Measurement ang Scaling Inourdaily life.as we are said to measure when we use some yardstick to determine weight, heig. 10 meaistire when we judge how well we like or some other features of a physical object. We song, «painting or the personilities of our friends. We, thus, measure physical objects as well abvrract concepts. Measurement is @ relatively complex and demanding task, specially so when; jitative characteristics are tas, concerns qualtans €or abstract phenomena, Other examples of qu homes. mielhgence, customer’ perception and brand loyatty etc. These charaeteristies are ah culled as constructs The feelings and perceptions of customers and employees are very importa! for a company fiiseas) o assign numbers in espectof properties of some objects. but itis relatively difficult respect of oters. For mstance, measuring such things as social conformity, intelligence, or mani {jn nuch esos and gues much cleseratention than measuring physical weigh jogical ae ors persons financial assets. In other words, properties like weight, height, etc..ce i be measured directly uh some standard unit of measurement, but it is not that easy to measur Forthe meaningful a ssessm properties ihe mot ation to succeed, ability to tand stress and the Hil {the qualitative characteristics 11s essential that they are also measured. 5.. QUANTITATIVE AND QUALITATIVE DATA { associating numbers or symbols to observations obtained sld be qualitative or quantitative: y quantitative data, For example, mean, 8 racteristics. Qualnative characterises can ‘chor must havewelear understanding of ft Fie observations on qualitative variable so person's marital status as 1, 2, 3.07 jowed or divorced. We can as well reco" cas | and 2 or perhaps as 59 and 60) In the tive) can be made mto numerical dill) the numbers we record as nominal dat /[Measurement and Scaling 67 {oniinal data is) numerical in name only, because they do not share any of the properties of the fiaatbers vic deal ar GRTIAATy BHR@NeATE. Por inten If we Te:s0ed raarital stgnus aw 1.2.3. or 48 stated above, we cannot write 4 > 2 or 3 <4 and we cannot write 3-1 =4~-2.143=4ord +2 In those situations when we cannot doanything except setup inequalities, we refer to the data as | data.)For instance, if one mineral can scratch another, it receives a higher hardness number | Mohs" scale the numbers from | to 10 are assigned respectively to tale, gypsum, calcite. fluorite, apatite, feldspar, quartz, topaz, sapphire and diamond, With these numbers we can write 5>2or 6 <9 as apatite is harder tan gypsum and feldspar is softer than sapphire, but we cannot write for example 10-9 = 5 ~4, because the difference in hardness between diam is actually much greater than that between apatite and fluorite. It would als that topaz is twice as hard as flu jond and sapphire 0 be meaningless to say ¢ simply because their respective hardness numbers on Mohs" scale are 8 and 4. The greater than symbol (i.e., >) in connection wit designate “happier than” “preferred to” and so on. —When in addition to setting up inequalities we can also form differences, we refer to the data as fmerval_data/Suppose we ate given the following temperature readings (in degrees Fahrenheit). 58°. 63°. 70°. 95°. 110°, 126° and 135°. In this case, we can write 100° > 70° of 95° < 135° which simply means that 110° is warmer than 70° and that 95° is cooler than 1352, We can also write for example 95° ~ 70° = 135° ~ 110°, since equal temperature differences are equal in the sense that the Same amount of heat is required to raise the temperature of an object from 70° to 95° or from 110° to 135°. On the other hand, it woul id not mean much if we said that 126° is twice though 126° = 63° = 2. To show the reas the first temperature becomes 5/9 (1 (63 ~ 32) = 17° h ordinal data may be used to as hot as 63°, even ‘on, we have only to change to the centigrade scale, where 26 — 32) = 52°, the second temperature becomes 5/9 and the first figure is now more than thr ree times the second. This difficulty arises from the fact that Fahrenheit and Centigrade scales both have artificial origins (zeros) i.e., the umber 0 of neither scale is indicative of the absence of whatever quantity we are trying to measure, When in addition to setting up inequalities and forming differences we can also form quotients sf. when we can perform all the customary operations of mathematics). we refer te such data as ratio data/n this sense, ratio data inéludes all the usual measurement (ord height, money amounts, weight, volume, area, pressures ete, The above stated distinction between nominal, ordinal nature of a set of data may suggest the use of particul, be quite alert about this aspect while measuring propert 2 fehl aa -H _ 5A CLASSIFICATIONS OF MEASUREMENT SCALES From what has been stated above, we can write U scales of measurement can be considered in terms