Rating ScaleRating Scale
Mukut HazarikaMukut Hazarika
ProfessorProfessor
Department of EducationDepartment of Education
Dibrugarh UniversityDibrugarh University
Rating ScaleRating Scale
• Rating scale refers to a scale with a set of pointsRating scale refers to a scale with a set of points
which describe varying degrees of thewhich describe varying degrees of the
dimension of an attribute being observed (Koul,dimension of an attribute being observed (Koul,
1997).1997).
• A rating scale is defined as a technique throughA rating scale is defined as a technique through
which the observer or rater categorizes thewhich the observer or rater categorizes the
objects, events or persons on a continuumobjects, events or persons on a continuum
represented by a series of continuous numeralsrepresented by a series of continuous numerals
(Singh,1997).(Singh,1997).
Types of Rating ScalesTypes of Rating Scales
• Numerical ScalesNumerical Scales
• Graphical ScalesGraphical Scales
• Percentage RatingPercentage Rating
• Standard ScalesStandard Scales
• Scales of cumulated pointsScales of cumulated points
• Forced choice scalesForced choice scales
Numerical ScalesNumerical Scales
• Numerical scale consists of a sequence ofNumerical scale consists of a sequence of
defined numbers which is supplied to thedefined numbers which is supplied to the
rater or the observer. The rater or therater or the observer. The rater or the
observer assigns to each stimulus, to beobserver assigns to each stimulus, to be
rated, an appropriate number as definedrated, an appropriate number as defined
or describes.or describes.
Graphic ScalesGraphic Scales
• The scales are presented graphically inThe scales are presented graphically in
which descriptive cues corresponding towhich descriptive cues corresponding to
the different scale steps are given. In thisthe different scale steps are given. In this
scale, a straight line is drawn vertically orscale, a straight line is drawn vertically or
horizontally with various cues to help thehorizontally with various cues to help the
rater. The line may be segmented in unitsrater. The line may be segmented in units
or continuous.or continuous.
Percentage RatingPercentage Rating
• This technique involves placing objects,This technique involves placing objects,
persons, etc among different specifiedpersons, etc among different specified
percentage groups or into differentpercentage groups or into different
percentiles or quartiles. For example: inpercentiles or quartiles. For example: in
the highest 5 per cent of the group, in thethe highest 5 per cent of the group, in the
middle 25 percent of the group, in themiddle 25 percent of the group, in the
lowest 10 percent, etc.lowest 10 percent, etc.
Standard ScaleStandard Scale
• A standard scale is one in which the raterA standard scale is one in which the rater
is presented with some standards withis presented with some standards with
pre-established scale values. Thesepre-established scale values. These
standards usually consist of objects ofstandards usually consist of objects of
same kind. For example: Handwriting-same kind. For example: Handwriting-
standards, Portrait-matching, man-to-manstandards, Portrait-matching, man-to-man
scale, etc.scale, etc.
Portrait Matching ScalePortrait Matching Scale
• In this technique a set of standards or verbalIn this technique a set of standards or verbal
sketches or portraits for any given trait on whichsketches or portraits for any given trait on which
rating is to be done, is prepared.rating is to be done, is prepared.
• For constructing the verbal portrait a largeFor constructing the verbal portrait a large
number of verbal statements describing that traitnumber of verbal statements describing that trait
are collected and each is written in a separateare collected and each is written in a separate
card.card.
• They are then rank-ordered by a group ofThey are then rank-ordered by a group of
experts.experts.
• Then a desired number of sketches areThen a desired number of sketches are
prepared which have the same average rank.prepared which have the same average rank.
• The portraits are given to another group ofThe portraits are given to another group of
judges for ranking them and mean rankjudges for ranking them and mean rank
becomes the scale value.becomes the scale value.
• Then the rater is given all the verbal portraits toThen the rater is given all the verbal portraits to
read and he names the persons who belong toread and he names the persons who belong to
the portraits.the portraits.
