Please recommend a test market design and address the associated questions for following scenarios. In your answer, be sure to label the Independent Variables (IV), the Dependent Variables (DV), and any control variables. Include the timeline and associated costs. Also, in answering the questions, you might consider using X and O to help describe your plan. See experimental designs starting on page 189. Be as specific as possible.

Please organize your response to each scenario using the following headings: 1. Objective of Test-Market: 2. Variables: Identify the main IV(s) and DV(s) 3. Control Variables: Identify these and explain why you need to control for them. 4. Test procedure details and justification: (briefly describe the plans to carry out the test. What will consumers do? What will the company do? How many test markets? How long should it last for? What are the associated costs? What information do you need to obtain to help answer the question of interest?) Be sure to justify your decisions. 5. Overall Recommendation: Given the associated direct and indirect costs of conducting your proposed test market, and keeping in mind the four main factors listed in the text to consider in determining whether or not to conduct the test market, (p. 196) do you recommend conducting the test market you outlined above? Please explain your answer.

Scenario: A soft drink company has determined in taste tests that consumers prefer the taste of their diet product when sweetened with Splenda in comparison to Equal. Now they are interested in determining how the new sweetener will play in the marketplace. Design a test market that will achieve this goal.

196 CHAPTER 9 The Concept of Measurement Exhibit 9.14 Stapel Scale Used to Measure a Retaller's Website Cheer En sy to Navigate Select a for you ik debe the we l y The more countly you think the word so the we b er you should choose Select a minus number for words you think niet de the w h oly Th u ney you think the word describes the web to the targete s you should co m e you can select any number from 5 for words you r very own the way to for words you think every inaccurate by contrast, reflects how dawl the descriptor adjective is the concepe being evaluated. An example of a Stape cale is shown in Exhibit 9.14 The primary advantage of the Supe sale is that it enable the researcher to avoid the ardu- ous task of creating bipolar adjective pairs. The scale may also permit finer discrimination in measuring attitude. A drawback is that descriptor adjectives can be phrased in a positive neutral or negative vein, and the choice of phrasing has been shown to affect the scale results and the person's ability to respond The Sapel sale has never had much popularity in commercial research and is used less frequently than the semantic differential Measurement scale in which the respondent species a level of agreement or disagreement with statements expressing ether a favorable or an unfavorable atitude toward the concept under study Likert scales The Likert scale is another scake that avoids the problem of developing pales of dichotomous adjectives. The scale consists of a series of statements expressing either a favorable or an unfavora- ble attitude toward the concept under study. The respondent is asked to indicate the level of her or his agreement or disagreement with each statement by signing it a numerical score. The scores are then totaled to measure the respondents atinde Exhibit 9.15 shows two Likert scales for an inte mer game site argeted toward teenagers Scale Ameasures tudes toward the registration process scale Bevaluates ses aminides towand advertising on the website With the likert scale the respondent is required to consider only one statement at a time, with the scale running from one extreme to the other series of statements (artitudes) can be examined, vel there is only a single set of uniform replies for the respondent to choose from. Ranns Likert arated this wile to measure a person's attitude toward concepes leg. unions). activities (eg, swimming, and so forth. He recommended the following steps in building the scale: 1. The teacher identifies the concept or activity to be caled. 