MODULE – 4 Identify the source of information Sources & Collection of data. > Select the data collection approach. > At this stage there is temptation to organize a field survey to collect me data. >Field survey should be resorted to only when all the secondary sources of data collection has been exhausted. >Survey is comparable to surgery. >It should be resorted to only after all other possibilities have been exhausted. Secondary & Primary Research • Secondary research or secondary data is any data that has been gathered earlier but which has not been specifically collected for the current marketing research. 1 • Primary research or primary data is data which has been collected first hand. This involves collecting data specifically for the purpose of the study on hand from the actual sources such as consumers, dealers, industries or other specific entities involved or other specific entities involved in the research. Secondary Data. • Secondary data resources include published sources of data, periodicals and newspaper reports, Internet etc. • • Primary data collected by one person becomes secondary data for another. It is sometimes possible to do a lot of good iiry research and get useful data. Yet the study may require primary data For instance, research report on a similar topic may yield relevant data. 2 But, the research on hand may require a lot of current data which may not be readily available from iiry sources. E.g. (1) Customer satisfaction level for a specific product / branch may not be reported any where in iiry sources. E.g. (2) The effectiveness of a particular advertisement may the evident from the sales which follow. But why people liked the advertisement may not obvious and can only be ascertained through primary research/interviews with consumer. E.g. (3) The methodology for the iiry data already collected may be unknown and therefore we may be unable to judge the reliability 3 & validity of the data Limitations/Disadvantages of iiry data 1. Reliability of data Reliability of iiry data depends on the purpose for which the data had been collected and it may not match the requirements of the present study. 2. Accuracy of data What degree of accuracy was employed? Was that achieved? 3. Suitability of data The data suitable for one enquiry may not necessarily be found suitable in another enquiry. Eg. (1) Units of measurement used in iiry data may be different from the one needed in the present research project. Say, size of firm can be expressed in several ways as in 4 Does it match the requirement of the current study? No. of employees; Capital & technology employed ; Gross sales & profit ; Net profit etc may not match “size” present study E.g.(2) Class boundaries may be different form the need of the researcher causing misfit. sufficiency or 4. Adequacy or sufficiency of data Sample size and depth of analysis of the iiry data may not suit the researcher’s requirement in the present study 5. Obsolescence of data Business/ marketing environment is dynamic wrt consumer taste, economic conditions, business trends etc., that the data may be a misfit in the current research requirement. Eg A study done on the need for telegram for communication may be obsolete in the present context. 6. Relevance of data 5 standadard in With changing technologies the iiry data may not be quite relevant in the current research context. Eg. A study on manual typewriters electric & & subsequent versions of wrt electronic typewriters consumer behavior may not be relevant in the study on CB w.r.t personal computers. 7. Proprietary & confidentiality barriers may prevent access & free use of such data desired by researcher. > The above mentioned limitations should not prompt the researcher to blindly discard the use of iiry data, especially if they are readily available from authentic sources and are also suitable or adequate for the present study > In that case, it will not be economical to spend time, energy & money in 6 field surveys for collecting primary information., > At times, there may be wealth of usable inform,. in the already which can be used available iiry data by an intelligent researcher but with due precaution. Advantages of iiry data 1) Low cost of data collection. At times they may offer insight and relevant sufficient problem. Eg. Size of the market for a particular product - the quality of the study may preclude the need for a new primary study 2) Saves time and helps in completion of research project. 