed by a series of response alternative for the respondent to express himself or herself about the statement.
  • 12.
    In a Likertscale the most common number of alternative is five because it often respondents a sufficient range of choice without requiring unnecessarily minute distinctions in attitudes The Likert Scale…… strongly agree, agree, uncertain, disagree, and strongly disagree
  • 13.
    The number ofalternatives may range from three to nine.  In the following example, from a study of food-shopping behavior, there are five alternatives :  In buying food for my family, price is no object. Strongly Disagree Uncertain Agree Strongly Disagree Agree (1) (2) (3) (4) (5) . The Likert Scale……….
  • 14.
  • 15.
    The Likert Scale…… Beginsby developing a series of statements relating to the variable being measured by using general criteria for statement . How much item should developed? Three times more the number of scale desired for the final scale. Generation of item pool
  • 16.
    The Likert Scale…… Generalcriteria for generating item  Should be -Simple ,clear, and direct. Should be unambiguous. Must strike a balance between generality and specificity in writing items. Should avoid constructing leading items. Should not be factual. Avoid items which require respondents to recall information from the distant past. Avoid double-barreled items, which is one item asking two questions. (height &weight) Should not use word like often, always, never, some time etc. Should avoid double negative items……………………..
  • 17.
    The Likert Scale…… Itemsscreening On the basis of expert advice- On the basis of statistical analysis— Discrimination power index/score - essentially tell us The degree to which each item differentiates between Respondents with high score and respondents with low score on the overall scale. Correlational technique/internal consistency item discrimination index(r 0.15≤) & Method of testing `the significance of difference between two proportions or percentage (1≤)
  • 18.
    Researchers assign scores,or weights, to each possible response. In this example, numerical scores of 1, 2, 3, 4, and 5 are assigned to each level of agreement, respectively.  REVERSE RECODING If a statement is framed negatively the numerical scores would need to be reversed. For example- “I carefully budget my food expenditures” The Likert Scale…..
  • 19.
    A method ofmaking sure all the items forming a composite scale are scored in the same direction. Negative items can be recoded into the equivalent responses for a non-reverse coded item. For a typical 1–5 scale, the formula of REVERSE RECODING REVERSE RECODING Xnew value = 6-Xold value
  • 20.
    Reliability of Likertscale Cronbach`s/Coefficient alpha(α) α = (n/n-1)×{ΣDi-Σ(SDi 2) }/SDi 2
  • 21.
    It is themost popular multiple-item formats because of the many advantages it processes. The Likert Scale…… ADVANTAGES Data produced by Likert scale at least ordinal level and many consider them Interval level, which enables us to use more powerful statistical procedure than with nominal data. Likert measure are fairly straightforward to construct.
  • 22.
    The Likert Scale…… DISADVANTEGES Onemust be careful in interpreting a single score based on a Likert Scale because it is a summary of so much information (separate response to a number of items ). To generate unambiguous and unbiased item.
  • 23.
    COMPOSITE SCALES A wayof representing a latent construct by summing or averaging respondents’ reactions to multiple items each assumed to indicate the latent construct. A Likert scale may include several scale items to form a composite scale. Each statement is assumed to represent an aspect of a common attitudinal domain.
  • 24.
    Likert Scale Itemsfor Measuring Attitudes toward Patients’ Interaction with a Physician’s Service Staff COMPOSITE SCALES….. 1. My doctor’s office staff takes a warm and personal interest in me. 2. My doctor’s office staff is friendly and courteous. 3. My doctor’s office staff is more interested in serving the doctor’s needs than in serving my needs. 4. My doctor’s office staff always acts in a professional manner. Item 3 is negatively worded and therefore it is reverse coded , prior to being used to create the composite scale. Example-
  • 25.
    Semantic Differential A measureof attitudes that consists of a series of seven point rating scales that use bipolar adjectives to anchor the beginning and end of each scale. Bipolar adjectives—“good” and “bad,” “modern” and “old fashioned,” or “clean” and “dirty” etc. Respondents are instructed to check the place that indicates the nearest appropriate adjective.
  • 26.
     Scoring ofthe semantic differential From left to right, the scale intervals are interpreted as: “extremely modern,” “very modern,” “slightly modern,” “both modern and old-fashioned,” “slightly old-fashioned,” “very old-fashioned,” and “extremely old-fashioned”: 7 6 5 4 3 2 1 Modern - - - - - - - Old fashioned” Semantic differential scale….. The semantic differential technique originally was developed as a method for measuring the meanings of objects or the “semantic space” of interpersonal experience.