• A person’s final rating is the average of all theA person’s final rating is the average of all the
portraits values that have been assigned to himportraits values that have been assigned to him
by all the raters.by all the raters.
Scales of cumulated PointsScales of cumulated Points
• The rating score of a person, in theseThe rating score of a person, in these
scales, is the sum of the individualscales, is the sum of the individual
ratings or points assigned to all the itemsratings or points assigned to all the items
of the scale.of the scale.
(a)(a)The Checklist Method:The Checklist Method:
(b)(b)The Guess-who Technique:The Guess-who Technique:
• Here is one who is always helping others.Here is one who is always helping others.
• Here is one who is always discouragingHere is one who is always discouraging
others.others.
• Here is one who is always happy.Here is one who is always happy.
Forced-Choice RatingsForced-Choice Ratings
• In forced choice rating scale the rater is given aIn forced choice rating scale the rater is given a
set of attributes in terms of verbal statements forset of attributes in terms of verbal statements for
a single item and he decides which one or ones,a single item and he decides which one or ones,
represent the individual being rated mostrepresent the individual being rated most
appropriately and accurately. The items ofappropriately and accurately. The items of
forced-choice scale may have severalforced-choice scale may have several
alternatives. The four alternatives are morealternatives. The four alternatives are more
common in forced-choice scale. In this form, thecommon in forced-choice scale. In this form, the
rater may be asked to select any two statementsrater may be asked to select any two statements
which are most descriptive and representative ofwhich are most descriptive and representative of
the persons being rated.the persons being rated.
Example of Forced-ChoiceExample of Forced-Choice
ScaleScale
• Is well informed in science.Is well informed in science.
• Can apply scientific fact and theory toCan apply scientific fact and theory to
practical situations.practical situations.
• Create confidence in those with whom heCreate confidence in those with whom he
deals.deals.
• Explains the reasons for hisExplains the reasons for his
recommendations.recommendations.
Distinction between Forced-ChoiceDistinction between Forced-Choice
Rating Scale and other types ofRating Scale and other types of
Rating ScaleRating Scale
• All other types of rating techniquesAll other types of rating techniques
consider one attribute at a time and placeconsider one attribute at a time and place
the ratee in any one of a set of categories.the ratee in any one of a set of categories.
Steps followed in Construction andSteps followed in Construction and
Standardization of Rating Scale:Standardization of Rating Scale:
• Definition of traitsDefinition of traits : Each trait should: Each trait should
be clearly defined so that traits may bebe clearly defined so that traits may be
clearly and uniformly understood by allclearly and uniformly understood by all
judges.judges.
• Degree of traitDegree of trait: The degree of trait: The degree of trait
should be defined.should be defined.
• Reliability:Reliability:
• Validity:Validity:
Methods of Determining Reliability:Methods of Determining Reliability:
• Reliability of rating scales are determinedReliability of rating scales are determined
by repeating judgments after a timeby repeating judgments after a time
interval; by determining correlationinterval; by determining correlation
coefficient between ratings of two or morecoefficient between ratings of two or more
judges; and by determining relationshipjudges; and by determining relationship
between judges’ ratings and self-ratings.between judges’ ratings and self-ratings.
Factors affecting Reliability ofFactors affecting Reliability of
Rating Scale:Rating Scale:
• Reliability of rating scale depends upon extentReliability of rating scale depends upon extent
variation of judges’ ratings. Judges rating anvariation of judges’ ratings. Judges rating an
individual on a specified trait will not alwaysindividual on a specified trait will not always
agree as to his score or rank. The mean or theagree as to his score or rank. The mean or the
median of all the judgments represent themedian of all the judgments represent the
nearest approximation to the true rating.nearest approximation to the true rating.
However, the averaging to be meaningful, theHowever, the averaging to be meaningful, the
variation of the judges’ rating should be small.variation of the judges’ rating should be small.
• Overt traits are more reliably rated than covertOvert traits are more reliably rated than covert
traits.traits.