of their mathematical properties. The most widely used cl: jassification of measurement scales are: (a) nominal scale: (b) ordinal scale: (c) interval scale: and (d) ratio scale. pmunalisca! preinabeco? jeterminations) of length, interval and ra data is important for the statistical techniques’, A researcher has to ies of objects or of abstract concepts. When data can be measured in units which are Kerchangeable ey. weights (hy rato scales), temperatures (by interval Jest that data rs sand ta he paramietnie and can be subjected fo most kinds of saistical and mathentatical processes, BUL When datas measured ay units which are not mnterchany alle. © 2 prontuct prcte ty be non-parametric and 1s suscepuble only to a linnted extent t Uy orsinal scales) the data is said= _A® Nominat sa Gomi ots \9 ye research Method! dat, Icis simply a system of assigning number symbols to ¢y, of numbers of baskey in order toJabelthg ze wsaferampre of this is the assignme: Players orderto Weftily then? Such numbers cannot be considered (0 be associated, border is of no consequence: the numbers are jUSt Convey aan ordered scale for their labely for the particular class of events aind as such have no quantitative value. No scales provide convenient ways of keeping track of people, objects and events. Oni c do much with the numbers invol ed{E nnot usefully average the num, on the back of a group of football pi ‘ul valuc. Nei, canrone usellly compare the numbers assigned one group with the numbers assigniy another, The counting of members in each group is the only possible arithmetic opera, ted ( USE Mode as 4 I scale is employed( Accordingly, we are tes' isno generally used measure of dispersion for nomi, tistical significance that can aumple, one ¢: J come bp with a nieani when 3 nomi ure of central tendency: Ther me scales. Chi-square test is the most common test of stat utilized. sures of correlation, the contingency coefficient can be work: out ment, It indicates no order or distay simply describes differene ply x them to categories. Nominstl dat counted data, 7 bout varying degrees of attitude, skill Mill very useful and are widy Nominal scale isthe least powerful level of measure! scale s. thus relationship and hay no arithmetic origin. A nominal between things by assign scale wastes uny information that we may ave understand used in surveys and other ex-past-facto research when de sub-groups of the population. / Ordinal scale: The lowes Jgvel of the ordered scale that is commonly used is the ordin scale ThoordinaTwale places events inorder, but there is no attempt to make the interal Of the scale equal m terme of some rule. Rank orders represent ordinals Hes and frequents used in rescarch relating to qualitative phenomena A students rank inh graduation class involves the use of an ordinal seale, One hits to be very arefull in makin les. For iistance. if Ram's position in his cla 11's position is Four times as gow _ 1, Ordinal scales only permitth no absolute values, and th syalerent about seores based on ordinal sca is 10 and Mohan’s position 1s 40, cannot be said that R asthatof Mohan The statement would make no senve at tl Jung of em from highest to lowest. Ordinal measures have 1 be said is that on: Lc: ise comparisons cannt real differences between adjacent ranks may not be equal. AMT tha person 1s higher or lower on the scale than another, but more prec be made. ns are restricted to various rank order methods significa ed to the non-parametne methods. ily are adjusted in terms of som cale, the inter naking the units equal. The units are equé ptions on which the rule 1s based. Interval scale Interval scale: In the case of inter rule that have been ¢ only in so far as one te, In spite of all this, nominal scales are n uta is being classified by maj 3\ &aes pz bave absolute zeros and as stich all minor traffic violations and all typing errors Ar ‘assumed to be equal in 2 t ACS zero" is not as precise wement and Scaling have an arbitrary zero, but it is not possible to determine for them what may be called an absolute 7ero or the unique origin. The primary limitation of the interval scale is the lack of a true Zero: it does not have the capacity to measure the complete absence of a trait or eristic. The Fahrenhcit scale is an example of an interval scale and shows similarities in what one can and cannot do with it, One can say that an increase in temperature trom 30° to 40" involves the same increase in temperature as an increase from 60° to 70, but one cannot say that the temperature of 60” is twice as warm as the temperature of 30 because both numbers are dependent on the fact that the zero on the scale is set arbitrarily at the temperature of the freezing point of water. ‘The ratio of the two temperatures, 30) and 60°, means nothing because zero is an arbitrary point Interval scales provide more powerful measurement