• Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
No Downloads

Views

Total Views
25,333
On Slideshare
0
From Embeds
0
Number of Embeds
0

Actions

Shares
Downloads
695
Comments
3
Likes
1

Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
    No notes for slide

Transcript

  • 1. ATTITUDE MEASUREMENT AND SCALING TECHNIQUES1
  • 2. Contents  About Consumer Behaviour  Factors effecting Human Decision Process  Attitude Measurement  About Attitudes  Methods of Attitude Measurement  Non-structured - Non-Disguised Methods  Non Structured - Disguised Methods  Structured Non - Disguised Techniques[scaling]  Structured - Disguised Methods  Multidimensional Scaling[MDS]2
  • 3. We Study:  Consumer Behaviour / Motivation Research: Try to look in to consumers thought processes to find WHY s/he buys something . This is a highly qualitative and subjective area. Marketing Research views the human mind as a black box (Opaque) as inputs and outputs are seen but what happens in between is a mystery even when there is a reason for every action.3
  • 4. FACTORS EFFECTING HUMAN DECISION PROCESS • Economic Factors: Man as a rational economic decision-maker who computes the marginal utility of each rupee and spends on products that maximise marginal utility. However we dont know how the consumer determines the marginal utilities.Yet most marketing phenomena can be explained by effect of prices, supply and demand. • Personality Factors: Most products have a personality profile and appeal to persons having / wanting that profile (e.g. Red and White). MR studies personality dimensions to profile products / persons to which the product appeals. This is also known as psychographics.4
  • 5. FACTORS EFFECTING HUMAN DECISION PROCESS • External Factors: i) Non - Marketing ControlledVariables: Economic, Social, cultural forces beyond control of the individual firms. ii) Marketing ControlledVariables: Consumers decision affected by combination of prices, products, promotion, service, distribution. Factors in market place. Companies test product versions, packages, prices etc. to attain profits.5
  • 6. Why study Consumer Behaviour?? To find out:  Which things appeal to which people.  Why consumers react to same stimuli differently.  Image vis a vis competitors.  To model an appropriate mix for the target market6
  • 7. MOTIVATIONAL RESEARCH TECHNIQUES: ATTITUDE MEASUREMENT  "Attitude is defined as the predisposition to respond to an idea or object, and in marketing it relates to the consumers predisposition to respond to a particular product or service".  Attitudes are composed of 1) Beliefs about the subject 2)Emotional feeling (like-dislike) 3) Readiness to respond behaviourally - i.e. buy7
  • 8. More about Attitudes  Attitudes are complex and not fully understood.  Attitudes can be changed but they tend to persist.  Attitude measurement tends to focus on beliefs and emotional feelings.  None of existing devices is very accurate.  General methods of attitude measurement include :  Questionnaire methods (i.e. self-reports, verbal reactions to situations),  Observation methods (actions or physiological reaction like sweating, pupils dilation etc).8
  • 9. Specific Methods of Attitude Measurement:  Non - structured methods: a) Disguised b) Non – disguised  Structured methods: a) Disguised b) Non-disguised A method is structured when there is a formal structure or procedure for the questioning. It is disguised when the respondent does not know the purpose of the interview.9
  • 10. Word Association Interviews: Sentence Completion  Focus Group Pictorial Techniques  In – Depth Guessing errors reveal Scales: attitudes  Ordinal  Interval10
  • 11. 1) Non-structured - Non-Disguised Methods: Good for exploratory studies: Depth Interviews: Conducted by psychologists - in-depth discussion around the subject and each aspect introduced. Respondent is encouraged to open up and depends on interviewer ‘s skill and imagination. It is costly and also gives errors due to small sample Focus Group Interviews: Here groups are interviewed and it reduces costs and stimulates respondents by group interaction. However, domination by one or more individuals is possible.11
  • 12. 2) Non Structured - Disguised Methods: These help to get responses in some cases where respondents would not open up. Projective techniques are used to see conscious / unconscious attitudes. These are vague / incomplete stimuli and it is believed respondents respond to reveal their attitudes: a) Word associations: First reaction to words related to the subject. Responses are timed so as to take in to account answers which are reasoned out. Its like a game but requires skill to interpret results. b) Sentence Completion: Respondent asked to complete incomplete sentences with first thought and timed. Here some respondents may be able to see thru the disguises: "A person who reads India today is……."12