3) Familiarity with the data may indicate deficiencies, gaps & pitfalls in the 7 information to solve which the may be marketing prompt previous study which may help refine primary data collection process, study design, survey design, data analysis in the proposed research 4) As the researcher explores the availability of secondary data relevant to his project, his understanding of the problem improves leading to some changes in his original ideas. 5) iiry data can be used as a basis for comparison with the primary data just collected. Sources of Secondary data iiry data can be obtained internally i.e. within the firm, or externally i.e.,from one or more outside agencies. Internal Sources of iiry Data 8 are those which are generated within the firm. Internal sources of iiry data • • • • Eg. Accounting Records. Sales Force Report Miscellaneous Reports Internal Experts Balance Sheet, P&L statements., Accounting Records measure performance & growth Sales Invoices can give details of customer, his address, items & quantity discounts allowed, total sales in the regon / territory during a period etc. Sales Force Reports In a proper format it can provide a wealth of information, sales force can be encouraged to provide accurate & comprehensive information, their comments & remarks on 9 the monthly sales, rapport with dealers & retailers, competitive activity, etc., Miscellaneous Reports Any etc., Internal experts Product Managers, Managers, PROs, Marketing. Res. Advertising Managers study done on earlier marketing problems of the company, special audits may home specialized knowledge relevant to marketingl problems. External Sources of Secondary data Govt. Publication 1) Registrar General of India – Population Census Report 2) Central Statistical Organisation (CSO)– Statistical Abstracts of India. industries Annual Survey of 10 - Estimates of National Product savings & capital formation 3) Planning commission reports Basic statistics relating to Indian economy 4) Reserve Bank of India Reports Wholesale price index Non Government. Publication 1) Indian Cotton Mills Federation Statistics on cotton textile industries 2) Stock exchange directories financial a/c, key profitability ratios of listed companies 3) Bureau of Coffee Board Coffee production statistics - 11 4) Coir Board, Rubber Board, Silk Board etcAnnual reports 5) The Steel Authority of India Ltd (SAIL) Stats for iron or steel I ndustry 6) FICCI – Fed. Ind. Ch. Com & Industry. Periodic Bulletins IACC - Ind. Am. Ch. Comm & Publications ACCII – Ass. Ch. Comm. &Industry of India QUESTIONNAIRE • • It is a communication method It is a tool or instrument to gather data from respondents situation • Definition in a survey 12 A questionnaire is a set of questions designed to generate the data necessary to accomplish a research projects. • Questionnaire can be classified into i) Structured non disguised /direct disguised non /indirect disguised four main types Questionnaire ii) Structured Questionnaire iii) Non -Structured /direct Questionnaire iv) Non-Structured disguised /indirect Questionnaire Questionnaire is a formal A structured list of questions • • Framed so as to get facts The interviewer asks the questions strictly in accordance with a pre – arranged order. 13 A structured Questionnaire is of two types a) disguised and b) non – disguised When the object or purpose of the survey is revealed it is called non – disguised type When the object or purpose of the survey is undisclosed / not revealed it is called as disguised type Structured Questionnaire non – disguised is one when the listing of questions is in a pre-arranged order and where the object of enquiry is revealed to respondent. Most marketing research studies use this type of questionnaire Advantages of structured non – disguised questionnaire or Advantages of a questionnaire 1) It facilities the collection of information in a systematic and orderly manner as 14 the questions have been formulated in advance 2) Since the questions asked by each interviewer happens to be identical and are asked in the same order, the information is generally not distorted on account of varying characteristics of different interviewers. 3) A structures approach calls for a straightforward and simple approach on the part of interviewers . As such, even less qualified interviewers can be deployed in the survey. 