2. The searcher assembles a large number of statements (5-100) concerning the publie's sentiments toward the concept or activity 3. Each sestem is assified by the recicher as generally "favorable or unfavorable" with ed to the attitude underly. No me is made to scale the items however, a petes is conducted that involves the full set of statements and a limited sample of respondents 4. In the press the spondent indicates agreement (or not) with writem, checking one of the following direction s descriptors 194 CHAPTER 9 The Concept of Measurement Here are some characteristics used to describen care products in gene Please tell me which characteristic in each persone mortant to you when wetgan care product Exhibit 9.11 Paired Comparison Scale for Sun Care Products h an a. Prevents burning a Good value for the money ins without burning b. Protects against burning and taining Does not stanothing Prevents burning a. Tans without buming Protects at buming and tag hoty Exhibit 9.12 Constant Sum Scale Used In Tennis Sportswear Study Below researcherstics of women and corr e cta 100 bets among the characters such that the doctor the port of each chc e to you. The more points you to a chadw the more important the chance to unimportant, you should not alocate p e tot When your vensted, please double check to make sure that your adds to 100 Characteristics of Tervis Sportswear Number of Points comoto durasi made by well known brand or sportsmantan made in the United States Has up-to-date styre Givestreadom of movement is a good for the money 100 points study of tennis sportswear is shown in Exhibit 9.12. Another advantage of the constant sum scale over a rank order or paired comparison sale is that if the respondent p owestwo characteristics to have equal value he or she can sodat A major disadvantage of this scale is that the respondent may have difficulty allocating the points to total 100 if there are a lot of characteristics or items. Most wachers for 10 items is the upper limit on a constant sum scale Measurement scale that m es the strengths concept by having the respondentrant betwen dichotomous Semantic Differential Scales The semantic differential was developed by Charles Osgood. George Saci and Percy Tannenbaum The focus of the original arch was the m o ment of meaning of an ob to a person The object mujhe beauvings and an association and the meaning in image among certain roup of people The construction of antic differential scale begins with determination of a concept bertad, such as the map of a company, brand, or so. The researcher selects dichotomous oppositel pairs of words or phrases that could be to describe the concepe. Respondents then rate the concepe una selecally to the of the wsponses for sach pui oladices is computed the p asa podelima and profile of an Arion savings and an a ction preved by c customers wie omers with family name SHO,000 and abo Aquick dance shows that the tim v iewed olishioned with her plain facilitie s web bl a nd baby w odel the parking and perhaps tyd gedical advertig is viewed as dismal prathatod be used to content the means of the re promage Attitude Measurement Scales 193 Eye Shadow Scales (A) Pre the following eye wadows, with being the brand that best meets the characteristic bong w ted and the worst and on the cha ristic being used. The six brands are listed on card HAND RESPONDENT CARD C) Let's begin with the idea of having high-quality compacts or containers Which brand would rank hang the highest quality compacts or corners which is second RECORD BELOW Q4. Having 0.49. Having Os Having High O ity Contin Applicator Eye Shadow Avon Exhibit 9.10 Series of Rank Order Scales Used to Evaluate Eye Shadows and Car Resale Values L'Oreal Revo Card C Cover GH Sephora Avon L'Oreal Este Car Resale Value Sale Based on your personal experience or what you have seen heard or read please make the following car brands aconding to the rest value percentage that is the brand that enables you to recover the largest wount percentage of your original purchase price of the vehicle Picea next to the brand that has the highest resav e percentage a 2 next to the brand that has the next higheste vale percentage and so forth Itamamber, no two cars can have the same ranking Chevrolet Toyota BMW Ford Paired Comparisons Paired comparison scales as a respondent to pick one of two objects from a wet based on some Sated crimi The respondent, therefore makes a series of praised judgments between objects Exhibit9.11 shows a paired comparison sale used in a national study for un care products. Only part of the scale is shown the data collection procedure typically requires the respondent to com pure all possible pain of objects. Paired comparison on several problems of traditional rank onder scales. First, it is easier for people to select one item from a set of two than to ranka large set of data. Second, the problem of onderbuis one there is no pattern in the ordering of items or questions a r e of bus. On the beaux all pouble pisane valuated, the number of paired compari s o me cytherofobo be evaluated in Carithmetic. Thus the number of objects to be evaluated should remain fauty small to preventiv e fatigue dows Measurement scales that the respondent 10 por