Network Analysis and the organisation of events

1 introduction

2 Research Methodology
2.1 Introduction
2.2 Literature Review
2.3 Research Methods
2.3.1 Qualitative Research: Action Research
2.3.2 Quantitative Research: Survey Research
2.3.2.1 Sampling
2.3.2.2 Questionnaire Design and Pre-testing
2.3.2.3 Reliability and Validity
2.3.2.4 Data Coding and Analysis
2.3.2.5 Ethical Issues
2.4 Summary

3 Conference Management
3.1 Introduction
3.2 Project Management Process: Key Phases
3.2.1 Strategic and Operational Planning
3.2.1.1 Strategic Planning
3.2.1.1.1 Vision and Mission
3.2.1.1.2 Objectives
3.2.1.1.3 Strategic Plan
3.2.1.2 Operational Planning
3.2.1.2.1 Single-use Plans
3.2.1.2.2 Standing Plans
3.2.2 Controlling and Monitoring
3.3 Meeting Planning Software
3.4 Summary

4 Project Network Techniques (PNT)
4.1 Introduction
4.2 The Origin of PNT
4.2.1 The Gantt Chart
4.2.2 Critical Path Method (CPM) and Program Evaluation and Review Technique (PERT)
4.3 How To develop a Cpm chart
4.3.1 The Logic behind CPM
4.3.2 Development Process
4.3.2.1 Project Plan and Work Breakdown Structure
4.3.2.1.1 Level of Network Detail
4.3.2.1.2 Graphical Representation ― Basic Rules
4.3.2.1.3 The Unique Network
4.3.2.2 Time Estimation
4.3.2.3 Basic Scheduling
4.3.2.3.1 Forward Pass Computations and Total Project Time
4.3.2.3.2 Backward Pass Computations: Path Float and Critical Path
4.3.2.4 Resource Allocation
4.3.2.5 Project Control
4.3.2.5.1 Time Control
4.3.2.5.2 Cost Control
4.4 Computer-drawn networks
4.4.1 Advantages
4.4.2 Comparative Features
4.5 Summary

5 Research report
5.1 Introduction
5.2 Milestones
5.3 Survey Data
5.3.1 General Data
5.3.1.1 Response Rates
5.3.1.2 Frequencies
5.3.1.2.1 Company-specific Frequencies: Nature of Organisation – Number of Employees – Events organised per Year
5.3.1.2.2 Respondent-specific Frequencies: Age – Sex – Position
5.3.1.2.3 Rates of Refusal
5.3.2 Reasons for not using PNT: Frequencies
5.3.3 Planning Techniques: Frequencies
5.3.4 Use of Meeting Planning Software: Frequencies
5.3.5 Meeting Planning Software Features: Frequencies
5.3.6 Cross-tabulations: Use of Meeting Planning Software – Demographic factors
5.3.6.1 Age
5.3.6.2 Sex
5.3.6.3 Position
5.3.7 Cross-tabulations: Use of Meeting Planning Software – Company-specific factors
5.3.7.1 Nature of Organisation
5.3.7.2 Number of Employees
5.3.7.3 Number of Events/ Year
5.3.8 Use of Computer-stored Tasklists – Use of Meeting Planning Software: Cross-tabulation
5.3.9 Analytical Portrait of the Organisation that uses PNT

6 Conclusion

REFERENCES

index

glossary of terms

APPENDIX

Chapter

Figure 2-1:

Research Methods

Figure 2-2:

Qualitative Research

Figure 2-3:

Quantitative Research

Chapter

Figure 3-1:

Project Management and PNT

Figure 3-2:

Project Life Cycle

Figure 3-3:

Strategic and Operational Planning

Chapter

Figure 4-1:

Gantt Chart

Figure 4-2:

Work Breakdown Structure and Project Plan

Figure 4-3:

Activity with Predecessor I

Figure 4-4:

Activity with Predecessor II

Figure 4-5:

Fully Calculated Network

Figure 5-1:

Analysis Steps

Table 5-2:

Response Rates

Table 5-3:

Nature of Organisation - Frequencies

Figure 5-4:

Number of Employees – Frequencies (Original Values)

Figure 5-5:

Number of Employees – Frequencies (Re-coded Values)

Figure 5-6:

Number of Events/ Year – Frequencies (Original Values)

Figure 5-7:

Number of Events/ Year – Frequencies (Re-coded Values)

Figure 5-8:

Age – Frequencies (Original Values)

Figure 5-9:

Age – Frequencies (Re-coded Values)

Figure 5-10:

Sex – Frequencies

Figure 5-11:

Position – Frequencies (Original Values)

Figure 5-12:

Position – Frequencies (Re-coded Values)

Table 5-13:

Reasons for not using PNT - Frequencies

Table 5-14:

Use of Meeting Planning Software*Reasons for not using PNT – Cross-tabulation

Table 5-15:

Planning techniques - Frequencies

Figure 5-16:

Use of Meeting Planning Software – Frequencies

Table 5-17:

Features Meeting Planning Software – Frequencies

Table 5-18:

Use of Meeting Planning Software*Age – Cross-tabulation

Figure 5-19:

Use of Meeting Planning Software*Age – Yes/No Distribution within Age Groups

Figure 5-20:

Use of Meeting Planning Software*Age – Distribution Age Groups within Yes/No Categories

Table 5-21:

Use of Meeting Planning Software*Sex – Cross-tabulation

Figure 5-22:

Use of Meeting Planning Software*Sex – Yes/No Distribution within Sex

Figure 5-23:

Use of Meeting Planning Software*Sex – Distribution Sex within Yes/No Categories

Table 5-24:

Use of Meeting Planning Software*Position – Cross-tabulation

Figure 5-25:

Use of Meeting Planning Software*Position – Yes/No Distribution within Position

Figure 5-26:

Use of Meeting Planning Software*Position - Distribution Position within Yes/No Categories

Table 5-27:

Use of Meeting Planning Software*Nature of organisation – Cross-tabulation

Table 5-28:

Use of Meeting Planning Software*Number of Employees – Cross-tabulation

Figure 5-29:

Use of Meeting Planning Software*Number of Employees – Yes/No Distribution within Number of Employees

Figure 5-30:

Use of Meeting Planning Software*Number of Employees – Distribution Number of Employees within Yes/No Categories

Table 5-31:

Use of Meeting Planning Software*Number of Events/ Year – Cross-tabulation

Figure 5-32:

Use of Meeting Planning Software*Number of Events/ Year – Yes/No Distribution within Number of Events/ Year

Figure 5-33:

Use of Meeting Planning Software*Number of Events/ Year – Distribution Number of Events/ Year within Yes/No Categories

Table 5-34:

Use of Meeting Planning Software*Computer-stored Tasklist – Cross-tabulation

Abbildung in dieser Leseprobe nicht enthalten

1 introduction

“In effect, arranging a conference is like drawing together to a single point a large number of threads of different lengths and from different places. People and suppliers of goods and services have to be briefed, informed, bullied and persuaded. Lead times are interdependent and variable. It is the perfect task for network analysis, but few would wish to adopt this method — perhaps justifiably, for it is not the plan but the result which counts in the end.”

(Seekings 1996, p. 363)

To cope with the task of planning and organising events, which includes the co-ordination of a multitude of different and interdependent activities, Seekings suggests using network analysis, or Project Network Techniques (PNT). These techniques, which are project management techniques and usually computer-based, have been applied, since their development, “with notable success [highlighted by the author] to research and development programs, all types of construction work, equipment and plant maintenance and installation, introduction of new products or services or changeovers to new models, development of major transportation and energy related systems, strategic long-term planning, emergency planning etc.”^[1]. Meeting planners, however, might be reluctant to make use of this highly developed planning and monitoring technique. It is undoubtedly due to this reluctance that conference management software does not provide network features. Only one supplier was found on the Internet who has integrated network analysis in its meeting planning product ― in the itinerary planning module. Obviously, project network techniques are not the conference organiser’s favourite methods. Yet, there are no statistics to support this hypothesis. To fill this gap, a questionnaire survey among British conference organisers was carried out within the framework of this thesis. The overall research questions the author tries to answer are how widespread the use of PNT is in conference management, on which internal and external factors the use of these techniques depends and which features are perceived as most useful. These general questions can then be split up as follows:

- Does the use of network techniques depend on:

§ The conference organisation:

- nature: corporate, association, government, hotel, independent PCO;
- size: number of employees, number of events/year;

§ The event:

- type: conference, incentive event, product launch, staff training, exhibition, etc.;
- size: delegate number, length, number of involved organisations/suppliers;
- regular or ‘one-off’;
- lead times/deadlines

§ The individual in charge of the project:

- demographic factors: age, gender, position;
- specialist skills: IT, project management;
§ The client who requires the conference organiser to work with a formal, standardised project management system?

- Are PNT’s used during the whole organisation process, including planning AND monitoring, or only during the initial planning phase?

- Which features are used, only the basic ones such as scheduling, or also the more complex ones such as cost and budget control?

One possible outcome of the research might be that meeting planners find it inappropriate to use a sophisticated planning tool like PNT for smallüscale or regular events the organisation of which they perceive to be routine work. They possibly prefer, especially if they lack the appropriate IT skills, a simple hand-written or typed administrative schedule which constitutes a time-table of what has to be done week by week and by whom.

The main body of this thesis consists of three parts:

¬ Research Methodology

This methodological chapter is dedicated to the detailed description of the research approach and techniques which have been used to answer the research questions.

­ Literature Review: Conference Management and Project Network Techniques

This theoretical part is designed to give an overview of existing literature in the field, aiming to put the research topic into context.

® Research Report

The report of the research study tells the story as it unfolded. It contains an analysis of the data gathered and includes accounts of problems that occurred during the implementation of the research project.

The main body is followed by the conclusion, or overall evaluation, which links the results of the survey to the research topics.

Abbildung in dieser Leseprobe nicht enthalten

2 Research Methodology

2.1 Introduction

This chapter defines the methods chosen for the research project and identifies strengths and weaknesses. It also discusses an alternative approach which was used by the author in the early stages of the research project, but later abandoned due to implementation problems. Some consideration will be given to the issues relating to sampling and selection, and to the limitations of the study. The concepts of reliability and validity of research data will equally be explained and applied to the specific research approach.

The research style to be used largely depends on the nature of the research project and of the questions being asked. Usually, the research questions and the rationale immediately suggest particular methods. This chapter will therefore deal with those methods which, for this specific research project, best promised to match the problem and proved to be the most suitable amongst the wide variety of styles available for designing, carrying out and analysing the results of research.

The structure of the following chapter is based upon the research process model proposed by Blaxter et al. (1996, p. 59). This model includes three dimensions called research families, or general strategies for performing research (quantitative, qualitative), research approaches (action research, case studies, experiments, surveys) and research techniques (documentary analysis, interviews, observation, questionnaires). To study the particular research question, the author focused, as the highlighted words demonstrate, on a quantitative research strategy. It should be noted that there will be overlaps in the description of the three dimensions because the alternatives chosen for the research project in hand are all from the ‘quantitative stable’ and therefore have a close affinity with each other.

As no one method can answer all questions and provide insights on all issues, a multi-method approach, referred to as triangulation, has been adopted in this thesis. To verify the validity and reliability of the information collected during the research process, the main method – that is the questionnaire survey, has been complemented by the exploration of existing literature on the research topics.

The following diagram is designed to show how the different parts of this chapter relate to each other. At the same time, the diagram illustrates which decisions were taken before the implementation of the final research action plan.

Abbildung in dieser Leseprobe nicht enthalten

Figure 2-1: Research Methods

2.2 Literature Review

The research project started with the exploration of existing literature, including monographs, journal articles and websites, in order to get a more detailed perspective of the chosen topics – ‘Conference Management’ and ‘Project Network Techniques’. The author restricted herself to using sources from 1990 to the present day make sure that the information they contain is recent. The background reading equally enabled the author to ask the ‘right’, or relevant, questions in the survey questionnaire.