than ordinal scales for interval scale also incorporates the concept of equality of interval. As such more powerful statistical measures cun be used with interval scales. Mean is the appropriate measure of central tendency, while standa of dispersion. Product priate and the generally used tests for statistical ard deviation is the most widely used measure moment correlation techniques are appro nif cance are the ‘t’ test and “F’ test. je: Ratio scales have an absolute or true zero of measurement. The tera absolute, as It was once believed to be. We can conceive of an absolute zero 'y We can conceive of an absolute zero of time. For example. the zero point on a centimeter scale indicates the complete absence of length or height. But an absolute zero of temperature is theoretically unobtainable and it remains a concept existing only in the scientist’s mind. The number of minor traffic-rule violations and the number of incorrect letters in a page of type script represent scores on ratio scales, Both these scales ‘“” of length and similar “an be ignificance. With ratio scales involved one can make statements like “Jyoti's™ typing performance was twice as good as that of "Reetu.” The ratio involved does have significance and facilitates a kind of comparison which i an interval scale. not possible in case of Ratio scale represents the actual mount of variables. Measures of physical dimensions such as weight, height, distance, etc, are examples. Generally, usable with ratio scales and all m: also be 1 statistical techniques are anipulations that one can carry out with real numbers can arried Out with ratio scale values. Multiplication and division can be used with this seale but not with other scales mentioned above. Geometric and harmonic means can be used as measures of central tendency and coefficients of variation may also be calculated. Thus, proceeding from the nominal scale (the least precise type of scale) to ratio seale (the most precise), relevant information is obtained increasingly. If the nature of the variables permits, the researcher should use the scale that provides the most precise description, Researchers in physical sciences have the advantage to describe variabley in ratio scale form but the behavioural sciences are generally limited to describe variables in interval scale form, a less precise type of measurementA measurement seale has to have certan aes. charac : : as geamness in measy characteristics under study. These qualities are described below: a (i) Validity Validity is the most ertieal erterion and indicates the degree to which an instrument measure, A it is supposed to measure. Validity can also be thought of as utility. In other words, validity, extentig which differences found with a measuring instrument reflect true differences among 4, being tested: Bur the question arises: how can one determine validity without diteet coffin, knowledge? The answer may be that we seek other relevant evidence that confirms the answey, have found with our measuring tool. What is relevant, evidence often depends upon the nature oj, research problem and the judgement of the researcher. But one can certainly consider three type, validity in this connection: (i) Content validity; (ii) Criterion-related validity and (iti) Construct vat, vient validity isthe extent to which a measuring instrument provides adequate cove, of the topic under study. Ifthe instrument contains a representative sample of the uniye, | the content validity is good. Its determination is primarily judgemental and intuitive, jo also be determined by using a panel of persons who shall judge how welll the measun instrument meets the standards, but there is no numerical way to express it. viterion-relared validity relates to our ability to predict some outcome or estimate ¢ existence of some current condition. This form of validity reflects the succes of measy used for some empirical estimating purpose. The concerned criterion must Possess following quali « Relevance: (A criterion is relevant if it is defined in ten measure.) «# Freedom from bias: (Freedom from bias is attaine' an equal opportunity to score well.) «# Reliability: (A reliable criterion is stable or reproducible.) i + Availabiliy: (The information specified by the criterion must be available.) (i ms we judge to be the prop 4 when the criterion gives each subj. In fact, a Criterion-related validity is a broad term that actually refers to @ Predictive ve predicting some Tutu and (ii) Concurrent validity. The former refers to the usefulness of atest in i performance whereas the later refers to the usefulness ofa test in closely relating to other measur of known validity. Criterion-related validity is expressed as the coefficient of correlation betwe. test scores‘ind some measure of future performance or between test scores and scores on anoth ‘measure of known