Chapter 19
Food Festival Research Methods and Approaches
The purposes of this chapter are twofold, to establish the special characteristics that distinguish food and drink festivals from a research perspective and to provide an overview of research methods for those involved with these festivals. Many of the characteristics identified as being peculiar to food and drink festivals may be relevant to other forms of festival, and the reader is invited to further explore the similarities and differences.
The research methods that can be used to research festivals are common to those employed in other disciplines. However, the approach advocated here is to look closely at the unusual characteristics of food festivals as they will have a major bearing on research methods appropriate in any particular situation. The ‘causal texture of the environment’ was a phrase coined by Emery and Trist (1965). In part their work was applied to an analysis of environmental turbulence, and in part this turbulence was related to the industry in which their studies were conducted. As their work related to the post-World War II era, it is unsurprising that their studies included British coal mining. Their understanding is however transferable. It is that the local or micro industry conditions (my words) are important to understanding a managerial situation. As such, the proposition of the contingent nature of research methods is based firmly in the research literature, and the distinctive characteristics of food festivals are explored in this chapter as a prelude to the consideration of the methods of research, or tools to be used in research.
Special Characteristics of Food and Drink Festivals
There are many competent books published that cover research methods for business, the social sciences, leisure, tourism, hospitality, and associated disciplines. This chapter will identify the special characteristics of food and drink festivals with the perspective of highlighting how the general principles and practices of research contained in the literature need to be reviewed in the light of these special characteristics.
The special characteristics of the food and drink festivals relate firstly to ‘food and drink’, and secondly to ‘festivals’. We will begin by looking at what makes festivals ‘special’; we do not use the word ‘unique’ as this would mean that they were a one-off having no points of difference with other activities. Such a proposition of uniqueness would be implausible.
The first attributes that distinguish festivals are that they are a service with all of the characteristics of a service (perishability, heterogeneity, intangibility, inseparability, e.g. Zeithaml and Bitner, 2002) which will be explained later. In addition, these services relate to a consumer product (the festival) that is unlike most others in that it is and can only be experienced briefly, typically for one or a few days a year. Festivals are unlike holidays which are also a typically short experience, but are taken over several days or weeks. Whereas many people take holidays and can be therefore researched both at home and on holiday, festival customers are hard to research other than when at the festival. In many cases there will be no opportunity to use a mailing list for research as even large festivals tend to rely on on-the-door purchase rather than pre-purchases. This differs from large exhibitions where prepurchases with early booking discounts are typical. Festivals such as York and Ludlow are still typically organized by their own committees which are lay, but may include professionals. In contrast exhibitions such as BBC Good Food are more typically run by professional organizers.
As the festival visit is likely to be short in duration and concentrated in experience, normally research will need to be undertaken at the festival, and will have to be either short if it involves interviews, or take no subject time and be in the form of observational research where people are observed and not interviewed. Off-site research would typically depend on the availability of an address database, possibly generated at the festival.
Many festivals have a seasonal identity. The underlying seasonal nature of food and drink is often reflected in the time of year when the festival occurs. Thus the opportunity to market winter or Christmas-specific products in spring, summer, or early autumn food and drink festivals may be difficult. Hence punch, toddy, gluvine (also Glühwein), Christmas pudding or cake, and turkey would be unseasonable at a summer festival, and may not inspire customers to sample or purchase.
Whereas food and drink exhibitions and shows held in large metropolitan exhibition halls may share a consumer expectation with festivals for seasonality of products, festivals held in more rural surroundings will be more closely linked by history or customer perceptions to the seasons. However, would people buy Easter Eggs at the November BBC Food and Drink Exhibition, despite being held at the large and anonymous National Exhibition Centre in Birmingham in England’s West Midlands region?
Organizers of metropolitan food and drink shows may call them festivals, so the metropolitan or rural location may be more important than the name, be it festival or show, for the purposes of the researcher. Some festivals may coincide with key times in the agricultural cycle, such as harvest or planting of crops. This real or perceived linkage to ‘nature’ may endow food festivals with attributes that transcend the more mundane characteristics associated with other products or services, and the religious festivals associated with seasons and the equinoxes often have a resonance with food and drink festivals.
For the researcher, these deeper relationships may be an important issue that could be explored, notwithstanding the risk that such an exploration risks the methodological pitfall of leading questions. A leading question is one where, for example, an interviewee may be fed an idea that had not occurred to them, such as the link between the festival and the season.