Literature about network techniques was primarily reviewed to find out what technical features they provide and which weaknesses and strengths they have. There do not seem to be any limits to their applicability, on the contrary: all the authors believe that network techniques can be successfully used for all sorts of different projects, irrespective of their size or nature. If that is really so, the survey results will tell. At least Rogers believes that the size of the event to be organised dictates appropriate planning tools:

“The more complex the event and the numbers involved in organising it, the more the need for some form of critical path analysis, mapping out the sequence of events in a logical order and within a realistic time-frame.”

(Rogers 1998, p. 137)

At any rate, there is one thing that all projects have in common: the need for project management, and PNT is one technique which may help to make the project management job easier. Lockyer (1996): Project Management and Project Network Techniques gives, as the title suggests, valuable insights in the inter-relationship between project management and project network techniques.

The conference business has not attracted significant attention from researchers and writers until comparatively recently; the availability of literature related to this topic is therefore limited. It was only in the last century that several text books were written and webpages created which provide up-to-date information, including statistics, about the meetings industry and that predict trends for the 21st century; an example is the comprehensive description of the conference business in the UK and Ireland found in Rogers (1998): Conferences - A twenty-first century industry. When reviewing the literature, the author found out that no logical link is made between conference management and project management, although conference management is actually project management. Texts about the organisation of events, such as Seekings (1997): How to Organise Effective Conferences and Meetings, only describe, in more or less detail, which activities are involved in the planning. These texts, which include the most diverse online planning checklists, do not give any information about how to reconcile and monitor all the different tasks involved in a project. True, it is commonly recognised that the conference management requires “common sense, forethought, meticulous planning and attention to detail, team work and sometimes crisis management”^[2], and that much of the administration can be enhanced by the use of conference management software packages, which are designed to reduce the routine work-load on the meeting planner: delegate registrations and correspondence, itinerary planning, invoicing, report production, and other aspects. However, the organisation of a conference does not seem to be considered a project that requires systematic project management. Such an approach includes not only operational planning, but also strategic planning (® chapter 3.3.1) and monitoring/evaluation procedures. Some authors, like Rogers, refer at least implicitly to a project management process:

“The organisation of a conference requires a similar strategic approach to that needed for planning and managing most other events. Clear objectives should be set from the beginning, a budget has to be prepared and the conference managed for its duration. Then, after it is over, final administrative details have to be completed and some evaluation of the conference should take place.”

(Rogers 1998, p. 135)

As theoretical background information about conference-related project management is not available, the author switched to studying literature about project management in general and then applied the key issues to the special requirements of conference operations. In addition to this, the author used Donnell et al. (1999): Festival and special event management, a textbook that makes an attempt to approach the planning and monitoring of special events systematically. To conclude: the conference-related literature does not tell the reader very much about conference management, the review only helped to define the role that planning plays in the organisation of an event and to establish which activities are involved in the planning.

It should be noted that the purpose behind the PNT-related literature review changed during the research process. The author started to approach the research study by doing action research which included creating a network for a particular conference. Literature about project network techniques was therefore primarily used for pragmatic reasons, i.e. to learn how to develop a network. The reader should keep this initial practical objective in mind when evaluating the chapter on PNT. This chapter will give a detailed introduction into the principles of network techniques, helping to understand how the use of PNT can make the task of planning and monitoring easier. It must be admitted that the above mentioned checklists, which have been criticised being somewhat ‘reductionist’, have proved very useful for the preparation of the network.

While secondary sources were particularly useful in the earlier stages of the preparation of this thesis, primary information forms the substance of the research itself. The need for primary data results from the lack of statistical information about conference organisers and their preferred planning tools and techniques. The following chapter will describe the approach and methods which were used to compensate this lack.