validity. (ii) Construct validity is the most complex and abstract. A measure is said to possess constr: validity to the degree that it confirms to predigted correlations with other theoreti Propositions. Construct lity is the degree to which scores on a test can be accounle for by the explanatory San a eae iors Fordetemnining contract Vatiaij associate a set of other propésitions with the results received from using our measureme instrument. If measurements on our devised scale correlate in a predicted way with thes other propositions, we can conclude that there is some construct validity.he above stated criteria and tests are mec win, we may state that our measuring instrument aR an ean i ve to look for more information: is valid and will result in correct measurement; otherwise we shall have to look t and/or resort to exe! eof judgement. TR Niability The test of reliability is another important test of sound measurement. A measuring instrument is ble if it prgvides cons’ ple instriment r esults, Reliable measuring instrument does contribute to validity. but ‘dot be 4 valid instrument. For instance. a scale that consistently overwel objects by five kys., is a reliable scale, but it does nor give a Valid me way is not true Le.. a valid instrument is always reliable. Accordi validity. but it is cass satisfied by an inst factors are not interfer areli ure of weight, But the other zly. reliability is not as valuable as # lo assess reliability in comparison to validity. If the qui nent. then while using it we can be confident that t ing ality of reliability is fhe transient and situational Tyo aspects of reliability viz., stability and equivg aspectis concerned with securing consistent result wit and with the same instrument. We usually determine th Of repeated measurements. The egitivatence aspect considers how muck error may get introduced by different investigators or different samples of the items being studied. A good way to test for the equivalence of measurements by two i events. Reliability can be improved i lence deserve special mention. The stability h repeated measurements of the sam? person e degree of stability by comparing the results (0) By standardising the conditior ensure that external sources o the extent possible. That will i (ii) By carefully designed directi by using wained and motivat sample of items used, This Practicality ns under which the measurement takes pl f variation such as boredom. fatigue. etc., improve stability aspect. ‘ons for measurement with no variation from group to group. ed persons to conduct the research and also by broadening the will improve equivalence aspect. lace i.€., we must are minimised to The practicality characteristic of a convenience and interpretabilit '. Fro to be practical i.e., it should be eco! Suggests that some trade-off is need: measuring instrument can be judged in terms of economy, 1m the operational point of view. the measuring instrument ought nomical, convenient and interpretable. Economy consideration are also dependent at times upon economic factors, Convenience test suggests that the measuring instrument should be easy to administer, For this purpose one shou give due attention to the Proper layout of the measuring instrument. For instance, a questionnaire, with clear instructions (illustrated by examples), is certainly more effective and easier to complete than one which lacks these features. interpretability consideration is specially important when Persons other than the designers of théfestareTo Tmterpret the results, ‘The measuring ium order to be interpretable, must be supplemented by (a) detailed instructions for, ae bi a Ee (b) scoring keys; (c) evidence about the reliability and (d) guides for using the test an results. \4fs hola etree nee SAK ITY ewupte> ——____ iv) Accuracy oO formar 4 S ale The characteristic of accuracy of «measurement Seale means it should be a true representati the observation of underlying characteristic. For example, measuring with an “inch” scale will pr, accurate value only upto one-eighth of an inch, while measuring with *cm* scale will Provide m, curate value. “4 SOURCES OF ERROR IN MEASUREMENT a Measurement shouldbe precise and unambiguous in an ideal research study. This objective, howe,, i J; cane i s provided. Categories are ordered A F . . ategory i 4 jnulus ject O1 ‘ems of scale positions. The respondents se! gn Commonly used itemized rating scales are e oh whereit'# l lon abe A: we ig appronch wher é awe analy PP ons ct izing the ite! d by utilizing Like scales or summated scales 26 ee by at isciiates PENT me . i is Paticular item is evaluated on the basis ©! those items oF statement ‘ota score is high and those whose score 'S strament ig : ia ; discrimination test are included in the Falyew re eS ee Researeh Metadalnp 7 _Thus, summated scales consist of number of statements which express either a