Research method

A two-stage research design, combining qualitative and quantitative phases, was considered appropriate for the

Tina Harrison and Kathryn Waite research questions being addressed. The qualitative phase was used to develop a list of factors that IFAs identified as being important to their initial decision to develop a company web site. The use of qualitative research captures the "social actors" points of view (Blaikie, 2000) and thus meets the objective of exploring the perceptions of the social system. Rogers (1995) notes that diffusion research has been criticised for its pro-innovation bias and that there exists a need for researchers to "see an innovation through the eyes of their respondents" (p. 111).

Data were collected through a total of 20 UK-wide individual interviews conducted from November 2003 to May 2004. Participants volunteered to take part in the research after a request was circulated through a professional network. Interviews were conducted in the major UK financial centres such as London, Bristol and Edinburgh together with other key areas such as the Home Counties and the North East. Participants were drawn from the different sizes of IFA companies ranging from those with turnovers of more than £5 million to sole traders and represented IFAs at different stages of web site development. Given the dynamic and, at times, complex relationship between internet development and social interaction, data collection followed the phenomenological tradition and respondents were asked to relate the story of web site development. (Thompson et al., 1989). The methodology facilitated the capture of the "processural view" of web site development and captured some degree of longitudinal data through participants' recollections (Bryman, 1988, p. 140).

The interviews were transcribed and then analysed using a method of content analysis informed by critical incidence technique (CIT). CIT is a classification technique employing the content of stories or "critical incidents" as data. Critical incidents are defined as "certain important facts concerning behaviour in defined situations" (Flanagan, 1954). It is recognised that these research methods come from what is perceived to be disparate philosophical traditions ofpositivism and interpretivism (see Denzin and Lincoln, 2000). The combined use of these techniques was considered appropriate given the financial industry's rising resentment towards

research and awareness of a fixed agenda as a consequence of successive government reviews (Sandler, 2002). Hence use was made of an open unstructured interview format to allow respondents maximum freedom of expression.

The 20 interviews lastest on average an hour each, ranging from 45 minutes to just over two hours. This resulted in interview transcripts of between 10 and 20 pages per interview. The "story of web site development" theme was one of several themes explored through the interviews and represented approximately one fifth of the interview time, or roughly around two pages of transcript per interview. Three of the 20 companies represented in the interviews had not developed a web site; hence, the analysis was based on the remaining 17 companies.

Events that were described as shaping web site development were clustered around themes. 54 statements were elicited from the interviews which described events that shaped web site adoption and development. The 54 statements clustered into 11 key themes (shown in Table I). Identification and classification of the critical incidents was performed by the authors, who had both carried out the interviews together. Each performed an independent assessment which was then compared. Discrepancies were resolved and both authors agreed with the final classification. Within each interview, between one and six different themes were mentioned as having an impact on web development in the company. The mean and modal number of themes per company was three. This initial analysis would suggest that factors shaping web development are multidimensional, and that certain influences are linked.

The 11 themes were developed into questionnaire statements, designed to measure their relative influence on the company's web development. The "client expectation" theme was developed into two statements to reflect the differences between individual and corporate clients. The frequency of the incidents mentioned in the qualitative phase may not have reflected the importance of the incidents, for example respondents may be reporting memorable events rather than critical incidents (Fountain, 1999). Therefore to test whether the incidents actually were "critical" respondents

Table I Classification of critical incidents

No. of times


theme mentioned


Strategic decisions


"We had an idea of really trying to enable clients to transact business online"

Spending on computers/IT


"The biggest step forward was us getting broadband to be honest"

Specific person leading web site


"I had a bit more time to really try and forge the way forward ... I saw it as part


of my role"

Key person outside company


"We had a guy ... who came in to run our web site"

Key person inside company


"It was driven by the owner"

Web site development services


"The technology has become available"

Software availability/providers


"I came across [software company] they were pushing their web site which had

all the functionality that we were trying to get"

Competitors' actions


"Everybody's got a web site ... we should just have one"

Internet boom


"It was the sort of internet boom that was going on"

Client expectations


"[academics]... they tend to have access to the internet on a day to day basis

... expect communication by email"

Provider developments


"They [provider] had developed a system where we as broker could input stuff

into their mainframes"


54 statements

17 companies

Tina Harrison and Kathryn Waite were asked to rate their importance on a five-point scale (Andersson and Nilsson, 1964). The extract from the questionnaire is shown in the Appendix. Measurement in this way allows for differences in the relative importance of the influences within any adopter categories to be assessed.

In addition to the critical incidents, the questionnaire also captured data on the web presence of the IFA including length of time on the web, how the web site is being used and key company and individual characteristics. The survey instrument was piloted among the same sample of interviewees for clarity, relevance and to estimate a response rate. The pilot was particularly useful in serving as a check of accuracy and relevance for the CIT categories, as well as resulting in a small number of changes to other items. Of the 20 companies interviewed, 16 responded to the pilot within the timeframe requested. The final questionnaire was mailed to a random sample of 5,000 IFAs stratified by size of firm. The sample was purchased from a commercial database. Questionnaires were distributed in June 2004, with a follow-up in July 2004. A total of 692 usable questionnaires were returned, yielding a 14 per cent response rate. There are 26,000 active advisors in the UK working in an estimated 5,000 firms, of these 37 per cent are operating as sole practitioners (Sandler, 2002).

Since the prime objective of the study is to provide an insight into the development and adoption of web sites by IFAs, only a sub-set of the sample was relevant to the analysis (217 firms). The sub-set therefore consisted of those firms that have a web site and have all been in business for at least the same length of time as the company which has had a web site for the longest length of time[1]. The longest any firm in the sample had a web site for was 10 years, hence the analysis focused on all companies who had been in business for at least 10 years. This represented 388 firms, out of which 44 per cent had a web site, 12 per cent were in the process of developing a web site, and 44 per cent did not have a web site. The final sample, thus, consisted of the 217 companies who either had a web site or were in the process of developing a web site.

0 0

Post a comment