The two types of survey taken into consideration were a mail survey and an e-survey. The former is reported to be the most effective method of surveying a membership organisation (Veal 1997). This delivery method was not undertaken though, due to the cost of printing 360 questionnaires and the accompanying letters, as well as purchase of envelopes and stamps. These expenses would increase substantially if reminders were to be sent.
Therefore, it was decided that an e-survey would be employed.
The advantages of an e-questionnaire included low cost of conducting the survey and the possibility of instant analysis of the collected data (Veal 1997). An additional benefit was the possibility of designing a visually attractive questionnaire. Furthermore, online surveys were selected as they are eco-friendly, owing to the non-use of paper (Survey Monkey 2007), an approach reflecting the basic principles of the Forestry Commission, the administrator of the locations which are a focus of this study. An e-survey is also more advantageous for the respondents as they are allowed to take their time to answer the questions. At the same time, completing a survey online is quicker in comparison with filling out a postal survey (Survey Monkey 2007).
The biggest disadvantage can be the general perception of some e-surveys as part of the increasing volume of ‘ junk e-mail’, which might play a role in response rates (Veal 1997). The spam and personal data protection concerns were addressed by selecting a legitimate and reputable organisation (IMBA U. K.) to distribute the e-questionnaire. Another disadvantage of a web-based survey, just like a postal one, is its limited administration, which can negatively affect the response rate (Survey Monkey 2007). Further drawbacks of using a web-based survey include technical faults and multiple submissions by the same respondent. Although it was impossible to prevent the former issue, the latter was solved by the use of a tool allowing only one response per computer.
2. 3. Survey instrument
A two-page, self-administered fully electronic questionnaire was designed using an online survey tool surveymonkey. com. The questionnaire design process offered by Veal (1997) was adopted in the present work. Firstly, literature review was conducted to identify conceptual problems and research questions. Secondly, a list of information required to address the issues as created. Thirdly, questionnaire was selected as a method which would meet the information requirements. The questionnaires used by Green (2003) in a survey on IMBA U. S. members, as well as a study by Cessford (1995) on New Zealand mountain bikers served as a blueprint in the questionnaire design process. The factors examined in the questionnaire can be divided into five groups corresponding to the research objectives:
1. mountain bikers general characteristics (type of mountain biking participated in, level of advancement, frequency of participation);
2. mountain bikers’ demographic and socio-economic characteristics (age, gender, household size, education, occupation and income);
3. the Forestry Commission locations used by mountain bikers (identification of the most popular mountain biking centres in Great Britain);
4. travel patterns of mountain bikers and trip-related factors in decision making (importance of promotional channels and of various destination features in destination choice, spend, means of transport and accommodation used, use of tour operators);
5. the importance of adventure components in mountain biking and the motivations of mountain bikers.
The e-questionnaire consisted of 21 questions, 15 of which concerned the respondent’s mountain biking experience, while the remaining 6 their demographics (see Appendix 4). The demographics section was presented on the second page, as inserting all 21 questions on a lengthy single page might discourage respondents from completing the survey. 19 questions were closed, while only 2 open-ended. It was decided that pre-coded questions are a more reasonable option since the questionnaire was respondent-completed and, as reported by Veal (1997), open-ended questions are too time consuming, which can negatively affect response rate. As regards question techniques, 2 kinds of rating scales were employed, namely Likert scale with simple YES/ NO answers, as well as semantic differentials indicating the degree of importance of a particular feature using 5-point scale (1 – not important, 2 – quite important, 3 – important, 4 – very important, 5 – extremely important). These two scales were not only easy to construct and administer, but also respondent friendly. Several questions were checklist or multiple choice type, and only 2 were open-ended, though they did not require descriptive answers. The ordering format was based on Veal (1997), who suggested sequencing questions in the following order: easy, relevant and personal. The questions were kept compact and simplified wherever possible, and clarity of layout was assured by using the online survey tool.
Introductory remarks specifying the purpose of the survey were included on top of the questionnaire, while confidentiality and anonymity were ensured in remarks preceding the demographic section of the questionnaire. In addition, a thank you page was created at the end of the questionnaire. These measures, coupled with the survey participation request posted on IMBA U. K. forum in advance, were employed to maximise response rate. In order to ensure validity of the questionnaire-based data, ‘ dummy’ categories were included. In question 4, listing the Forestry Commission managed mountain biking centres, 3 non-existent locations were added – Mammoth (ENG), Gutter Valley (SCO) and Badger Trail (WAL). In question 8, the same feature was repeated twice under different wording (‘ Strong mtb community/ culture’ and ‘ Strong mountain biking community/ culture’). This approach is suggested by Veal (1997) as a tool of measuring the degree of error in responses.
In the pre-testing stage of the questionnaire design, a 21-question draft was e-mailed to 10 members of IMBA UK forum who had earlier declared that they would be interested in completing it. The purpose of the pre-test was to determine whether the instructions and questions were interpreted in a manner which had been intended (Finn et al. 2000). Specifically, wording, sequencing and layout of the questionnaire were to be tested (Veal 1997). Most of those who replied described it as “ easy to complete and not time consuming”. Nevertheless, problems in two questions were indicated. Firstly, extending the list of Scottish mountain biking sites in question 4 was suggested. Upon reviewing the Internet sources, a total of 8 sites were added to the list. Secondly, it was signalled that question 13 asking about an approximate spend per trip was not precise enough. This issue was addressed by specifying the areas of spending that the respondents should consider (travel, food, drink, accommodation).