favourahy, unfavourable attitude towards the given object 10 which the respondent is agked to react. The respon. indicates his agreement or disagreement With each statement in the instrument. Each response given a numerical score. indicating its favourableness or unfavourableness, and the Scores are totay, to measure the respondent's attitude. In other words, the overall score represents the responden st position on the continuum of favourable-unfavourableness towards an ‘Most frequently used summmated scales in the study of social attitudes foll by Likert. For this reason they are often referred to as Likert-type scales. In a Likert scale, y ked to respond 10 each of the statements in terms of several degrees. Usually fj, degrees (but at times 3 of 7 may also be used) of agreement or disagreement. For example, wig asked to expressopinfon whether one considers his job quite pleasant, the respondent may respong) any one of the following ways: (strongly agree. (i) agree, (ii) undecided, (iv) disagree. (V) strong jow the pattern devia, respondent is a disagree ‘We find that these five points constitute the scale. At one extreme of the scale there is stron, agreement with the given statement and at the othe ereement, and between them |, intermediate points. We may illustrate this as under: strong disa stony Agee Undecided Disagree Stronaly wet) 8) @) ia) disagree (5) Fig. 5.3 Each point on the scale carries a score. Response indicating the least favourable degree of jo: satisfaction is given the least score (say I) and the mnost favourable is given the highest score (say 5) These score_values are normally not printed on the instrument but are shown here just to indicay the scoring patter, The Likert scaling technique. thus, assigns a scale value to each of the fi responses. The same thing is done in respect of each and every statement in the instrument. Th way the instrument yields total score foreach respondent, which would then measure the respondents favourableness toward the given point of view, If the instrument consists of, say 30 statements, th: following score values would be revealing. 30x 5 = 150 Most favourable response possible 30x 3=90 A neutral atitude 30x | =30 Most unfavourable attitude, The scores for any individual would fall bet 490. shows favourable opinion tothe given point of view, a score of below 90 would mean un} opinion and a score of exactly 90 would be suggestive of a neutral attitude. ween 30 and 150. If the score happens to be abow favourabk Procedure: The procedure for developing a Likert-type scale is as follows: (AS a first step. the researcher collects a large number of statements which are relevant te 4 the attitude being studied and each of the statements expresses definite favourableness unfavourableness to a particular point of view or the attitude and that the number favourable and unfavourable statements is approximately equal81] [Heasurement ‘and Sealing eu ered to a number of ér the statements have been gathered, a trial test should be administered toa numberof Sects In other words, a small grouf of people, from those who are goi fin ly. are asked to indicate F response to each Siatement by checking one of the categories of agreement or disagreement using a five point scale as stated above, hhe response to various statements are scored in such a way that a response indicative of the most favourable attitude is given the highest score of 5 and that with the most unfavourable attitude is given the lowest score, say, of 1, (iv) Prien the total score of each respondent is obtained by adding his scores that he received for separate statements. ») Thenext step is to array these total scores and find out those statements which have ahigh inatory power. For this purpose, the researcher may select some part of the highest and the lowest totarscores, say the top 25 per cent and the bottom 25 per cent. These two extreme groups are interpreted to represent the most favourable and the least favourable autitudes and are used as criterion groups by which to evaluate individual statements, Thig way we determine which statements consistently correlate with low favourability and which with high favourability. Only those statements that correlate with the total test should be retained in the final instrument and all others must be discarded from it, ‘ype scale in comparison to Thurstone-type “Can be performed without a panel of judges, Likert-type scale is considered more reliable by statement included in the in jecause under it respondents answer each Strument, As such it also provides more information and data than does the Thurstone-type scale (c) Bach statement, included in the Likert-ty ch pe scale, is given an empirical test for discriminating ability and as such, unlike ‘Thurstone-type scale, the Likert-type scale permits the use of statements that are not manifestly related (to