2.3 Research Methods

Two different research styles would have been suitable to answer the research questions raised in this thesis: quantitative and qualitative research. The author first decided in favour of the qualitative method which included action research in a meetings industry-related organisation and a brief case study of the chosen project and organisation. After initial commitment on the part of the conference organiser selected for the research project, the project had to be abandoned before the implementation of the computer-based network due to lack of resources in terms of time, IT skills and equipment/software. The early end of the project made the author think about to what extent one can make general assumptions based on this single case for the whole conference industry. To answer this question, the author finally decided to switch over to quantitative research and to conduct a questionnaire survey among conference organisers in the UK.

2.3.1 Qualitative Research: Action Research

Abbildung in dieser Leseprobe nicht enthalten

Figure 2-2: Qualitative Research

(Extract from figure 2-1: Research Methods)

The action research approach, which is a qualitative approach, had been selected for both theoretical and personal reasons. Action research, which deals with small samples, is an approach very much based on practice, having a practical purpose in view. It is an attempt to actively change, or improve a situation in a defined context.^[3] At the same time, action research is very much an interactive process which is based upon the close collaboration and co-operation of the researcher and the practitioners, enabling the investigator to gain an insider’s view of the field.^[4] Action research was therefore the approach that best matched the author’s ‘learning for life’ attitude. But the decision to use the action research approach was not only motivated by personal ambition, it was also based on the belief that a full and rounded, holistic understanding of the network implementation process, however unrepresentative it may be, is of more value, especially in terms of identifying strengths and eventual pitfalls, than a more limited understanding gained from studying a large, representative number of examples.

At different stages of the action research project, the author used, or planned to use, different techniques for data collection; some of them ― the structured interviews ― were even taken from the quantitative stable. In the preliminary stages, when it came to the case study, promotional material and websites served as sources of current and historical data about the conference organiser. In addition to this, informal interviews were conducted with the head of the office. To be able to structure the project and to create the network, online checklists and records of the previous conference were traced and studied. Apart from the document analysis, several interviews with the project manager were conducted. The interviews took place both face to face and remotely, by e-mail. Considering the purpose of the interviews ― to get access to detailed information about a particular event and the activities of which it consisted ― these interviews were not in-depth interviews, the kind of interview one would expect within the framework of a qualitative research study. Rather than taking the form of a free-flowing conversation, they were tightly structured and guided by a series of formal, pre-determined questions (‘oral questionnaire’).

To get a feedback from the person who was supposed to work with the network, the author had planned to followed up the research project with a post-conference debriefing in the form of a semi-structured interview, including formalised close-ended questions, as well as the more flexible open-ended questions which allow for fuller answers. As the data collected during qualitative research projects usually are not readily quantifiable, data would exclusively have been presented in report form.

Validity and Reliability

The problem of adequate validity and reliability is a major criticism placed on qualitative research methods. Because of the subjective nature of qualitative data and its origin in single contexts, it is difficult to apply conventional standards of reliability^[5] and validity^[6]. Contexts, situations and interactions cannot be replicated to any extent nor can generalisations be made beyond the particular case. Rather than replicability, reliability in action research is more focused on creating results make sense and are agreed on by all concerned. In action research, ‘theories’ depend not so much on ‘scientific’ tests of ‘truth’ as on their usefulness in helping people to function more efficiently. The theories are validated through practice, and generalisability is often left up to the reader, who may ask to what extent he can relate to the content of the study to his own situation.^[7] To improve reliability and enable others to replicate the work, the steps and procedures are clearly and explicitly documented in the final report. Ways of improving reliability and validity in action research involve triangulation, meaning the use of multiple sources of evidence to demonstrate convergence of data from all sources.^[8] The author therefore planned and started to use a diversified range of secondary and primary data, the latter ranging, as already mentioned, from files and documents provided by the conference organisers to structured and semi-structured interviews.