4) Such a questionnaire makes it far easier to edit, tabulate & interpret the data. 5) Such a questionnaire can be pretested so that suitable modifications can be made in phrasing sequence or both. 15 question or in the LIMITATION Such a questionnaire to elicit is not in a position answers from respondent re. motivation & personal reasons for his behaviour. This is done by disguised questionnaires in qualitative studies. Eg. In-depth interview Characteristics of a good questionnaire 1) Has a good heading along with object of enquiry clearly stated. Carries necessary instructions Well arranged sequence of questions answers capable of tabulation 2) 3) 4) Elicits objective Has clarity 5) 6) Has limited no. of questions 16 7) Has good format-size, with appearance, printing margin, spacing to look good on good quality paper good motivating respondents to answer. Objectives of a Questionnaire 1. It must translate the information needed into a set of specific question that the respondents can & will answer 2. It must motivate & encourage the respondent to become involved, cooperate & complete the interview. 3. It must minimize response error arising from inaccurate, misrecorded or misanalysed answers. Questionnaire Design process - Steps STEP 1) Specify the information needed 2) Specify the type of interviewing method 3) Determine questions 17 the content of individual 4) Design the questions to overcome the respondents’ inability or unwillingness to answer 5) Decide on the question structure 6) Determine the questions wording 7) Arrange the questions in proper order 8)) Identify the form & layout 9) Reproduce the questionnaire in limited nos for pretesting 10) Eliminate bugs by pretesting and bulk reproduction 1. Specify the information Needed. (i) Review research questions, hypothesis and objectives to ensure information to be obtained addresses the problem. 18 (ii) Prepare a dummy table to describe how the analysis will be structured when the data have been collected. (iii) Check if questions match target group eg. Questions for college student may be inappropriate for housewives. iv) Language – What language is the respondent respond. The going to understand & must be so (iv) questionnaire designed that it can be used in any language. Eg. A Questionnaire printed in English could be administered to the respondent in the local language he speaks, by a trained interviewer who records the answers in the questionnaire. 2. Type of interviewing method 19 In personal interviews, respondents see the questionnaire & interact face to face with the interviewer. Thus lengthy, complex & varied questions can be asked. In telephone interviews, respondents do not see the questionnaire. Hence, questions have to be short & simple. Mail questionnaires are self administrated, so questions must be simple & detailed instructions must be provided. E-mail Questionnaires also have to be simple. 3. Individual Question content Is the question necessary? Eg. Questions unrelated to the problem. Are several questions needed instead of one eg. – Do you think Coca Cola is tasty & refreshing soft drink? – incorrect barreled question Instead 2 separate questions may be asked 20 double i) Do you thick Coca Cola is a tasty drink? ii) Do you think Coca Cola is a refreshing drink? 4. Overcoming inability & unwillingness to answer. Is the respondent informed or familiar with the topic ? Eg – Asking husband about monthly grocery bill when wife does the purchase. Can the respondent remember Eg- What did you have for lunch last week? Can the respondent be expressive? Eg- Describe interior of a dept. store you would like to shop in? Such a question may not get the desired response if respondent is not expressive. 5. Choose question structure 21 A. Unstructured Questions - Open-ended questions that respondents answer in their own words. EgWhat is your occupation ? ---------------------Why do you think people shop in department stores.?----------------------------------------------------------------------------------------------------B. Structured questions i) Multiple choice questions Please tick the brand of Shampoo you normally use. a) Clinic, b) Chik, c) Sunsilk & shoulder, e) Any other. Please specify__________________ii) Dichotomous Questions – has only two response alternatives. Eg – Yes, No ; Agree, Disagree etc. d) Head 22 Sometimes neutral alternative needs to be included Eg- Don’t Know, both or no opinion Do you intend to buy a new car within the next six months? ____yes’ _____No, _______ don’t know iii) Scales Do you intend to buy a new car within the next six months? Definitely Probably Will not buy will buy 1 5 2 Probably Definitely will not buy will buy 3 4 Undecided 6. Choosing question wording - avoid ambiguous words 23 eg. Do you think distribution of soft Have you ever faced out of Yes ; No ; This can be drink is adequate? Instead: Drink ? stock situation of Brand X soft further simplified - use ordinary words Simplify – Do you think soft drinks are readily available when you want to buy them? - avoid leading questions Do you think as a patriotic Indian you should buy Gilette? Better- When you buy a razor do you choose Indian or Imported brands? 