• Reliability of trait estimates is affected byReliability of trait estimates is affected by
desirability or undesirability of the trait. Indesirability or undesirability of the trait. In
self ratings there is a tendency forself ratings there is a tendency for
individuals to overrate themselves inindividuals to overrate themselves in
respect to traits regarded as sociallyrespect to traits regarded as socially
desirable. In rating other persons,desirable. In rating other persons,
especially friends, some judges may beespecially friends, some judges may be
similarly influenced, even while trying tosimilarly influenced, even while trying to
be careful. There is also a generalbe careful. There is also a general
tendency toward generosity, rather thantendency toward generosity, rather than
the reverse.the reverse.
• Some persons are more accurately rated thanSome persons are more accurately rated than
others. Extroverted individuals are more reliablyothers. Extroverted individuals are more reliably
judged than introverted.judged than introverted.
• Degree of certainty of ratings should be stated.Degree of certainty of ratings should be stated.
With each rating, it is desirable to have theWith each rating, it is desirable to have the
respondent state his degree of certainty that isrespondent state his degree of certainty that is
very strong, strong, moderate, etc. Judges arevery strong, strong, moderate, etc. Judges are
usually in closer agreement on ratings at theusually in closer agreement on ratings at the
extremes because extreme deviants are mostextremes because extreme deviants are most
clearly distinguishable from others. Thus suchclearly distinguishable from others. Thus such
terms as ‘introverted-extroverted’ apply clearly toterms as ‘introverted-extroverted’ apply clearly to
individuals who are manifestly at one pole or theindividuals who are manifestly at one pole or the
other.other.
Validity of Rating Scale:Validity of Rating Scale:
• The validity of a rating scale is the extentThe validity of a rating scale is the extent
to which the judges’ understand theto which the judges’ understand the
meaning of the traits being evaluated andmeaning of the traits being evaluated and
the extent of accuracy in rating them.the extent of accuracy in rating them.
• The principal indication of the validity ofThe principal indication of the validity of
some rating scale is the extent to whichsome rating scale is the extent to which
the person using them find them useful.the person using them find them useful.

Research tool.rating scale

  • 1.
    Rating ScaleRating Scale MukutHazarikaMukut Hazarika ProfessorProfessor Department of EducationDepartment of Education Dibrugarh UniversityDibrugarh University
  • 2.
    Rating ScaleRating Scale •Rating scale refers to a scale with a set of pointsRating scale refers to a scale with a set of points which describe varying degrees of thewhich describe varying degrees of the dimension of an attribute being observed (Koul,dimension of an attribute being observed (Koul, 1997).1997). • A rating scale is defined as a technique throughA rating scale is defined as a technique through which the observer or rater categorizes thewhich the observer or rater categorizes the objects, events or persons on a continuumobjects, events or persons on a continuum represented by a series of continuous numeralsrepresented by a series of continuous numerals (Singh,1997).(Singh,1997).
  • 3.
    Types of RatingScalesTypes of Rating Scales • Numerical ScalesNumerical Scales • Graphical ScalesGraphical Scales • Percentage RatingPercentage Rating • Standard ScalesStandard Scales • Scales of cumulated pointsScales of cumulated points • Forced choice scalesForced choice scales
  • 4.
    Numerical ScalesNumerical Scales •Numerical scale consists of a sequence ofNumerical scale consists of a sequence of defined numbers which is supplied to thedefined numbers which is supplied to the rater or the observer. The rater or therater or the observer. The rater or the observer assigns to each stimulus, to beobserver assigns to each stimulus, to be rated, an appropriate number as definedrated, an appropriate number as defined or describes.or describes.
  • 5.
    Graphic ScalesGraphic Scales •The scales are presented graphically inThe scales are presented graphically in which descriptive cues corresponding towhich descriptive cues corresponding to the different scale steps are given. In thisthe different scale steps are given. In this scale, a straight line is drawn vertically orscale, a straight line is drawn vertically or horizontally with various cues to help thehorizontally with various cues to help the rater. The line may be segmented in unitsrater. The line may be segmented in units or continuous.or continuous.