Turning to the characteristics of services, perishability is an important issue. At its simplest, an empty table in a restaurant NOW cannot be sold at another mealtime or on another day. Hence services are even more perishable than the freshest food. For a food festival, perishability is a key concept for researchers. A festival has a carrying capacity of visitors, and these numbers may be negotiated with Health and Safety Inspectors and be limited by licensing authorities (the mechanism for setting actual limits will vary depending on the legal jurisdiction). However, individual elements of the festival will have limits to participants that are perishable. If there is a ‘logical’ route around the festival stands or exhibits, then the crush of entrants at busy times may make certain stands or exhibits unapproachable. Later in the day they may be accessible, and it is an important research issue to see if visitors will re-trace their path to see things that they missed first time around.
Similarly there are times when visitors will want to purchase meals at the festival, and engineering a mechanism to stagger demand is worthy of research, as the capacity of food service is unlikely to match the total number of visitors at a single sitting. Managing the catering should be such that all who want to buy food will be served, yet significant redundant capacity engineered out of the system.
Free food and drinks are also typical of food and drink festivals and exhibitions. If used, they are provided as a sales promotional tool. As with mass catering, the traders wish to see a festival where visitors are spread across the opening times, and where they have enough time to sample and can be persuaded to purchase. So again the perishability of the service is crucial. As will be argued later, many of the foods that are sampled are also perishable – in that they are subject to melting (ice cream), microbiological deterioration (spoilage of cooked meats and dairy), etc. Hence, whereas perishability is an issue for all services, it is unusually complex to model and manage in food festivals. In an art festival an empty seat at a ballet performance is a loss that is at least easy to comprehend, but there will be rarely a very large number of concurrent events. Whereas a food festival may have hundreds of concurrent opportunities represented by exhibitors, talks, demonstrations, and other activities.
Heterogeneity is a reflection that services are delivered to a person, often by another person. The experience of each customer is therefore truly unique because the relationship between customer and server is inseparable. The two characteristics of heterogeneity and inseparability are distinct but closely linked. A person providing a sample of cheese to a food festival visitor is trying to generate a sale. They are trying to select a cheese for the visitor to sample that they have reason to believe will please them so that they will buy. This is a complex relationship and reflects traditional sales practices rather than those typically found in supermarkets, where the unique experience of shoppers tends to be increasingly designed out of shopping.
At a broader level, the experience of each festival visitor is unique. Researchable sources of heterogeneity include:
Inseparability is widely discussed in the context of businesses such as restaurants and hairdressers. However, a food and drink festival is a much more complex environment in which to explore the idea. In a restaurant, the customer may typically meet a waiter and a wine waiter. In a hair dresser they meet the stylist and hair wash person. In a food festival they meet and at some level interact with perhaps scores of exhibitors and officials, and hundreds of other visitors. Yet their experience of the festival is coloured by a mixture of individual encounters plus an overall feel of the event. This is a challenging research area, and my own insights and experience suggests two different lines of investigation. Firstly at the individual experience level, I have encountered answers to open ended questions, where respondents have highlighted specific experiences such as the positive and helpful attitudes of stewards. In answer to other open ended questions, respondents have commented specifically on the ‘buzz’ of a festival. Clearly these observations are ad hoc and superficial, and the topic is open to rigorous investigation.
Intangibility is another attribute of a service. Hence the cheese that is sampled may be tangible (literally able to be touched), but the visitor’s state of mind and propensity to buy cheese is probably generated by their intangible experience of the festival as well of being a function of their relationship with the cheese vendor. I recently sampled some very strong smoked cheese which ruined my next sampling of a subtly smoked salmon. So the immediate experience in the festival has a major impact on purchasing behaviour. Researching the visitor experience so that it can be managed to ensure that they have a propensity to sample and buy is a great challenge for festival organizers and researchers.
So far we have looked at the special characteristics of festivals with an emphasis on food and drink. Next we will consider what makes food special, and reflect on this from a research perspective.
What Makes food Special?
Just as service is perishable, so too is some of the food supplied to, and sold at, festivals. However, it is not all perishable. In fact the traditional stance of food festivals means that much of what is sold has been subject to traditional forms of food processing and preservation, such as bottling (canning), salting, smoking, preserving with sugar, fermentation, and in alcohol or oil. Some fresh produce is also sold, typically covering traditional (and commonly rare) breeds of animals and var-ieties of fruit and vegetables which are often subject to more modern forms of preservation such as chilling, freezing, and irradiating as with spices.
Hence the level of physical perishability of foods is limited, but often the regulatory authorities apply standards to food storage that is appropriate for ‘modern foods’ that are lower in their salt and sugar content. Examples include salted meats such as ham, where chill cabinet temperatures are required to be set at a level appropriate to factory processed meats rather than those that are traditionally cured. Hence perishability is on occasion a legal rather then an innate concept. A major public health issue and emerging research need is to monitor the understanding of consumers of the storage needs of foodstuffs arising from their innate characteristics. At the extreme, traditional hams are stored in cool ambient rooms for months or years (and sometimes hung from the ceiling in shops and bars). Whereas packaged ham has a short shelf life even in the refrigerator.