one of two objects in a set based on some stated area Constant Sum Scales To avoid long loss of paired items, marketing teachers use constant sum scales mom often than pand compar a nt w cale the respondent se divide a given number of pois typically 100 m g two or more but based on their impone to him or her Respondents must value each item lative to all other woms. The number of points allocated chalematices the ranking and by the respondent as well as the cavemas de of achamative proved by the respondent. A constant sum scale used in a nacional Measurement scales that the respondent to divide a gun number of points, vocally 100 among two or more atributos based on The importance to hun or her 192 CHAPTER 9 The Concept of Measurement EXHIBIT 9.9 Selected Ite mized Rating Scales Characteristic of interest Rating Choices Purchase intent Definitely will buy Probably will buy Probably will not by Definitely will not buy Level of Agreement Strongly agree Somewhat agree Neither or no Somewhat d oe Strongly agree Quality Dependability Good Somewhat dependable Nother good or bad Not very dependable Far Not dependable at Very good Completely decedat Very stylish Style Somewhat stylish Not very stylish Complete unistys Satisfaction Completely satisfied Somewhat satisfied Nother side Completely disshished Cost Extremely expensive expensive Sahityinexpensive Nerther expensive expensive Very expensive Ease of Use Color Brightness Very easy to use Extremely bright Somewhat easy to use Very light Not very easy to use Somewhat bright i t to use Slightly bright Not bright at Modernity Very modern Somewhat modem He thermocon Somat Two-Stage Format "How diective do you believe Senator Foghom is in having your money stay in the community How tect Would be veryosomwa Efective ineffective No opinion Very Somewhat Mesutement scales in which judgment is made without reference to another object concept, or person M oment calen ondert Rank-Order Scales Itemined and graphic sales are considered to be noncomparative scales because the respondent makes a judgment without reference to another object, concept, or person Rank order scales, by contrast, are comparative scales because the respondent is asked to compare two or more items and rank each item. Rank-order scales are widely used in marketing research for several reasons They are easy to use and give ordinal measurements of the items evaluated. Instructions are to understand, and the process typically moves at steady pace. Some rescascher daim thank der scales force respondents to evaluate concepts in a real manner. Exhibit 9.100AH trates a series of rank-onder scales taken from a study on eye shadows Exhibit 9.10(1) shows an online scale on automobile real value percentage Rank-onder les postess several disadvantages. If all of the alternatives in a respondent's choice we are not included, the results could be misleading. For example, respondentist choice on all dimensions in the eye shadow seady might have been Mineral Fusion, which was not induded. A con problem is that the concepe being anked may be completely outside choice , thus producing meaningless dura Perhaps a respondent doesnt use eye shadow and fed that the producer appropriate for my woman. Another limitations that the sake gives che searcher only ordinal dara. Nothing is learned about how far apart the kems stand or how intensely the respondientes about the ranking of a m . Finally, there archer does not know why the respondent anked the items as he or she did. compares two or more items and ranks them Measurement scales in which one object concept. or person is compared with the on a scale 190 CHAPTER 9 | The Concept of Measurement Scale A Uncomfortable Comfortable Scale B 10 20 8000 0 Uncomfortable 100 Comfortable Scalc Very uncomfortable very comfortable Exhibit 97 Four Types of Graphic Rating Scales Not at all important Fromely important Exhibit 9.8 offered, how would you be to use the following areas on this site? Scale A itemized Rating Scales Used in Internet and Mall a. Auctions Extremely 01 02 03 04 05 b. Foe-based ducation Not at all likely to 5 c. Event registration Not at all likely to use d Online shopping markets Not at all key to use e Recruiting Notatky so 5 Extremely likely to use Extremely likely to use Extremely likely 5 BB 8888 8 8 8 8 8 82 588 383 3838 5 8888888 Extremely likely ou 5 Extremely likely Notatayo Trang com Not at you h Trainings Not at all likely to 5 Extremely likely Measurement Scales 189 phic ratings Measurement scales that inchide a graphic continuum, anchored by twares Graphic Rating Scales Graphic rating sales terrespondents graph e Exhibit 9.7 deplats three types of graphic main cale h Boy racinen. Sak wpresents the