  • 6.
    Percentage RatingPercentage Rating •This technique involves placing objects,This technique involves placing objects, persons, etc among different specifiedpersons, etc among different specified percentage groups or into differentpercentage groups or into different percentiles or quartiles. For example: inpercentiles or quartiles. For example: in the highest 5 per cent of the group, in thethe highest 5 per cent of the group, in the middle 25 percent of the group, in themiddle 25 percent of the group, in the lowest 10 percent, etc.lowest 10 percent, etc.
  • 7.
    Standard ScaleStandard Scale •A standard scale is one in which the raterA standard scale is one in which the rater is presented with some standards withis presented with some standards with pre-established scale values. Thesepre-established scale values. These standards usually consist of objects ofstandards usually consist of objects of same kind. For example: Handwriting-same kind. For example: Handwriting- standards, Portrait-matching, man-to-manstandards, Portrait-matching, man-to-man scale, etc.scale, etc.
  • 8.
    Portrait Matching ScalePortraitMatching Scale • In this technique a set of standards or verbalIn this technique a set of standards or verbal sketches or portraits for any given trait on whichsketches or portraits for any given trait on which rating is to be done, is prepared.rating is to be done, is prepared. • For constructing the verbal portrait a largeFor constructing the verbal portrait a large number of verbal statements describing that traitnumber of verbal statements describing that trait are collected and each is written in a separateare collected and each is written in a separate card.card. • They are then rank-ordered by a group ofThey are then rank-ordered by a group of experts.experts.
  • 9.
    • Then adesired number of sketches areThen a desired number of sketches are prepared which have the same average rank.prepared which have the same average rank. • The portraits are given to another group ofThe portraits are given to another group of judges for ranking them and mean rankjudges for ranking them and mean rank becomes the scale value.becomes the scale value. • Then the rater is given all the verbal portraits toThen the rater is given all the verbal portraits to read and he names the persons who belong toread and he names the persons who belong to the portraits.the portraits. • A person’s final rating is the average of all theA person’s final rating is the average of all the portraits values that have been assigned to himportraits values that have been assigned to him by all the raters.by all the raters.
  • 10.
    Scales of cumulatedPointsScales of cumulated Points • The rating score of a person, in theseThe rating score of a person, in these scales, is the sum of the individualscales, is the sum of the individual ratings or points assigned to all the itemsratings or points assigned to all the items of the scale.of the scale. (a)(a)The Checklist Method:The Checklist Method: (b)(b)The Guess-who Technique:The Guess-who Technique:
  • 11.
    • Here isone who is always helping others.Here is one who is always helping others. • Here is one who is always discouragingHere is one who is always discouraging others.others. • Here is one who is always happy.Here is one who is always happy.
  • 12.
    Forced-Choice RatingsForced-Choice Ratings •In forced choice rating scale the rater is given aIn forced choice rating scale the rater is given a set of attributes in terms of verbal statements forset of attributes in terms of verbal statements for a single item and he decides which one or ones,a single item and he decides which one or ones, represent the individual being rated mostrepresent the individual being rated most appropriately and accurately. The items ofappropriately and accurately. The items of forced-choice scale may have severalforced-choice scale may have several alternatives. The four alternatives are morealternatives. The four alternatives are more common in forced-choice scale. In this form, thecommon in forced-choice scale. In this form, the rater may be asked to select any two statementsrater may be asked to select any two statements which are most descriptive and representative ofwhich are most descriptive and representative of the persons being rated.the persons being rated.
  • 13.
    Example of Forced-ChoiceExampleof Forced-Choice ScaleScale • Is well informed in science.Is well informed in science. • Can apply scientific fact and theory toCan apply scientific fact and theory to practical situations.practical situations. • Create confidence in those with whom heCreate confidence in those with whom he deals.deals. • Explains the reasons for hisExplains the reasons for his recommendations.recommendations.
  • 14.