Abbildung in dieser Leseprobe nicht enthalten

2.3.2 Quantitative Research: Survey Research

Figure 2-3: Quantitative Research

(Extract from figure 2-1: Research Methods)

While qualitative research tends to deal with a small number of examples, the setting is considered an integral part of the primary data, quantitative research involves the analysis of relatively large-scale and representative sets of data in numeric form. As the research project in hand required gathering quantifiable data from the specific population of professional conference organisers, the author decided to carry out a questionnaire-based survey among a sample of this target-group. The target-group included:

§ Individuals: independent or in-house

§ Conference planning organisations.

These conference organisers were drawn from associations, corporations, the government, hotels, conference centres, educational institutions, or other sectors. For the purpose of the survey, the term ‘conference’ was used as a generic term for all gatherings of any size and nature, including even incentives and exhibitions, which are often attached to a conference. The organisation of family parties, however, was excluded.

The survey was limited to Great Britain. This country had been chosen for the simple reason that the author focused on the conference business in the UK in her study of the relevant literature; this knowledge of the structure and the characteristics of the British conference industry was required for creating the questionnaire.

The questionnaires were completed by the respondent, and initially delivered to the intended respondents by e-mail. It was assumed that the target-group were highly likely to have e-mail since fast, paperless communication ― with clients, delegates and suppliers is at the heart of the conference business. During the data collection process, however, the medium had to be changed due to the intended respondents’ fear of macro viruses; the questionnaires were therefore sent by fax. Respondent completion was the only practicable approach because the intended respondents are based in many different locations throughout the country, which would have made a face-to-face survey impossibly time-consuming and expensive. For the same cost reasons, a telephone survey was another option that was rejected for cost and time reasons. However, self-completion often results in low response rates and poorer answers because there is nobody in persona to convince people to take part in the survey nor to clarify questions for respondents. Low response rates can introduce bias in the results because the people who choose not to respond or are unable to respond may differ from those who do respond.^[9] There is one major advantage of respondent completion the author considered being likely to substantially increase the response rate: the respondents are free to answer in their own time and in their own environment. Furthermore, self-completion is the best way to avoid the results of the survey being influenced by the personal appearance, mood or conduct of an interviewer.^[10]

Abbildung in dieser Leseprobe nicht enthalten

2.3.2.1 Sampling

Abbildung in dieser Leseprobe nicht enthalten

Questionnaire surveys usually involve only a proportion, or sample, of the target-group in which the researcher is interested. The framework for the sample consists of the following directories which were all updated in 2000:

- International Association of Professional Conference Organisers (IAPCO)

at: www.iapco.org

- British Association of Professional Conference Organisers (BAPCO)

at: www.bapco.co.uk/members.htm

- British Universities Accommodation Consortium (BUAC)

at: www.buac.co.uk

- Connect Venues

at: www.connectvenues.co.uk

-
Yellow Pages UK

at: http://search.yell.com/search/DoSearch

The group taken from the Yellow Pages UK, which is the largest one, consists of two sub-groups, the ‘Conference organiser’ sub-group and the ‘Conference centre’ sub-group^[11], the latter containing conference venues of any nature. The use of the Yellow Pages raises the following problems:

§ The sub-groups contain entries from related services, such as conference technology. As far as possible, the author tried to exclude such organisations by checking the organisation’s website. This does not rule out the possibility that related services are included into the sample. Of course, this will affect the response rate because people who do not feel they are the intended respondents will not complete the questionnaire.

§ The Yellow Pages include only a small amount of ‘one-man / one-woman’ organisations so that this category is under-represented. Conference organisers working in or for the government or an association are also under represented because the Yellow Pages listings is mainly comprised of independent PCOs and conference venues.

§ The ‘Conference centre’ sub-group may include in-house staff who organise conferences ‘on the side’, planning meetings comprising only of a small segment of the individual’s job description. The author solved this problem by clearly stating in the cover letter that the questionnaire should be filled out by people whose sole responsibility consists in organising conferences. The questionnaire asked for the job-description of the person responding, as conference organisers in venues such as hotels may have names other than ‘conference organiser’.