simple fom e sale. Rea d to mark their response on the continuum. Atendens led a ined by dividing the line inte as many c r iss desired and assigning a based on the category in which the mark has been placed. For example, if the line were inches inch might represent a category. Soal Botfen the respondent Wightly more ser i ng numbers along the sale Responses to graphic rating sales are not limited to simply placing a mark on a continua as wale ch e , Scle Chas been successfully by many researchers speed up inter views. Respondents are asked to touch the thermometer on the screen or mobile device the best depicts their feelings. Graphik rating sales are also sometimes called "Widers" when used in online and mobile surveys. The respondent simply uses his or her finger to slide the scale along a continuum. On some slides, a number will pop up above the sider as it is moned along the scale (nee Scale in Exhibit 9.7). Graphic rating clean be connected easily and we simple to use. They enable escarchet to discern fine distinctions. ming that there has adequate discriminatory abilities Numerical data obtained from the scales are typically treated as interval dan One disadvantage of graphic rating sales is that every drome anchos tend to force respondents toward the middle of the sake. Als, some research has suggested that such scales not as reliable sitemindesting scales. madrating sa Measurement scales in which the respondent selects an answer from aimed rumber of ordered categories Itemized Rating Scales Itemised rating scales are similar to graphic rating scales, except that respondents must select from a limited number of ordered categories rather than placing a mark on a continuous scale. Purists would argue that scale in Exhibit 9.7 is an itemid rating scale) Exhibit 9.8 shows come cuamples of itemized rating sales taken from nationwide marketing research surveys Scaring items a otted on cach questionnaire to diminate the order is that might arise from arting with the same item each time. Scale A was used by a dor.com company in determining what features and services it should add te is website. Scale B was used in measuring satisfaction with an online travel site. Scale C was und by an e-commerce music retailer to better understand how people select a music website. Scale D was also an Internet survey conducted by a producer of customer relationship manage ment software Examples of other itemized rating scales are shown in Exhibit 9.9. Although itemized rating sales do not allow for the fine distinctions that can be achieved in a praphk mating sak, they are easy to construct and administer. And the definitive categories found in emited rating sales sally produce more reliable ratings When reachers for some are interested in the most extreme position views, they might dect to use a two-stage format R anch has shown that a fomut can provide bener data quality in detecting extreme views than a single-stage itemwed rating scale. The follow ing is an example of the two-stage approach Traditional One-Stage Format "How effective do you believe Senator loghorn is in having your money stay in the community?" Very effective Somewhe effective Somewhat ellective 196 CHAPTER 9 The Concept of Measurement Exhibit 9.14 Stapel Scale Used to Measure a Retaller's Website Cheer En sy to Navigate Select a for you ik debe the we l y The more countly you think the word so the we b er you should choose Select a minus number for words you think niet de the w h oly Th u ney you think the word describes the web to the targete s you should co m e you can select any number from 5 for words you r very own the way to for words you think every inaccurate by contrast, reflects how dawl the descriptor adjective is the concepe being evaluated. An example of a Stape cale is shown in Exhibit 9.14 The primary advantage of the Supe sale is that it enable the researcher to avoid the ardu- ous task of creating bipolar adjective pairs. The scale may also permit finer discrimination in measuring attitude. A drawback is that descriptor adjectives can be phrased in a positive neutral or negative vein, and the choice of phrasing has been shown to affect the scale results and the person's ability to respond The Sapel sale has never had much popularity in commercial research and is used less frequently than the semantic differential Measurement scale in which the respondent species a level of agreement or disagreement with statements expressing ether a favorable or an unfavorable atitude toward the concept under study Likert scales The Likert scale is another scake that avoids the problem of developing pales of dichotomous adjectives. The scale consists of a series of statements expressing either a