    Distinction between Forced-ChoiceDistinctionbetween Forced-Choice Rating Scale and other types ofRating Scale and other types of Rating ScaleRating Scale • All other types of rating techniquesAll other types of rating techniques consider one attribute at a time and placeconsider one attribute at a time and place the ratee in any one of a set of categories.the ratee in any one of a set of categories.
  • 15.
    Steps followed inConstruction andSteps followed in Construction and Standardization of Rating Scale:Standardization of Rating Scale: • Definition of traitsDefinition of traits : Each trait should: Each trait should be clearly defined so that traits may bebe clearly defined so that traits may be clearly and uniformly understood by allclearly and uniformly understood by all judges.judges. • Degree of traitDegree of trait: The degree of trait: The degree of trait should be defined.should be defined. • Reliability:Reliability: • Validity:Validity:
  • 16.
    Methods of DeterminingReliability:Methods of Determining Reliability: • Reliability of rating scales are determinedReliability of rating scales are determined by repeating judgments after a timeby repeating judgments after a time interval; by determining correlationinterval; by determining correlation coefficient between ratings of two or morecoefficient between ratings of two or more judges; and by determining relationshipjudges; and by determining relationship between judges’ ratings and self-ratings.between judges’ ratings and self-ratings.
  • 17.
    Factors affecting ReliabilityofFactors affecting Reliability of Rating Scale:Rating Scale: • Reliability of rating scale depends upon extentReliability of rating scale depends upon extent variation of judges’ ratings. Judges rating anvariation of judges’ ratings. Judges rating an individual on a specified trait will not alwaysindividual on a specified trait will not always agree as to his score or rank. The mean or theagree as to his score or rank. The mean or the median of all the judgments represent themedian of all the judgments represent the nearest approximation to the true rating.nearest approximation to the true rating. However, the averaging to be meaningful, theHowever, the averaging to be meaningful, the variation of the judges’ rating should be small.variation of the judges’ rating should be small. • Overt traits are more reliably rated than covertOvert traits are more reliably rated than covert traits.traits.
  • 18.
    • Reliability oftrait estimates is affected byReliability of trait estimates is affected by desirability or undesirability of the trait. Indesirability or undesirability of the trait. In self ratings there is a tendency forself ratings there is a tendency for individuals to overrate themselves inindividuals to overrate themselves in respect to traits regarded as sociallyrespect to traits regarded as socially desirable. In rating other persons,desirable. In rating other persons, especially friends, some judges may beespecially friends, some judges may be similarly influenced, even while trying tosimilarly influenced, even while trying to be careful. There is also a generalbe careful. There is also a general tendency toward generosity, rather thantendency toward generosity, rather than the reverse.the reverse.
  • 19.
    • Some personsare more accurately rated thanSome persons are more accurately rated than others. Extroverted individuals are more reliablyothers. Extroverted individuals are more reliably judged than introverted.judged than introverted. • Degree of certainty of ratings should be stated.Degree of certainty of ratings should be stated. With each rating, it is desirable to have theWith each rating, it is desirable to have the respondent state his degree of certainty that isrespondent state his degree of certainty that is very strong, strong, moderate, etc. Judges arevery strong, strong, moderate, etc. Judges are usually in closer agreement on ratings at theusually in closer agreement on ratings at the extremes because extreme deviants are mostextremes because extreme deviants are most clearly distinguishable from others. Thus suchclearly distinguishable from others. Thus such terms as ‘introverted-extroverted’ apply clearly toterms as ‘introverted-extroverted’ apply clearly to individuals who are manifestly at one pole or theindividuals who are manifestly at one pole or the other.other.
  • 20.
    Validity of RatingScale:Validity of Rating Scale: • The validity of a rating scale is the extentThe validity of a rating scale is the extent to which the judges’ understand theto which the judges’ understand the meaning of the traits being evaluated andmeaning of the traits being evaluated and the extent of accuracy in rating them.the extent of accuracy in rating them. • The principal indication of the validity ofThe principal indication of the validity of some rating scale is the extent to whichsome rating scale is the extent to which the person using them find them useful.the person using them find them useful.