7. Determining order of questions 1. Opening questions should be simple, interesting & check eligibility of the candidate to answer questions 24 Eg- Suppose the respondent is required to be a graduate, then the opening question can eliminate non graduates. 2. Difficult questions to be included later & should be a few in number 3. Logical order – sequence and flow to make interview smooth 4. Avoid crowding of questions that confuse respondent 8. Format & layout - Format, spacing & positioning should be comfortable to read. - Questionnaire may be in several parts if long. - Preferably it is best to have just one page 9. Reproduction – in good quality paper for pre-testing & main Survey 25 10. Pretesting respondents improving questions > – for The the testing of a of by questionnaire on a small sample of purpose the questionnaire identifying & eliminating problematic Misunderstood questions to be reworded or eliminated. > Length to be trimmed. > Information content to be reviewed. SCALES OF MEASUREMENT USED IN BUSINESS/MKTNG RESEARCH Nature System Nos. have certain basic properties. First nos. follow a rank order 9 > 3, 12 < 24 and characteristic of Number 26 Second The difference pairs of or nos intervals can be between compared. The Interval between 6 & 3 is the same as the interval between 9 & 6. The Interval between 36 & 24 is twice the interval between 12 & 6. Third We can divide one no by another and interpret the relative magnitudes of the two nos by the resulting ratios. 6/3=2 Thus 6 is twice as large as 3. 12 is one third of 36 because 36x1/3=36/3=12 Fourth We can compute & interpret numbers because the number system has Unique Zero Point The properties of numbers are of special significance in the context of 27 quantifying survey responses esp. when measuring subjective aspects such an emotions, attitudes etc on a scale. Scale -A device constructed with a set of members that serves as a yardstick or means of measurement. Measurement is the assignment of numbers to responses based on a set of guidelines. Eg yes is 1 ; No is 2 Scaling Techniques. Marketing research is often concerned with the behavior of consumers. variables. Why is this important ? They want to assess their attitude to different 1) If the attitude of the public towards a company or its product is unfavorable, the company cannot survive too long. 28 So, a company should know public’s attitude towards it. Also, the company must ensure that people have a favorable attitude towards its present products/services. The company must also look into the future to anticipate the preference of the public. Thus, a study of attitudes becomes very relevant to marketing researchers. 2) Attitudes also plays a major role in advertising. A marketing researcher would like to decide the copy message the advertisement should carry, how it should be conveyed, the effective media that can take this message to the consumer. Eg. Complan with height, Junior Horlicks with growth in intelligence with an interesting storyline suited 29 to TV 3) A manufacture may embark on an attitude survey to reveal the likes & dislikes of consumers towards certain brands that would be helpful in his product development. Attitude The sum total of a person’s inclination and feelings, prejudice or bias, preconceived notions, ideas, fears, threats and convictions about any specific topic. Attitude is, therefore, a subjective and personal opinion. Attitude influences behaviour. Purchase decisions are based upon attitudes. Attitudes can change over time Types of Scales / Scaling Techniques There are basically four types of scales of measurement or scaling techniques 1. Nominal Scales 3. Interval Scales 2. Ordinal Scales 4. Ratio Scales 30 1. Nominal Scales Nominal sales are more widely used than any other scale for research in social sciences. In a nominal scale, numbers serve as labels etc. to identify persons, objects events, different categories of responses Eg. ( ) A. What is your