favorable or an unfavora- ble attitude toward the concept under study. The respondent is asked to indicate the level of her or his agreement or disagreement with each statement by signing it a numerical score. The scores are then totaled to measure the respondents atinde Exhibit 9.15 shows two Likert scales for an inte mer game site argeted toward teenagers Scale Ameasures tudes toward the registration process scale Bevaluates ses aminides towand advertising on the website With the likert scale the respondent is required to consider only one statement at a time, with the scale running from one extreme to the other series of statements (artitudes) can be examined, vel there is only a single set of uniform replies for the respondent to choose from. Ranns Likert arated this wile to measure a person's attitude toward concepes leg. unions). activities (eg, swimming, and so forth. He recommended the following steps in building the scale: 1. The teacher identifies the concept or activity to be caled. 2. The searcher assembles a large number of statements (5-100) concerning the publie's sentiments toward the concept or activity 3. Each sestem is assified by the recicher as generally "favorable or unfavorable" with ed to the attitude underly. No me is made to scale the items however, a petes is conducted that involves the full set of statements and a limited sample of respondents 4. In the press the spondent indicates agreement (or not) with writem, checking one of the following direction s descriptors 194 CHAPTER 9 The Concept of Measurement Here are some characteristics used to describen care products in gene Please tell me which characteristic in each persone mortant to you when wetgan care product Exhibit 9.11 Paired Comparison Scale for Sun Care Products h an a. Prevents burning a Good value for the money ins without burning b. Protects against burning and taining Does not stanothing Prevents burning a. Tans without buming Protects at buming and tag hoty Exhibit 9.12 Constant Sum Scale Used In Tennis Sportswear Study Below researcherstics of women and corr e cta 100 bets among the characters such that the doctor the port of each chc e to you. The more points you to a chadw the more important the chance to unimportant, you should not alocate p e tot When your vensted, please double check to make sure that your adds to 100 Characteristics of Tervis Sportswear Number of Points comoto durasi made by well known brand or sportsmantan made in the United States Has up-to-date styre Givestreadom of movement is a good for the money 100 points study of tennis sportswear is shown in Exhibit 9.12. Another advantage of the constant sum scale over a rank order or paired comparison sale is that if the respondent p owestwo characteristics to have equal value he or she can sodat A major disadvantage of this scale is that the respondent may have difficulty allocating the points to total 100 if there are a lot of characteristics or items. Most wachers for 10 items is the upper limit on a constant sum scale Measurement scale that m es the strengths concept by having the respondentrant betwen dichotomous Semantic Differential Scales The semantic differential was developed by Charles Osgood. George Saci and Percy Tannenbaum The focus of the original arch was the m o ment of meaning of an ob to a person The object mujhe beauvings and an association and the meaning in image among certain roup of people The construction of antic differential scale begins with determination of a concept bertad, such as the map of a company, brand, or so. The researcher selects dichotomous oppositel pairs of words or phrases that could be to describe the concepe. Respondents then rate the concepe una selecally to the of the wsponses for sach pui oladices is computed the p asa podelima and profile of an Arion savings and an a ction preved by c customers wie omers with family name SHO,000 and abo Aquick dance shows that the tim v iewed olishioned with her plain facilitie s web bl a nd baby w odel the parking and perhaps tyd gedical advertig is viewed as dismal prathatod be used to content the means of the re promage Attitude Measurement Scales 193 Eye Shadow Scales (A) Pre the following eye wadows, with being the brand that best meets the characteristic bong w ted and the worst and on the cha ristic being used. The six brands are listed on card HAND RESPONDENT CARD C) Let's begin with the idea of having high-quality compacts or containers Which brand would rank hang the highest quality compacts or corners which is second RECORD BELOW Q4. Having 0.49. Having Os Having High O ity Contin Applicator Eye Shadow Avon Exhibit 9.10 Series of Rank Order Scales Used to Evaluate Eye Shadows and Car Resale Values