  • 13. 2) Non Structured - Disguised Methods: c) Story completion: give ending to a story d) Pictorial Techniques: Here pictures are used as stimuli : • Thematic Apperception Test: TAT shows series of ambiguous pictures and respondents are asked to tell a story. It is held that in describing the characters / actions the respondent indirectly tells about himself. • Cartoon Tests: Modification of TAT & simpler to show, Cartoons pertinent to the problem are shown one / more balloons is left open for the respondent to fill. Cartoons should be carefully chosen to reflect situations in which respondent can project themselves easily.13
  • 14. 3)Structured Non - Disguised Techniques : [Scaling] Since disguise makes measurement indirect and comparisons become difficult hence development of standardised attitude measuring instruments is necessary. Scaling is used to measure attitudes objectively. Basically there are two types of scales: • Ordinal Scales: Rank respondent according to some characteristics or ranks items / brands in order of preference. These do not measure degree of like / dislike rankings or distance between rankings. • Interval Scales: Give items / individuals rank order plus measure distance between rank positions in equal units.14
  • 15. Types of Scales 1. Self Rating Scales:  Graphic rating scales  Semantic differential scales 2. Ranking 3. Multiple item scale  Thurstone scale  Likert scale15
  • 16. Structured Non - Disguised Techniques a) Self-Rating Scales: Respondents classify themselves in one or two categories (or do not know) categories. Further refined to give additional alternatives in degrees of like or dislike. Two variations of self-rating scale are: i) Graphic Rating Scales: Widely used, Respondents asked to rate himself by checking a point on a scale between two extremes. There are different types of such scales in use. They can vary from a 3 point scale or 11 pt. It can be odd/even, measured numerically / verbally or by both. Such ordinal scales rank points and say nothing about relative distance between points.16
  • 17. ii) Semantic Differential Scales: Respondents are asked to rank brands (or other subject) according to attitude being studied. 17
  • 18. ii.) Semantic Differential Scales:  Popular and main use in brand and company image studies as it permits development of descriptive profiles that facilitate comparisons. Bipolar scales are used to rate any product, moving to / from bipolar adjectives. Researchers can develop scales as per antonyms or opposite.  It gives a good basis of comparisons of image profiles. It is simple to use, easy and fast to administer and sensitive to small differences to attitude, highly versatile, reliable and generally valid.18
  • 19. b) Ranking  Respondents can be asked to rank brands or other subjects of interest according to attitude being studied. Such rankings do not give absolute ratings. For example : For each of the products listed in the next slide four different brands and generic categories are listed. For each product rank the brands 1,2,3 or 4 according to how likely you would be to purchase the brand when you next buy that product. Put #1 by the brand you would be most likely to buy; #4 by the brand you would be least likely to buy.19
  • 20. Paper Towels Peanut Butter Coffee Canned Peas- Scot towels - Heritage House - Generic - DelMonte- Bounty - Skippy - Folgers - Jewel- Jewel - Generic - Eight O’clock - Generic- Generic - Jif - Maxwell House - Green Giant20
  • 21. c) Multiple Item Scales:  To facilitate summarising all elements in to one measure. Two most popular multiple item scales are: i) Thrustone Scale: Method of equal appearing intervals based on the concept that people can tell difference between the attitude represented by 2 different statements and can identify items approximately half way in between. Here we (a) Collect large number of statements related to the attitude in question. (b) Over 20 judges put them in to 11 piles negative in pile one; neutral in 6 and positive in 11 piles. (c) Drop statements that different judges have given scattered ratings.21
  • 22. Thrustone Scale (d) Select one statement from each pile for final scale. Following statements show how to measure attitudeto Tv commercials on a 11pt scale : 1. All TV commercials should be prohibited by law. 2. Watching TV commercials is a complete waste of time. 3. Most TV commercials are pretty bad. 4. TV commercials are monotonous. 5. TV commercials do not interfere too much with enjoying TV. 6. I have no feeling one way or the other about mostTV commercials. 7. I like TV commercials at times. 8. Most TV commercials are fairly interesting. 9. I like to buy products advertised on TV whenever possible. 10. Most TV commercials help people select the best products available. 11.TV commercials are more fun to watch than the regular programs.22