Using membership lists or directories as data source raises an important question: How well does the sample group represent the eligible target-group? In other words, what proportion of the eligible target-group is represented? For the research project in hand, the target-group definition and the sample group do not match because people who are not members of the associations mentioned above or who are not listed are excluded. It is not possible to find out about the percentage of the target-group missing nor about the behaviour and attitudes of those who are not included in the frame. Of course, this will result in bias in the survey.^[12] Given the fact that there is no official statistical information about conference organisers in Britain, the author had to ignore the bias that results from excluding ‘non-members’. Data about the target-group was only found in Rogers 1998, pp. 79f.: There are an estimated 10,000 conference organisers in the UK, many of these combine conference organising with other responsibilities. Full-time conference organisers, whose sole responsibility is the management of conferences and other events probably number less than 1,000. As the intended respondents of the survey in hand are only full-time conference organisers, independent from the job title they may have, the number of 1,000 may serve as reference number.

The sampling technique used to select the sample for this research project is referred to as convenience sampling, where information is gathered from those who can be most easily accessed, i.e. those prospective respondents with an e-mail address.^[13] This sampling method is not probability-based. The weakness of non-probability sampling is that, as opposed to probability sampling^[14], the subjects might not be representative of the target-group. Consequently, it is not possible to make any valid generalisation about the target-group from which the survey sample was drawn. Furthermore, the use of non-probability based methods of sample selection indicates that the inquiry results have lower scientific validity (® Reliability & Validity, p. 33).^[15]

Abbildung in dieser Leseprobe nicht enthalten

2.3.2.2 Questionnaire Design and Pre-testing

Abbildung in dieser Leseprobe nicht enthalten

When designing a questionnaire for respondent-completion, great care must be taken with the layout and the way in which the questions are asked. A carefully prepared and totally self-explanatory questionnaire will pay dividends in terms of the level, accuracy and completeness of response. The statements have to be clear and simple, since instructions and questions must be uniformly understood by a wide variety of respondents. A lack of understanding by respondents can and does affect the quality of the survey responses. If respondents do not understand the task or view it as difficult, they will probably not complete the questionnaire or will give unintentionally misleading answers.^[16]

One of the first tasks in designing a questionnaire, however, is convincing potential respondents that the study is important enough for them to devote their personal resources of time and effort to it.^[17] The study in hand was therefore introduced with a cover letter which explains the purpose of the survey and asks for the respondent’s assistance. Encouraging co-operation is crucial to a high response rate. One other major function of the introductory remarks is to ensure the respondents that their replies would be treated confidentially.

A range of formats exists for obtaining information about respondent characteristics, behaviour and attitudes. To ease data analysis, the questionnaire generally consists of questions in a closed format in which a set of numbered alternative responses are specified. Open-ended questions, where respondents can provide free expression of opinions, have not been included in the questionnaire design for three reasons. First, response coding and interpretation difficulties would arise because data from open questions is essentially narrative. Second, the comparability of the information between respondents is difficult to assess. Third, the response rates to open-ended questions can be very low: people are often less inclined or too busy to write out free-form answers.^[18]

The following formats, described in Burns (2000), pp. 572-573 and Blaxter et al. (1996), p. 162, have been used for the questionnaire:

- Multiple choice

Respondents are given a range of possible answers to choose from.

A high proportion of the survey questions are in multiple choice format because this format best matched the purpose behind the survey which consists not so much in finding out about attitudes but about behaviour.

- Ranking

Respondents are asked to rank questionnaire items in order of importance.

-
Scale responses

Respondents are required to indicate their agreement or disagreement with a proposition, or the importance they attach to a factor, using a standard set of responses, for example: very important, important, not important.

The above mentioned, close-ended formats have the advantage of achieving greater uniformity of measurement and of being more easily coded. The likelihood of making the respondents answer in a manner fitting the response category is also greatly increased. However, disadvantages include the possibility of annoying respondents who find none of the alternatives suitable, or forcing responses that are inappropriate. The researcher therefore has no scope to find out the beliefs, feelings or perceptions of the respondents that do not fit into the pre-ordained response categories.^[19] To defuse this situation, most of the questions contain an ‘Others’ category where the respondents were allowed to give answers not fitting into the given response categories.