have a direct relationship) to the attitude being studied. (d) Likert-type scale can easily be used in respondent-centredand stimulus-centred studies ie., through it we can study how responses differ between people and how responses differ between stimuli. Likert-type scale takes much less time to construct Deis frequently used by the students of pinion research. Moreover, it has been reported in various research stuiles: that then n high degree of correlation between Likert-type scale and Thurstone-type scale (e) Limitations: There are several limitations of the Likert-type scale as well. One ‘important limitation is that, with this scale, we can simply examine whether r respondentyare more or less favourable to a topic) but we cannot tell how much more or less they ar&."Phere is no basis Tor belief thatthe five A.L. Edwards and K.C. Kenn Journal of Applied Pyychology, 30,72 comparison of the Thurstone and Likert techniques of atitude seale construction 3, 1946.a nn Research ene Research MeTiatiy does not rise to a stature more than that of an ordinal scale, whereas the designers of Th igners of Thury Ther disadvantage is that ofc ten}, since a given total score can, scale claim the Thurstone scale 10 be an interval scale. One total score of an individual respondent hay Little clear m tired by a variety of answer patterns. tis unlikely that the respondent ean validly reset to ay, Sinemenr ena printed form mthe absence of real-life quhtying situations, Moreover. there ey d poses that people nay answer according 10 what they think they should feel rather than fe ther do feel ” This particular weakness oF the Likert-type seale is met by using @ cumulative sa, which we shall take up later m this chapter. tanons. the Likert-type summated scales are regarded as the Most usefy ble to compare the respondent's score with a distribution of scores fi. nme In pute ofall the lr 4 situation wherein 1118 po aaene defined group. They are equally useful when wwe are concemed with a progi Ghange or improvement in which case we €an use the scales to measure attitudes before and af, ~ the programme of change or umpeovement in order 10 assess whether our efforts have had », deaied effects. We can as well correlate scores on the scale to other MeaSUTS without any conc, vn the abecloe salug of wiht Lavounsble and what ts unfavourable All this accounts for y, popula of Likerttype wales in social states relating fo measur pf attitudes. ts) we the SD. scale developed by Charles E. Osgood. G.J. Suci and Ph, han 1957) 1san attempt to measure the psychological meanings 3 object to an individu: 1 hused on the presumption that an object can have different dimensions of comnotati. m h .elncaied im mulladimensional property space, oF what can be called the seman. TST ale This scaling consists of a set of bipolar scales, usually of points. by which one or more respondents rate one or more concepts on each seale item, For instane he $.D_ scale tems tor ves for leadership position may be shown as under: comash unsuccessful _ Lenient wes Light J Cold | Regressive | Weak +1 False . * Unsocisble r83 [Measurement and Scaling Candidates for leadership position (along with the concept—the “ideal” candidate) may be compared and we may score them from +3 to —3 on the basis of the above stated scales. (The letters, E, PA showing the relevant factor viz. evaluation, potenc along the left side are not written in actual scale.) and activity respectively. written in actual scale. Similarly the numeric values shown are also not write Osgood and others did produce a list of some adjective pairs fo attitude research purposes and concluded that semantic space is multidimensional rather than unidimensional. They made sincere efforts and ultimately found that three factors. viz., evaluation, potency and activity, contributed most to meaningful judgements by respondents. The evaluation dimension generally accounts for 1/2 and 3/4 of the extractable variance and the other two factors account for the balance. arious steps involved in developing S. follows: Tl) First of all the concepts to be studied are selected, The concepts are usually chosen by ae personal judgement, keeping in view the nature of the problem. (b) The next step is to select the scales beari the c1 Procedure: Va D. scale ar ing in mind the criterion of factor composition and ‘crion of seale’s relevance to the concepts being judged (tis Common practice to use at least three scales for each factor with the help of which an average factor score hive to be worked out). One more criterion to be kept in view is that scales should be stable across subjects and concepts. (c) Then a panel of judges are used to rate the various s timuli (or objects) on the various selected scales and the responses of all judges would then be combined to determine the composite scaling. To conclud “the S.D. has a number