  • 23. Thrustone Scale Respondents indicate agree/disagree with each statement and generally they will agree with only one / few statements (take median) indicating direction of attitude. Disadvantage: 1) Not very popular due to time consuming task of preparing. 2) Judges attitude may influence their ratings. 3) Resp. can get same score even while agreeing with different with different statement items (as median is taken). 4) It does not get information on degree of intensity of agreement with different items (an medians is taken).23
  • 24. ii) Likert Scale Here Respondents are asked to indicate the degree of agreement or disagreement with each statement and each degree of agreement disagreement is given a numerical score and total score is computed by summing these scores from all statements.  Here we : (a) Get large number of statements relevant to the attitude in question which are clearly identifiable as favourable or unfavourable. (b) Select a series of responses that represent various degrees of agreement / disagreement. (five, 7 or 9 variation). (c) Administer statements to a rep. group. (d) Compute each individual score by summing scores of the response to each question. (e) Drop those statements that do not discriminate between the high and low scorers on the total test24
  • 25. Disadvantages of Likert Scale 1) Likert scales are of ordinal type and dont measure difference between attitudes. 2) Same problems as preparing Thurstone scales. 3) Similar scores can be achieved through varying combination of responses.25
  • 26. 4) Structured - Disguised Methods:  Since non - structured techniques are (i) slow, costly (ii) data collection / interpretation is subjective / open to bias, structured methods overcome the problem.  It is held that people tend to know more about things they favour / like and if asked to guess factual information they will guess in a direction favourable to their ideas. Respondent given questionnaire that they are not likely to answer correctly and so forced to guess.  The extent and direction of these guessing errors is assumed to reveal their attitudes as the subject.26
  • 27. Multidimensional Scaling [MDS]:  In earlier scales researchers knew in advance what attitude dimensions are relevant. In MDS computer based techniques are used to present an object in multidimensional space based on one or more respondents perceptions towards the object.  Here neither the respondent nor researcher can accurately identify the number or nature of dimensions a respondent utilises in the evaluation. MDS uses simple data - similarly or preference for brands, Companies etc and attempts to disclose dimensions that underlie these judgements.27
  • 28. Multidimensional Scaling [MDS]: MDS helps to answer following questions: 1. What are major attributes of a product class (e.g. soft drinks) which consumers perceive viewing the product and by which they compare different brands. 2. Which brands compete most / least with each other. 3. Would consumers accept a new brand with a combination of characteristic not found in the market. 4. What is consumers ideal (point) combination of attributes 5. What sales / advertising messages are compatible with brand perceptions.28
  • 29. Multidimensional Scaling [MDS]:29
  • 30. Multidimensional Scaling [MDS]: In MDS since a no. of attributes are being analysed simultaneously it is difficult to diagram the relationships but we can understand the concept from a 2-D comparison of 11 cars based on two attributes. Note: MDS techniques do not provide names of dimensions. The Researcher has to identify them using experience, nature of data, additional information etc. Data from many consumers in sample are collected and efforts are made to find groupings that tend to give similar configurations suggesting market segments that deserve attention.30
  • 31. Multidimensional Scaling [MDS]: Other Uses: (1) Market segmentation. (2) Perception at different stages of product life cycle. (3) Advertisement media selection. (4) Supplier evaluation of purchase managers.31
  • 32. Multidimensional Scaling [MDS]: Limitations: (1) Definition of similarity and preference imperfect - conceptual problems. (2) Empirical problems in subjective identification of relevant dimensions or bias in data collection (3) Computational problems - Most Computational programs assume linear distance because what is the best distance function is not known.32
  • 33. 33