Before the survey was carried out, the questionnaire was pre-tested on a small number of friends and others directly involved in conference organising to get a feedback on questionnaire layout, questionnaire wording, response categories and question sequencing.^[20] After the draft had been modified in the light of the responses received, the data collection process started. The questionnaire was sent to the intended respondents by e-mail as an attachment. This file contained macros in the form of questionnaire forms, programmed to facilitate data entry. The respondents were given two weeks to complete and return the questionnaire. Those who refused to open the questionnaire, or who were not able to open it, received the document by fax. Non-respondents were sent an e-mail reminder one week after the two weeks deadline. The questionnaire was not attached again, but the author did offer of sending the questionnaire per fax. Five weeks after this follow-up mailing, the author stopped collecting data.

Abbildung in dieser Leseprobe nicht enthalten

2.3.2.3 Reliability and Validity

Abbildung in dieser Leseprobe nicht enthalten

Reliability is the extent to which research findings would be the same if the research were to be repeated at a later date or with a different sample of subjects. This model is taken from the natural sciences where, if experimental conditions are properly controlled, a repetition of an experiment should produce identical results. This is rarely the case in social sciences, because they deal with human beings in everüchanging social situations. This means that the author must be very cautious when making general statements on the basis of the survey conducted within the framework of this thesis. While the use of standardised procedures ― the questionnaires ― ensures a degree of generalisability, any research findings relate only to the subjects involved, at the time and place the research was carried out.^[21] To be reliable, questionnaires need to be repeated with the same individuals after a period of time. This is usually not possible. Furthermore, some answers to questions may change legitimately over time. Other sources of discrepancies in the survey results are measurement errors, non-response and lack of coverage (® Sampling, p. 27). Thus, the researcher and the user of research results should always bear in mind the nature and source of the data, and not fall into the trap of believing that, because information is presented in numerical form and is based on more or less large numbers, it represents an immutable ‘truth’.^[22]

[...]

^[1] Moder et al. 1983, p. 4

^[2] Carey 1997, quoted in Rogers 1998, p. 155

^[3] Blaxter et al. 1996, p. 64

^[4] Burns 2000, p. 448

^[5] Reliability may be defined as the accuracy, stability and relative lack of error in a measuring instrument (Burns 2000, p. 337).

^[6] Validity is the extent to which the information collected by the researcher truly reflects the phenomenon being studied (Veal 1997, p. 35). Internal validity is the property of an investigation which means that the conclusions are correct within that particular investigation. External validity is that property of an investigation which enables its results to be generalised (Preece 1998, p. 105).

^[7] Blaxter et al. 1996, p. 64; Burns 2000, pp. 443, 474, 476

^[8] Burns 2000, pp. 475-476

^[9] Veal 1997, p. 147

^[10] Burns 2000, pp. 581-582

^[11] The terms in quotation marks are the business categories the author searched for entries.

^[12] Czaja and Blair 1996, pp. 16-17

^[13] The sample selected for the survey in hand only comprises of those individuals or organisations who are listed with their e-mail address or whose e-mail address the author could find out over the Internet.

^[14] Probability, or random sampling gives every population member an equal non-zero chance of inclusion in the sample, ie the samples are selected randomly (Czaja and Blair 1996, p. 6)

^[15] Graziano and Raulin 1997, pp. 83, 86; Veal 1997, p. 205

^[16] Veal 1997, p. 182; Czaja and Blair 1996, pp. 33-34, 37

^[17] Czaja and Blair 1996, p. 78

^[18] Burns 2000, p. 424, Veal 1997, p. 167

^[19] Burns 2000, pp. 424, 572-573

^[20] Czaja and Blair 1996, p. 20; Blaxter et al. 1996, p. 163

^[21] Veal 1997, pp. 35-36; Burns 2000, p. 581; Czaja and Blair 1996, p. 94

^[22] Veal 1997, p. 145; Czaja and Blair 1996, p. 37