英语论文网

rganisation at least two respondents were chosen from each category of executives, viz. Type A and Type B. Type A respondents include senior level executives like general managers, marketing heads, regional managers, and product managers. These executives can exert direct influence on the decision to introduce modifications in the product. Type B respondents include the senior and middle level executives who indirectly influence the decisions to introduce modifications in the product. Area sales managers, senior officers’ in-charge for rural markets, and rural marketing executives are some of the Type B executives.

RESULTS

Perceived Rurality in Rural Markets

Alpha value of 0.736 given in Table - 2 for the thirteen item scale used to measure perceived rurality of rural markets indicates high levels of internal consistency in the items. The value of 0.877 indicates a higher level of predictive validity of the scale. Mean values of the thirteen item scale measuring perceived rurality of rural markets ranged from 3.03 to 4.22, indicating that the perceived rurality is moderate to high. Mean value of overall social representation for rural markets is 2.49 indicating moderate level of response from the executives. Among the respondents drawn from four different product categories, executives from banking and insurance and hair care product categories perceived higher rurality when compared to television and pharmaceutical categories.

Cluster Analysis

Two stage clustering process as advocated by Punj and Steward (1983) and Hair et al (1998) is employed to generate a manageable number of groups of perceived rurality. Later the cluster solutions would be used to examine product adaptation in rural markets. As a first step, hierarchical clustering using Ward’s method for minimizing within cluster differences, Squared Euclidean distance as similarity measure were used. Agglomeration Schedule suggested that the percentage change in the coefficients is large when going from two to one clusters (32.26) followed by when going from three to two clusters (15.03). Idea at this stage was not to interpret the clusters but to ensure that the cluster solutions are truly distinctive. If the results were restricted to only two cluster solution, then two quite different categories of executives’ perception on rural markets would emerge (Table 3). Same is the case for three cluster solution. However, increased number of clusters exhibits a moderate improvement in representing distinct groups that may reflect an underlying structure (refer to F values in Table 3).

K-Means clustering (non hierarchical) was employed as second step with an objective to consider the practical significance of the clusters derived from hierarchical clustering method. Centroids of the items were employed as initial seed points from the results in Step 1. F values of two cluster solution indicates that one item (helping nature) is insignificant, two items (languages and competition) are significant at 0.05 level and all the other items are significant at 0.01 level. In case of three cluster solution; F values of one item i.e. competition is significant at 0.05 level and other items are significant at 0.01 level (Table 3).

As a validity check for the stability of the cluster solution, a second K means hierarchical solution was performed by allowing the procedure to ran本论文由英语论文网提供整理，提供论文代写，英语论文代写，代写论文，代写英语论文，代写留学生论文，代写英文论文，留学生论文代写相关核心关键词搜索。