gender 1. Male Female ( ) 2. B. Which of the following influences your purchasing decision the most. 1.TV media ( ) ) 2. No labels:31 ( ) 2.Radio ( ) 3.Print ( 4.Interntet ( ) 1. Yes ) C. Do you like ice cream? ( 4. Any set of nos can be used as 05 MBA 26 05 Signifies yearof intake and 26 subject code of BRM Arranging books in a library – Physics 48, Chemistry 92 etc Room No 406 in a hotel signifies 4th Floor 06 – room no. Limitations Nominal scale is the simplest of the four scales. It does not imply any ranking of the responses. Only serves as labels The nos. do not have any arithmetic origin ie., no zero origin. Hence no mathematical operation is possible There is no distance relationship in the scale employed. Hence, unsuitable for determining relationships. However, Nominal Scales are useful in research (esp. Exploratory). where 32 broad dimensions (frequency) of variables need to be determined. ie., How many said YES, how many said NO Statistical Implication: Calculation of Std.Deviation and Mean is not possible. Mode is the most appropriate central tendency scale. to determine the most frequently checked response in nominal 2. Ordinal Response. Ordinal Scale posses of /Order all the Scaled unique plus scales characteristic nominal scales possess the property of rank order. Hence, more powerful than nominal scales. Eg. (1) Consider the followin Question How long do you spend reading new papers on a typical weekday ? 33 A. B. do --C. do --D. do --- Less than 5 min. 5 min to less than 15 min. 20% 15 min to less than 30 min. 25% 30 min or more. 5% Actual --------- response 40% Rearranging above and ranking them from lowest to highest response rate, we have 1. D 2. B 3. C 4. A 5% 20% 25% 40% Eg. (2) Five types of house holds with monthly income House hold Income (Rs. ‘000) 6 A 10 5 B C D E 12 11 34 If the households are ranked highest to the lowest we have 1 2 3 4 D Rs 12,000 Rs 11,000 Rs 10,000 Rs 6,000 5,000 E B A 5 C The point to emphasise here is that the exercise is simply rank ordering lowest to highest or highest variable Consider to lowest the given and does not reflect the internal another example Scale Response Ranking +2 Lowest +1 2 10% 4 25% 35 differences between the categories. where % acceptability of a soft drink is involved. I strongly like to I like it somewhat To Highest 0 5 1 5% 45% 3 I am indifferent -1 -2 15% I dislike it somewhat I strongly dislike it Please rank the following domestic airlines from 1 to 5, being most preferred and 5 least preferred in terms of price. Rank Indian Spice jet Deccan Jet Airways Kingfisher 3 2 1 5 4 100 Respondents (nos) 200 250 300 150 Mode & Median 300 Two measures of central tendency are meaningful for ordinal-scaled responses: Mode & Median (category percentile response falls) 36 where 50th Difference between Nominal & Ordinal Scale - In nominal scale numbers can be interchanged as it serves the purpose of Numbers in only labeling and counting. cannot be interchanged 3. Interval Scales >This has all the properties of ordinal scale plus the differences between scale interpreted. >This characteristic makes interval scaled responses more powerful than nominal or ordinal scaled responses. >Strictly speaking, variables such as attitudes, opinions and preferences cannot be quantified to yield exact interval scales. However, within the scale equal intervals are assumed between categories. 37 Ordinal Scale have more meaning and values can be meaningfully >So an a particular scale, equal differences indicate equal differences in value with regard to that scale only. EgHow likely are you to buy a new automobile within the next six months? (Please tick the most appropriate category) 1. 2. 3. 4. 5. 6. Will definitely not buy Extremely unlikely to buy Unlikely to buy Likely to buy Extremely likely to buy Will definitely buy >Values assigned to this set of responses run from 1 to 6 (these nos. can also be in reverse) and strictly speaking from only an ordinal scale. >But they are considered to be interval scale under the assumption that respondents will see the connection between adjacent responses and 38 will treat the differences between adjacent response categories as being equal, especially since the categories are physically separated by equal distance. >Unfortunately, verifying whether the respondents perception of the differences between adjacent response categories are equal is very difficult, if not impossible. >However, the MR & BRM, such scales are more often (than not) treated