response rate. Whilst lower than the expected 28% for executive level research this was considered sufficient for subsequent scale development. The completed sample was randomly split into calibration (n=211) and validation (n=210) samples. The calibration sample was used to develop the scale whilst the validation sample was used to verify the scale dimensionality and establish its psychometric properties.
. Scale development
operationalize B2B service brand identity, scale development procedures were guided by the extant scaling literature. With service brand identity defined and a priori dimensionality postulated, it was necessary to generate a pool of items that scaled each dimension. A total of 119 items were generated from the literature that sampled the domains of the five postulated dimensions. Table 2 provides details of the relevant literature used to scale each dimension.
Table 2 Source of B2B service brand identity scale items
DimensionMeasurement focusItems culled / adapted fromMarketing cultureImportance placed on marketing culture.Webster (1995; 1993; 1990), Kotter and Heskett (1992) Corporate visual identity systemsImportance placed on Corporate visual identity.Simoes et al. ( 2005), Stuart (Stuart, 1999), Melewar & Saunders (1998, 2000), Van den Bosch et al (2006), Baker and Balmer (1997) Integrated marketing communicationsDegree to which marketing communications is integratedReid (2005) Duncan and Moriarty (1997), Ewing and de Bussy (2000), Low and Lamb (2000) Customer relationship managementThe quality of the customerbrand relationship managementFournier (1994), Roberts, Varki, and Brodie (2003), Iacobucci et al. (1995), Sirdeshmukh et al. (2002), Aaker, Fournier, and Brasel (2004), Boon and Holmes (1999), Sin et al. (2005) Brand personalityThe strength, favourability and uniqueness of the brand personality association.Keller (2003) with the scaling literature multiple items were developed for each dimension.initial pool of items was then subject to an expert panel review. The panel consisted of three leading brand academics and three senior consultants working at global brand agencies based in the UK3. Each panel member was asked to rate the extent they thought each item «represented» the domain of service brand identity. Panel members were also invited to suggest new items to enhance content validity. A series of pretests with MBA students working in the IT services sector then followed to assess participants « understanding of questions. This step was informed by face validity concerns. The survey was then piloted with fifty individuals randomly selected from the sampling frame to obtain response rate estimates and »test run 'the survey process. This expert panel, pretests and pilot resulted in the initial pool of 119 items being reduced to 50. This represented the service brand identity items in the final survey.
5. Analysis
, item-to-total and scale reliability estimates were obtained to assess the internal consistency of the scale and remove «garbage» items. Cronbach alphas for each subscale ranged from 0.83 to 0.89. These reliability estimates exceeded 0.7 and indicated potential item redundancy cf. attenuation paradox. Given the early stage of scale of development all items were retained. Items th...