L'Oreal Revo Card C Cover GH Sephora Avon L'Oreal Este Car Resale Value Sale Based on your personal experience or what you have seen heard or read please make the following car brands aconding to the rest value percentage that is the brand that enables you to recover the largest wount percentage of your original purchase price of the vehicle Picea next to the brand that has the highest resav e percentage a 2 next to the brand that has the next higheste vale percentage and so forth Itamamber, no two cars can have the same ranking Chevrolet Toyota BMW Ford Paired Comparisons Paired comparison scales as a respondent to pick one of two objects from a wet based on some Sated crimi The respondent, therefore makes a series of praised judgments between objects Exhibit9.11 shows a paired comparison sale used in a national study for un care products. Only part of the scale is shown the data collection procedure typically requires the respondent to com pure all possible pain of objects. Paired comparison on several problems of traditional rank onder scales. First, it is easier for people to select one item from a set of two than to ranka large set of data. Second, the problem of onderbuis one there is no pattern in the ordering of items or questions a r e of bus. On the beaux all pouble pisane valuated, the number of paired compari s o me cytherofobo be evaluated in Carithmetic. Thus the number of objects to be evaluated should remain fauty small to preventiv e fatigue dows Measurement scales that the respondent 10 por one of two objects in a set based on some stated area Constant Sum Scales To avoid long loss of paired items, marketing teachers use constant sum scales mom often than pand compar a nt w cale the respondent se divide a given number of pois typically 100 m g two or more but based on their impone to him or her Respondents must value each item lative to all other woms. The number of points allocated chalematices the ranking and by the respondent as well as the cavemas de of achamative proved by the respondent. A constant sum scale used in a nacional Measurement scales that the respondent to divide a gun number of points, vocally 100 among two or more atributos based on The importance to hun or her 192 CHAPTER 9 The Concept of Measurement EXHIBIT 9.9 Selected Ite mized Rating Scales Characteristic of interest Rating Choices Purchase intent Definitely will buy Probably will buy Probably will not by Definitely will not buy Level of Agreement Strongly agree Somewhat agree Neither or no Somewhat d oe Strongly agree Quality Dependability Good Somewhat dependable Nother good or bad Not very dependable Far Not dependable at Very good Completely decedat Very stylish Style Somewhat stylish Not very stylish Complete unistys Satisfaction Completely satisfied Somewhat satisfied Nother side Completely disshished Cost Extremely expensive expensive Sahityinexpensive Nerther expensive expensive Very expensive Ease of Use Color Brightness Very easy to use Extremely bright Somewhat easy to use Very light Not very easy to use Somewhat bright i t to use Slightly bright Not bright at Modernity Very modern Somewhat modem He thermocon Somat Two-Stage Format "How diective do you believe Senator Foghom is in having your money stay in the community How tect Would be veryosomwa Efective ineffective No opinion Very Somewhat Mesutement scales in which judgment is made without reference to another object concept, or person M oment calen ondert Rank-Order Scales Itemined and graphic sales are considered to be noncomparative scales because the respondent makes a judgment without reference to another object, concept, or person Rank order scales, by contrast, are comparative scales because the respondent is asked to compare two or more items and rank each item. Rank-order scales are widely used in marketing research for several reasons They are easy to use and give ordinal measurements of the items evaluated. Instructions are to understand, and the process typically moves at steady pace. Some rescascher daim thank der scales force respondents to evaluate concepts in a real manner. Exhibit 9.100AH trates a series of rank-onder scales taken from a study on eye shadows Exhibit 9.10(1) shows an online scale on automobile real value percentage Rank-onder les postess several disadvantages. If all of the alternatives in a respondent's choice we are not included, the results could be misleading. For example, respondentist choice on all dimensions in the eye shadow seady might have been Mineral Fusion, which was not induded. A con problem is that the concepe being anked may be completely outside choice , thus producing meaningless dura Perhaps a respondent doesnt use eye shadow and fed that the