Seasonality of food is innate and rooted in natural cycles of sowing and harvesting on land or sea. Religious festivals frequently celebrate these seasons, and emphasize how deeply grounded culture is in natural cycles. However, one major theme of contemporary food supply is that foods are sourced from around the world. Hence any food is available as it is in season somewhere in the world, or in a controlled factory farming environment. Food festivals celebrate those foods that are in season, and how foods are processed at times of glut to ensure that they are available in lean times. Hence the management of perishability through the traditional forms of preservation dealt with above (bottling, salting, etc.) is also typically celebrated. How far contemporary lives mean that customers perceive the underlying natural cycles is an important challenge for those of us who wish to see a return to ‘real’ and ‘local’ foods.
This dimension of seasonality is highlighted when we consider that the major religions of the world have festivals that coincide with natural rhythms of sowing and harvesting crops. As these religions are often from the northern hemisphere, they are anachronistic in the southern hemisphere, where seasons are reversed.
Celebration may be at its zenith when encountered as alcohol. Grains, fruits, and occasionally vegetables (such as potato spirits) are fermented when crops are available, and then they are stored and aged for celebration at other times of the year. Sampling such products is an agreeable and popular activity at many food and drink festivals. How purchasing behaviour at festivals is affected by such sampling is a legitimate research activity. Furthermore, the prevalence of sampling alcoholic drinks at festivals means that the role of the researcher is affected by an imbibulous sample. Clearly not all visitors drink alcohol, and many who do are prudent in their consumption. Yet the bonhomie engendered is both a facilitator for the success of the festival, and also a risk for researchers for whom some results may be biased.
‘Involvement’ is a characteristic that considers the extent which festival visitors are merely ‘surfing’ the event, and how far they feel an ideological commitment to any espoused values. Typical of these values usually relate to local, whole, rare, and traditional foods. Thus some festival visitors may share an almost religious commitment to the values claimed by some festival organizers, who seek to maintain and develop local food traditions. This may differ from visitors who have a hedonistic drive to visit. Of course many (maybe most) visitors go for a ‘good day out’.
Establishing psychographic market segmentation models is a very valid form of theory building research for festival researchers. There are several forms of approach to psycho-graphic segmentation which can be found in most texts on consumer behaviour, typical of which is VALs, standing for Values, Attitudes, and Lifestyles. Clearly knowing the age, gender, and other demographics, and also the socio-economic profile of festival visitors is useful. However, a more complex typology that describes archetypes may be more useful for festival organizers who wish to ensure that visitors’ needs are met and that exhibitors are properly briefed.
You may wish to consider several possible research ideas that can be explored in food and drink festivals. These are additional to ideas previously discussed.
Project ideas
• Describe the types of visitor that you would expect to find at a rural food and drink festival. Producing such a typology may help organisers’ to improve their advertising.
• Identify the psychographic market segments that a festival addresses using factors such as values, attitudes and lifestyles (Business Dictionary 2007).
• Have a look at multivariate analysis to see some advanced approaches to measuring responses.
So far we have identified possible research themes linked to what makes food festivals a differentiated research site. We next take this into a problem ownership context by considering the special needs of the various interest groups who will either sponsor or facilitate festival research.
A Research Agenda for food Festivals
The many research tools available to any social scientist may be relevant to those who work for food festival businesses. A lack of a large and coherent body of research into food festivals is a barrier to the development of the field, but is typical of new areas of work. However, the rapid growth of the field means that there is a large body of good practice that has been reported, and will be referred to in other chapters in this book.
The focus of this part of the chapter is to suggest ways in which the research base of food festivals can be expanded in a structured way. Normally, new fields develop over a protracted period of time as researchers publish their individual contributions.
What is proposed is that we should develop a tradition of publishing research findings that reflects the work undertaken over decades in medicine. In that field research often took the form of findings being published as case studies in which individual or small numbers of patients were studied. Such studies provided insights, but the purposes of research – model building or model testing – were hard to justify on the strength of such small numbers of observations.
A large body of such published case study findings does, however, provide an opportunity for statistical analysis. Hence researchers undertake what is called case s studies such as in Lyons (2005), where multiple case studies are analysed. For this to be relevant, the provenance of case studies must be clear, and the replicability of studies ensured by developing relevant formats in which data are clearly and precisely outlined. In contemporary medical research, we see the publication of the so-called meta-research studies, in which suites of case study research are aggregated and analysed, and results represent a large number of individual investigations.
