The corrected item-to-total correlations of the final scale ranged from 0.3792 to 0.7204, which is above the minimum recommended level of 0.35 for inclusion of the items in a scale.The final scale came to include 42 positively stated items. 4.3 Factor Analytic Results In order to provide a more parsimonious interpretation of the results, 42-item scale was then Factor analyzed using the Principal Component method with Varimax rotation. However, before applying factor analysis, the data was tested for its appropriateness.
International Journal of Business and Social Science Vol. 2 No. 18; October 2011223 In the present study, Kaiser-Meyer-Oklin (KMO) Measure of Sampling Adequacy (MSA) and Bartlett‟s test of Sphricity were applied to verify the adequacy or appropriateness of data for factor analysis. In this study, the value of KMO for overall matrix was found to be excellent (0.918) and Bartlett‟s test of Sphericity was highly significant (p< 0.001). The results thus indicated that the sample taken was appropriate to proceed with a factor analysis procedure. Besides the Bartlett‟s Test of Sphericity and the KMO Measure of Sampling Adequacy, Communality values of all variables were also observed. The extraction value of the Communalities of all the variables was sufficiently above 0.50 except variable 24; this variable was removed from the instrument as per the recommendation of Hair et al.(2010). Further, for defining the factors clearly, two criteria have been employed. First, it was decided to delete any variable having loading below ± 0.50. Second, it was decided that a factor must be defined by at least two variables. This criterion is consonant with the observations made by Rahtz et al.(1988).With this criterion in mind, a series of factor analysis was performed on the data. Following each analysis, items which did not meet the criteria were deleted from the analysis. After these preliminary steps, Factor Analysis with Principal Component Analysis as an extraction method has been performed on the remaining 41-item scale. Furthermore, it was observed that the variable 21 was cross loaded in F1 and F4; that variable too was eliminated (as per the recommendation of Hair et al.2010) from the instrument. Factor Analysis was rerun on the remaining 40-item scale. Ultimately, the final factor solution, which met the criteria, included 34-items defined by seven factors. Consequently, life insurance service quality in the present study composes seven factors namely, Proficiency; Media and presentations; Physical and ethical excellence; Service delivery process and purpose; Security and dynamic operations; Credibility; and Functionality. The initial instrument (as proposed by Sureshchandar et al.,2001) was adjusted to account for seven factors rather than five factors of service quality.