of specific way to secure attitudes from direction and intensity. The to meanir advantages. It is an efficient and easy 4 large sample. These attitudes may be measured in both tal set of responses provides a comprehensive picture of the = OF an object. as well as a measure of the subject doing the rating. Itis a standardised luc that is easily repeated, but gseapes many of the problems of response distortion found with more direct methods. apel Scale + le was developed by John Stapel. This is a unipolar rating scale w ed from —S to 45, Th ith usually 10 categories s scale does not have zero or the neutral point, Respondents rate how W describes the object by selecting the appropriate number. Positive number means the term describes the object accurately, while negative number implies that the term describes the object inaccurately. +5 means the highest degre accuracy ~ each ter of accu while 5 meany the highest degree of in For example. consider the phrases (i) Tasty food. (ii) Fast § nd (ii) Good ambience for arestaurant. A respondent is asked to rate how accurately these terms or phrases describe a specified restaurantR Th Methoa, , ee) 7 +5 +4 +4 N #3 43 #2 #2 Gy +1 Service Good Ambience Tasty Food -1 1 “1 -2 ae 3 -3 -3 -4 -4 -4 -5 -5 -5 ervice’ and +3 for ‘good ambience accurate with a minimum degree, The respondent selects +1 for ‘tasty food,” = 2 for ‘fa This means the respondent thinks that the phrase ‘tasty food” accuracy. phrase “fast service’ is some what in accurate, while “good ambience’ is sufficieny accurate, So according o the respondent the ambience of the restaurant is good, food tast is okay, service is slow This method is applicable when the responses are rated on a single dimension. The method, very economic and data can be collected over telephonic interview also. Some researches think th, the metiod is confusing and of not much use. ingle/Multiple Category Scales 1ese scales are also known as dichotomous scales. Here we have two or more mutually exclusiy, responses. For example. ~ *Yes’ and ‘No’, “True” and ‘False’. The respondent has to choose onl, ‘one out of the given categories. It would be clear from the following example questions of; ‘questionnaire: {Do you play cricket ? > Yes > No 2 What is your marital status? » Unmarried > Married » Divorce » Widower 3 What is your employment type? » Salaried » Self employed professional » Retired » Self employed business » Student » Home mi ce more than one responses, For example, r What do you own? Car 0 Computer oO85 | Measurement and Scaling 2. What are you interested in? Music Oo Films oOo Sports TC —CsReading oO Travelling oO ‘What do you like to do on internet? Social networking [7] E-mail oO News reading o Knowledge gaining = [7] Gaming Te. It should be ensured that the provided choice exhaust almost all possible answers of the asked question. The choice ‘others’ may be given along with other choices. The collected data is on nominal scale. The method is easy and very popular on internet surveys. wer Frequency Scale # scale is used when the respondent is unable or unwilling to give the exact numbers in the answer. For example, How often do you eatout? 1. Frequently 2. Sometimes 3. Rarely 4. Never This scale provides only an approximation of frequency and so the data is on ordinal scale. Dimensions: In unidimensional scale, onl: multidimensional scale considers that an obj the popularity of a restaurant can be measur multiple dimensions like food taste, 'y one attribute of the object is measured, While, a ject is described with several dimensions. For : example, ed by a single measure food taste. It can be defined on service, cleanliness, ambience, etc, Number of scale categori collected data. However, categories is Larger number of scale categories will it would become inconvinient for the respo: provide better precision in the a trade-off between precision and convinience. It shoul ndent. So, the number of scale Id be chosen with care, » number of favorable and unfavorable categories qual in an unbalanced scale. The scale should be balanced in general collect an unbiased opinion. However, if the distribution of responses i v likely to be skewed on one direction, an unbalanced scale having more categories on the opposite direction is used Balanced and unbalanced scales: In balanced scale, are equal, while they are une 5.9 MULTIDIMENSIONAL SCALING Multidimer ional scaling (MDS) is relatively more complicated scaling device, but with this sortor scaling one can scale objects, individuals or both with a minimum of information, Multidimensional scaling (or MDS) can be characterized as a set of procedures for portraying perceptual or affective dimensions of substantive interest, MDS is used when all the variables (whether metric or non- metric) in a study are to be analyzed simultaneously and all such variables happen to be independ