Having identified the amateur writing community as one which could benefit from ad- ditional semantic data and services a number of methods were identified to garner in- formation about user requirements.
Direct interaction with members of the community through observation, questionnaire and interviews was used to gain understanding of the specific difficulties and issues that community members face. Following initial discussions with community members as part of a preliminary assessment, a questionnaire was used to gain greater insight into the issues that had been highlighted. This method was chosen for the initial data gathering because it allowed us to involve a larger section of the community than would have been possible through alternative methods and it allowed us to retain the most control over the way that the questions were presented and the ways in which they could be answered (Andrews et al., 2003).
Because of the nature of the community and the broad range of people involved it was considered necessary to get the opinions of people outside the community who might have an interest in it. The most obvious group who fit this profile were the parents and guardians of the younger members of the community. The questionnaire sought to cover a number of areas from general information about the respondents to their reading habits, their social involvement within the community and their preferences and opinions on the access and distribution of different types of material within the community. The questionnaire was deliberately designed to be broad in scope to allow opportunity for
the respondents to identify the issues which were important to them. This was intended to minimise the influence of the researcher’s preconceptions of the respondents’ relative priorities. The questions themselves were divided into seven groups, with some sections aimed at specific participants. The full questionnaire is available in Appendix A.
No. Section Description Respondents
1 Questions related the respondent’s familiarity
with amateur fiction online, their age and the age of adulthood in their region
Everyone
2 Questions related to preferences for story ac-
cess and display and level of community in- volvement
Readers
2b Questions related to access controls Under-age Readers
3 Questions related to practice regarding story
access and display and publishing methods
Writers
4 Questions related to how people outside the
community came into contact with amateur writing online
Interested Others
5 Questions related to specific sub-types of am-
ateur fiction and opinions on the access of “adult” content
Everyone
6 Questions related to technical knowledge and
ability
Everyone
7 Questions relating to blocking and filtering
amateur fiction and the collection of personal information
Everyone
An eighth section was added shortly after the questionnaire was made public to allow anonymous feedback and comments since while contact details were made clear they were via e-mail and therefore not as anonymous as the questionnaire. 206 responses to the questionnaire were recorded before this section was made available and 912 after. Of those 912 people who had the option, as part of the survey, to leave an anonymous comment, 157 exercised the option to do so. Given the comparatively low rate of response to this option we do not consider that the initial option, additional comment by e-mail only, adversely affected the integrity of the survey.
The draft version of the questionnaire was shown to the members of the Fan Fiction
Ontology Yahoo Group6. This group was created by the author on 13rd April 2004 to
facilitate discussion with and within the community about the creation and use of an on- tology for the description of fan fiction and other creative fan endeavours. Members were
6
also concerned with defining and comparing the vocabulary used within the community. At that time, 12 November 2004, the group had just over fifty members drawn mainly from the fan writing community. As of 21st September 2007, the group has eighty-nine members. Changes were made to the questionnaire based on feedback received from those members. The survey was announced to a few select people on 3rd December 2004. This small group which consisted of known adults and children responded to the questionnaire over the subsequent weekend. One possible issue was raised as a result of this preliminary trial, however the problem proved to be down to user error rather than a fault of the code running the survey pages. The line numbers relating to this query and resubmission were noted so the superseded result could be removed during analysis. Having been reassured that the survey was in full working order it was advertised to the public starting on 7th December 2004. Invitations to participate took the form of posts on LiveJournal, e-mails to persons known to have an interest and posts to the mailing lists and bulletin boards of related interest groups. A number of archives and fan sites mostly related to Harry Potter were also contacted with a request that they publicise the survey to their users. Since those contacted directly were mostly adults or adult dominated groups, the general Potter archives and websites were contacted in the hope of getting more younger fans. This last received a limited response, but at least two archives did post notices. Notices were also posted in a number of places both around the university campus and at other locales where it was thought they might be seen. This was done to solicit opinions from those who were not directly involved in the amateur writing community but had some contact with it via friends or relations and therefore wanted to express their views as a outside observer.
Due to the high number of responses, the decision was made to harvest the first set of results after the survey had been running for two weeks so as to ease processing and analysis. It was estimated that by that point there would be in the region of 1000 responses and this was seen as a reasonable dataset to work with. When harvested, this first set of data was found to comprise of 1118 responses, of these one was known to have been superseded and, on investigation of the data, one was shown to be a double. The remaining 1116 responses included 4 from prior to the 7th December 2004 but otherwise spanned the 7th December 2004 till noon on 21st December 2004. Analysis of the IP address logged as part of the duplication identification process suggested that responses had come from over thirty countries. While America, Great Britain, Australia and Canada ranked the highest of the known contributing nations, the international nature of the interest can be seen in the contributions from countries such as Finland, Russia, Brazil, Singapore, Japan, Estonia, Israel, India and Argentina as well as most of the nations in western Europe. The basic analysis of the first wave of results has been completed. The information gained from this has been used to inform the design of the ontologies and hypothesised applications described in the rest of this thesis.
produce an easily readable result (see Appendix B.2 for full, aggregated results).