producer appropriate for my woman. Another limitations that the sake gives che searcher only ordinal dara. Nothing is learned about how far apart the kems stand or how intensely the respondientes about the ranking of a m . Finally, there archer does not know why the respondent anked the items as he or she did. compares two or more items and ranks them Measurement scales in which one object concept. or person is compared with the on a scale 190 CHAPTER 9 | The Concept of Measurement Scale A Uncomfortable Comfortable Scale B 10 20 8000 0 Uncomfortable 100 Comfortable Scalc Very uncomfortable very comfortable Exhibit 97 Four Types of Graphic Rating Scales Not at all important Fromely important Exhibit 9.8 offered, how would you be to use the following areas on this site? Scale A itemized Rating Scales Used in Internet and Mall a. Auctions Extremely 01 02 03 04 05 b. Foe-based ducation Not at all likely to 5 c. Event registration Not at all likely to use d Online shopping markets Not at all key to use e Recruiting Notatky so 5 Extremely likely to use Extremely likely to use Extremely likely 5 BB 8888 8 8 8 8 8 82 588 383 3838 5 8888888 Extremely likely ou 5 Extremely likely Notatayo Trang com Not at you h Trainings Not at all likely to 5 Extremely likely Measurement Scales 189 phic ratings Measurement scales that inchide a graphic continuum, anchored by twares Graphic Rating Scales Graphic rating sales terrespondents graph e Exhibit 9.7 deplats three types of graphic main cale h Boy racinen. Sak wpresents the simple fom e sale. Rea d to mark their response on the continuum. Atendens led a ined by dividing the line inte as many c r iss desired and assigning a based on the category in which the mark has been placed. For example, if the line were inches inch might represent a category. Soal Botfen the respondent Wightly more ser i ng numbers along the sale Responses to graphic rating sales are not limited to simply placing a mark on a continua as wale ch e , Scle Chas been successfully by many researchers speed up inter views. Respondents are asked to touch the thermometer on the screen or mobile device the best depicts their feelings. Graphik rating sales are also sometimes called "Widers" when used in online and mobile surveys. The respondent simply uses his or her finger to slide the scale along a continuum. On some slides, a number will pop up above the sider as it is moned along the scale (nee Scale in Exhibit 9.7). Graphic rating clean be connected easily and we simple to use. They enable escarchet to discern fine distinctions. ming that there has adequate discriminatory abilities Numerical data obtained from the scales are typically treated as interval dan One disadvantage of graphic rating sales is that every drome anchos tend to force respondents toward the middle of the sake. Als, some research has suggested that such scales not as reliable sitemindesting scales. madrating sa Measurement scales in which the respondent selects an answer from aimed rumber of ordered categories Itemized Rating Scales Itemised rating scales are similar to graphic rating scales, except that respondents must select from a limited number of ordered categories rather than placing a mark on a continuous scale. Purists would argue that scale in Exhibit 9.7 is an itemid rating scale) Exhibit 9.8 shows come cuamples of itemized rating sales taken from nationwide marketing research surveys Scaring items a otted on cach questionnaire to diminate the order is that might arise from arting with the same item each time. Scale A was used by a dor.com company in determining what features and services it should add te is website. Scale B was used in measuring satisfaction with an online travel site. Scale C was und by an e-commerce music retailer to better understand how people select a music website. Scale D was also an Internet survey conducted by a producer of customer relationship manage ment software Examples of other itemized rating scales are shown in Exhibit 9.9. Although itemized rating sales do not allow for the fine distinctions that can be achieved in a praphk mating sak, they are easy to construct and administer. And the definitive categories found in emited rating sales sally produce more reliable ratings When reachers for some are interested in the most extreme position views, they might dect to use a two-stage format R anch has shown that a fomut can provide bener data quality in detecting extreme views than a single-stage itemwed rating scale. The follow ing is an example of the two-stage approach Traditional One-Stage Format "How effective do you believe Senator loghorn is in having your money stay in the community?" Very effective Somewhe effective Somewhat ellective