as having interval properties. Consider above eg: Response as follows Resp. Category 1 2 3 No. & % of response 10 (5%) 10 (5%) 70 (35%) 39 4 5 6 60 (30%) 20 (10%) 30 (15%) 200 100% > Interval scales are more powerful than nominal and ordinal scales. > They to apply. >In a particular interval scale, equal differences indicate equal differences in value with regard to that scale only. >Numbers forming an interval scale, in addition to possessing ordinal scale attributes, permit computation of the Mean and Standard. Deviation. are quicker to complete & researchers find them more convenient 40 In above example, the mode & median are 3 & 4 respectively. The Mean is 3.8 on a scale of 1 to 6. (It is assumed that the unit of measurement is constant throughout) >Suppose, we administer the same question to a different sample of respondents with lower income levels than the earlier sample. Say mean is computed to be 1.9. Although 3.8 is twice as large as 1.9. we cannot say that the first sample is twice as likely to buy a new automobile than the second sample. -This is because, the starting or zero point is arbitrary. Hence, the two samples cannot be linked. 41 -Compare with a thermometer that has a common zero point for all the measurements. Workings = Calculation of Mean. (1x10+2x10+3x70+4x60+5x20+6x30) 200 = = 3.8 response % 1 2 10 10 10 (5%) 10 (5%) 42 / (10x20+210+240+100+180) / 200 2nd Sample No & 120 (1x120+2x40+3x10+ (60%) category 4x10+5x10+6x10) – 200 40 (20%) =(120+80+30+40+50+60) 3 (5%) = 1.9 (5%) 5 6 4 200 (100%) Statistical Implication: We can compute mean, median, mode Difference between Interval and Ordinal Scale Ordinal Scale gives only the ranking of the alternatives, namely one is greater than the other. other 4. Ratio Scales > Ratio Scales possess the power of the preceding three scales plus the concept of absolute zero or origin. > Thus they have the order, distance/interval and unique origin. > Ratio scales are the most superior amongst all the scales. Interval Scales provide information about the difference between one and the 43 > Quantified responses forming a ratio scale analytically are the most versatile. >The ratio numbers on these scales have meaningful interpretation. >Data on certain demographic or descriptive attributes, if they are obtained through open ended questions, will have ratio scale properties. Consider following questions: (i) What is your annual income before taxes? Rs. ________________ (ii) How far is your workplace from your home? _________________km. Answer to these questions have a natural, unambiguous starting point, namely zero. 44 Since the starting point is not chosen arbitrarily, as in interval scale, computing & interpreting ratios make sense. In this scale it is possible to say how many times greater or smaller one object is compared to the other Eg. – A respondent earning Rs. 40,000 p.m. is earning twice as much as a respondent earning Rs. 20,000/- p.m. >Non metric data - are data with only nominal or ordinal properties > Metric Data – are data with interval or ratio properties > The interval between nos. have precise meaning for metric data but not for non metric data ATTITUDE SCALES Attitude is the predisposition of the individual to evaluate some object or 45 aspect of his world in a favorable or unfavorable manner. Attitude Scales are sets of rating scales constructed to measure one or more dimension of an individual’s attitude toward some object (issue) Attitude scales are constructed using Likert, Semantic Differential, Paired Compassion & Thurston Scales Criteria for a Good / Effective scale 1. be Reliability- A scale or test is said to reliable to the extent that repeat made by it under measurements constant conditions will give the same result (assuming no change in attitude). 2. Validity – The scale measures what it sets out to measure and any 46 differences differences. represent scale’s the true to 3. Sensitivity-The ability detect subtle / minute differences in the attitudes being measured. It should have sufficient numbers range to facilitate medasurement of fine variations in attitude. This is achieved better by a multi item scale. 1.Likert Scale > It has an ordinal scale format . > It enables ranking of attitude > It consists of a series of attitude related statements (multi-item scale format) Respondents are asked to rate these statements on a 5 point item scale. 