As shown in Figure 21.2, input data obtained from the respondents may be related to perceptions or preferences. Perception data, which may be direct or derived, is discussed first.
PERCEPTION DATA: DIRECT APPROACHES In direct approaches to gathering perception data, the respondents are asked to judge how similar or dissimilar the various brands or stimuli are, using their own criteria. Respondents are often required to rate alI possible pairs of brands or stimuli in terms of similarity on a Likert scale. These data are referred to as similarity judgments.
For example, similarity judgments on all the possible pairs of toothpaste brands may be obtained in the folIowing manner
The number of pairs to be evaluated is n(n – I )/2, where n is the number of stimuli. Other procedures are also available. Respondents could be asked to rank-order all the possible pairs from the most similar to the least similar. In another method, the respondent rank-orders the brands in terms of their similarity to an anchor brand. Each brand, in turn, serves as the anchor. In our example, the direct approach was adopted. Subjects were asked to provide similarity judgments for all 45, (10 X 9/2), pairs of toothpaste brands, using a 7-point scale. The data obtained from one respondent are given in Table 21.1.5
PERCEPTION DATA: DERIVED APPROACHES Derived approaches to collecting perception data are attribute-based approaches requiring the respondents to rate the brands or stimuli on the identified attributes using semantic differential or Likert scales. For example, the different brands of toothpaste may be rated on attributes such as these:
Sometimes an ideal brand is also included in the stimulus set. The respondents are asked to evaluate their hypothetical ideal brand on the same set of attributes. If attribute ratings are obtained, a similarity measure (such as euclidean distance) is derived for each pair of brands
DIRECT VERSUS DERIVED APPROACHES Direct approaches have the advantage that the researcher does not have to identify a set of salient attributes. Respondents make similarity judgments using their own criteria. as they would under normal circumstances. The disadvantages are that the criteria are influenced by the brands or stimuli being evaluated. If the various brands of automobiles being evaluated are in the same price range. then price will not emerge as an important factor. It may be .difficult to determine before analysis if and how the individual respondents’ judgments should be combined. Furthermore. it may be difficult to label the dimensions of the spatial map.
The direct approaches are more frequently used than the attribute-based approaches. However.it may be best to use both these approaches ill a complementary way. Direct similarity judgments may be used for obtaining the spatial map, and attribute ratings may be used as an aid to interpreting the dimensions of the perceptual map. Similar procedures are used for preference data
PREFERENCE DATA Preference data order the brands or stimuli in terms of respondents’ preference for some property. A common way in which such data are obtained is preference rankings. Respondents are required to rank the brands from the most preferred to the least preferred. Alternatively, respondents may be required to make paired comparisons and indicate which brand in a pair they prefer. Another method is to obtain preference ratings for the various brands. (The rank order, paired comparison. and rating scales were discussed in on scaling techniques.) When spatial maps are based on preference data. distance implies differences in preference. The configuration derived from preference data may differ greatly from that obtained from similarity data. Two brands may be perceived as different in a similarity map yet similar in a preference map. and vice versa. For example. Crest and Pepsodent may be perceived by a group of respondents as very different brands and thus appear far apart on a perception map. However, these two brands may be about equally preferred and appear close together on a preference map. We will continue using the perception data obtained in the toothpaste example to illustrate the MDS procedure and then consider the scaling of preference data.
Select an MDS Procedure
Selection of a specific MDS procedure depends upon whether perception or preference data are being scaled, or whether the analysis requires both kinds of data. The nature of the input data is also a determining factor. Nonmetric :IIDS procedures assume that the input data are ordinal, but they result in metric output. The distances in the resulting spatial map may be assumed to be interval scaled. These procedures find. in a given dimensionality, a spatial map whose rank orders of estimated distances between brands or stimuli best preserve or reproduce the input rank orders. In contrast. metric MDS methods assume that input data are metric. Because the output is also metric. a stronger relationship between the output and input data is maintained, and the metric (interval or ratio) qualities of the input data are preserved. The metric and nonmetric methods produce similar results.s
Another factor influencing the selection of a procedure is whether the MDS analysis will be conducted at the individual respondent level or at an aggregate level. In individual-level analysis, the data arc analyzed separately for each respondent, resulting in a spatial map for each respondent. Although individual-level analysis is usefui from a research !,’Crspective, it is not appealing from a managerial standpoint. Marketing strategies are typically formulated at the segment or aggregate level. rather than at the individual level. If aggregate-level analysis is conducted, some assumptions must be made in aggregating individual data. Typically, it is assumed that all respondents use the same dimensions to evaluate the brands or stimuli, but that different respondents weight these common dimensions deferentially
The data of Table 21.1 were treated as rank ordered and scaled using a non metric procedure. Because one respondent provided these data, an individual-level analysis was conducted. Spatial maps were obtained in one to four dimensions and then a decision on an appropriate number of dimensions was made. This decision is central to all MDS analyses therefore. it is explored in greater detail in the following section