  • 27.
    Scoring of theSemantic Differential…. To reduce response set (i.e., a subject responding in a consistent manner that does not represent his or her “real” sentiments) several score weights are reversed. For example: (a) Eager (7)……………….. Indifferent (1) (b) Useless (1)………………Useful (7)
  • 28.
    Scoring of SemanticDifferential…….. Which type of data it will produce? This item format produces interval level data and parametric statistics may be applied. All of the Measures of central tendency and MCV`s are appropriate. The number and percentage of respondents endorsing each response option is also reported.
  • 29.
    Advantages and disadvantagesof semantic differential scale SDs are much easier and less time consuming to construct than other scales. Adjective pairs are easier to develop than are unambiguous and unbiased statement. Adjective pairs from prior studies are more readily adaptable to new studies because of their general and nonspecific nature. Identifying the abstract dimensions tapped by the adjective pairs is somewhat subjective and judgmental. The validity of the conclusion drawn is only as good as the judgment of those who identify the dimension.
  • 30.
    Numerical Scales An attituderating scale similar to a semantic differential except that it uses numbers, instead of verbal descriptions, as response options to identify response positions. Now that you’ve had your automobile for about one year, please tell us how satisfied you are with your Ford Taurus. Extremely Dissatisfied 1 2 3 4 5 6 7 Extremely Satisfied
  • 31.
    Stapel Scale  Asimplified version of the semantic differential scale in which a single adjective or descriptive phrase is used instead of bipolar adjectives. Characteristics 1. The scale measures both the direction and intensity of the attribute simultaneously. 2. It has properties similar to the semantic differential. Example: Model A -3 -2 -1 Durable Car 1 2 3 -3 -2 -1 Good Fuel Conaumption 1 2 3
  • 32.
    Response Bias Source ofresponse bias- Response set Response pattern anxiety Social desirability effect
  • 33.
    Hard to attacha verbal explanation to response Visual impact, easy for poor readers Choose a visual picture8. Graphic scale-picture response No standard answersVisual impact, unlimited scale points Choose a point on a continuum 7. Graphic scale Endpoints are numerical, not verbal. Easier to construct than semantic differential Choose point on scale with 1 center adjective 6. Stapel scale Bipolar adjectives must be found, data may be ordinal, may be interval Easy to construct, norms exist for comparison, e.g. profile analysis Choose points between bipolar adjectives on relative dimensions 4. Semantic differential and numerical scales Hard to judge what a single score means Easiest scale to constructEvaluate statements on a 5-point scale 3. Likert scale Ambiguous items, few categories, only gross distinction. Flexible, easy to respondIndicate a response category 2.Category scale 1. Simple attitude scaling DisadvantagesAdvantagesSubject must:Rating Scale Characteristics Different Types of Rating Scales
  • 34.
    Issues Related toRating Scale 1. Whether to use single or index measure. 2. Whether to use a ranking, sorting, choice, or rating scale. 3. Whether to use monadic or comparative scale. • Monadic rating scale is one in which respondents evaluate an object in isolation • Comparative scale s one in which the object is evaluated in relation to other objects • Construction and labeling is different for monadic and comparative scales 4. Whether to use category labels or not. 5. If the decision is to use category labels, what labels to use.
  • 35.
    Issues Related toRating Scale 6. Number of response options (scale categories) to use, i.e whether to use 2, 3, 4, 5, etc response categories • In general, the larger the number of categories the more sensitive the scale is; but also the more difficult it is for respondents to answer 7. Whether to use balanced or unbalanced scale. • A balanced scale has an equal number of points to the left and right of a mid-point. An unbalanced scale has more response options on one side than the other 8. Whether the scale should force choice among the response categories, i.e should the scale contain a “neutral” or “don’t know” category.
  • 36.

Editor's Notes

  • #25 In Likert’s original procedure, a large number of statements are generated, and an item analysis is performed. The purpose of the item analysis is to ensure that final items evoke a wide response and discriminate among those with positive and negative attitudes. Items that are poor because they lack clarity or elicit mixed response patterns are eliminated from the final statement list. Scales that use multiple items can be analyzed for reliability and validity. Only a set of items that demonstrates good reliability and validity should be summed or averaged to form a composite scale representing a hypothetical construct. Unfortunately, not all researchers are willing or able to thoroughly assess reliability and validity. Without this test, the use of Likert scales can be disadvantageous because there is no way of knowing exactly what the items represent or how well they represent anything of interest. Without valid and reliable measures, researchers cannot guarantee they are measuring what they say they are measuring.