47 Strongly Agree agree Disagree Disagree agree Nor Disagree Neither Strongly 1 4 Maruti 800 is the best car ( ) developed by Maruti Udyog It is light but sturdy ( ) ( to take on bad roads in India ) 5 2 3 ( ) ( ) ( ) ( ) ( ) ( ) ( ) 48 It has got excellent fuel efficiency ( ) It meets middle class aspirations ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) After sales service is countrywide & ( ) ( ) excellent Cheaper R & M ( ) ( ) > The statements are separate ideas yet linked to a common object (see 49 ( ) ( ) ( ) ( ) ( ) ( ) above example) > Single – item scale formats have single statement format. >The central tendency of single statement indicates the average attitude > The central tendencies in multi-item format gives a consolidated avge > The central tendencies in format gives a consolidated average. The individual averages & consolidated average may indicate a pattern in the relationship – pictorial profile. > This is one of the widely used scales in MR. 2.Semantic Differential Scale > It is an Ordinal Scale similar to Likert Scale. > It consists of a series of items to be rated – multi item scale. 50 multi – item > Single item scale may be a stand alone attitude statement. > Items are presented as a bipolar adjective phrases or words at two extremes of the scale. > The scale in – between may have 5 – 7 points. > The respondent marks one category for each scale that best represents the attitude. > A more common application of the semantic differential scale is to develop a pictorial profile of the attitude based on mean ratings. > The profiles exhibit the relative strengths & weaknesses as perceived by customers. This is one of the widely used scales in marketing research. Maruthi 800: 51 1.Best car Worst car 2.Sturdy car Weak car 3Excellent Fuel Worst Fuel Efficiency Efficiency 4 Meets Doest not Middle Class Middle lass aspiration aspiration 5.Excellent Worst 52 After sales After sales Service Service. 6.Cheap Expensive R&M M Thurston Scale > It is an Ordinal Scale > This applies the logic that attitudes are best measured when comparative statements are arranged one below another in a sequence; To get neutral statements towards halfway point ; and Extreme statements opposite in character beyond the halfway point. > The statements are arranged in 11 piles for the final scale. These 53 R & statements are distilled version of many statements one would have studied and finally short listed. Eg Attitude towards TV commercials can be studied by using Thurston scale as follows: Scale Value 1) 2) 3) 4) 5) 6) 7) 8) All TV commercials should be prohibited by LAW Watching TV commercial is complete waste of time Most TV commercials are pretty bad TV commercials are monotonous TV commercial do not interfere with regular TV watching I have no feelings whatsoever I like TV commercials at times Most TV commercial are fairly interactive 54 9) I like TV to buy products advertised an TV 10) Most TV commercials help people select best products 11) TV commercials are fun to watch than regular programmes > Respondents are required to agree or disagree with the statement. > In a general trend if respondents agree with statements of scale values 8,9 & 10, they may disagree with 1, 2, 3, & 4. Neutral positions will be 5, 6 & 7. >Median of the item nos is calculated of 8, 9 & 10 favorably disposed toward TV commercials >1, 4, 7 & 11 agreement would indicate no organized attitude toward TV commercial 55 > Not applied in MR frequently owing to tedious procedure. Disadvantage 1) Cumbersome to develop 2) Many bias can arise shortlisting & making average statements 3) Degree indicated. Paired comparison scale > This requires respondents to pick two objects of a set with regard to an attitude. > A series of paired judgments between objects preference. > Suppose the respondent is asked to show his preference amongst 56 in of agreement or disagreement is not is made by the respondent on the basis of his Five brands of tea A, B, C, D, E with respect to its flavor. > He has to select one brand out of all paired combination of the five brands A&B A&C A&D A&E B&C B&D B&E C&D C&E so, 10 D&E paired comparisons n(n-1)÷ 2 = 5 (5-1) ÷ 2 = 10 > In each pair, the respondents are asked to divide 100 points on the basis of how much they liked one as compared to the other eg. between A & B liking, 60 points to A 40 points to B; between A & C liking 70 points to A 30 points to C etc >When combinations are large paired comparison is not preferred. When 57 There are 15 brands to be evaluated, we will have 105 paired Comparisons which is the major limitation of this method 58