The remainder of the article will outline some key research areas and strategies for creating case studies. Reading other chapters in this book will suggest other research topics, and the approach proposed here will be equally relevant.
A Problem Owner Approach
The festival organizer’s needs are a useful focus. Their needs revolve around a requirement to have knowledge of their visitors, exhibitors, competitors, and staff (particularly volunteers), alongside research needs associated with local planners and regulatory authorities. We will consider these groups as each may provide research opportunities.
Customers or visitors are most readily researched at the festival. Hence permission from the organizers is essential. What the organizers need to know is, who the visitors are demographically (primarily, age and gender plus the relationship of group members, e.g. a family), because this identifies needs for facilities and helps to inform exhibitors. Next socio-economics are important as they may affect marketing. However, our experience is that the foodie group seems to span a wide range of demographic and socio-economic groups.
Socio-economics are important if measured and evaluated carefully. A question about household income is always sensitive and if asked should be left to the end of any interview. However, care must be taken in evaluating such data. Thus a high salary for a new graduate living at home with parents may be rather poor salary for a family with children, and be moderate for a retired home owner with a small modern house (an old or large house may reduce the retired person to penury).
A surrogate to socio-economic questions is often useful. Thus a car park survey where the value of motor cars used by festival visitors can be helpful particularly if benchmarked against some other festival or supermarkets. This provides a methodology, where sensitive and unreliable questions about income can be avoided, and where objective data can be collected (despite limitations associated with, say, company cars, second cars, etc. be used to visit the festival). At its simplest a car park survey requires the collection of data on the make, model, year, and condition of cars. These data are then used in conjunction with a used car price guide to value vehicles. A comparison of value with other retail activities is a useful method to establish the economic place of visitors. For a more sophisticated analysis it is possible to look at cars as a reflection of VALs such as the prevalence of iconic vehicles such as Volvo’s, or life style choices such as 4WD’s, and people carriers. Prestige marques may be another indicator as may utilitarian vehicle choices.
Where food and drink festival visitors come from is of great importance as it has implications for accommodation needs of visitors, and for advertising to new or existing geographical areas for visitors. The UK postcode is our preferred method of collecting homes as it links to publicly available databases that also provide socio- and demographic data (e.g. from http://www.upmystreet.com). Similar data exist in other countries, although Michels (2007) denies the value of such data in a company-focused article based on Canada.
New forms of psychographic segmentation are important to festival visitors. A major longitudinal study that we have been involved in is the Ludlow Food and Drink Festival. We have identified that some visitors to the Ludlow Food Festival are what we call ‘foodies’, whereas others visit because they want a good day out. We have encountered church groups, singles holidays, group reunions, and pub group outings alongside foodies who cannot afford hotels or B & B’s, so they camp to afford the goodies, and others who visit both the festival and also local Michelin starred restaurants whilst staying at expensive hotels.
Our own experience of festival on-site research is that conventional market research interviewers (and students) can collect short structured questionnaires, as visitors are generally relaxed and communicative. However, for less structured research and probing, there is a need for mature and experienced researchers. We have collected many hundreds of interviews as mature experienced researchers who are prepared to undertake fieldwork, and our commitment and knowledge to the field mean that the data that we have collected is very rich. Similarly, we have undertaken many focus groups in which we have explored questionnaire results more fully, which again require experienced interviewers or facilitators.
Supplier or exhibitor research is a little different from customer research because it carries a commercial dimension. Many traders regard their replies as having implications for exhibitor charging, and the allocation of pitches. We have found that on-site research is helpful for probing immediate issues, such as when a new closing time is being piloted. However, exhibitor research is challenging and probably best approached personally by phone or pre-arranged visit at some time other than at the festival. The ethical dimension of researching traders is a little difficult as an outside researcher cannot commit the organizers to not changing their pricing policy of stands based on research data. In fact, the only commitment that the external researcher can be sure of delivering is anonymity. For university researchers, the Research Ethics Code of their university is clearly tested by such studies, as is the professional code of commercial market researchers.
Some festivals depend on volunteers. Where this is the case the individual volunteers may be more crucial to the organizers than individual customers because the pool of prospective volunteers is more limited than festival visitors. Knowing as much as possible about volunteers and their satisfaction with the festival is crucial. The typical volunteer in our study in Ludlow is retired, and how new retirees can be attracted is important as is the recruitment of new people who will serve on committees and in leading the festival.
So far this analysis has been focused on the festival organizer as problem owner. However, the analysis above indicates many of the stakeholders in a festival, to whom can be added the local community (possibly local or regional government authorities) and others such as academics.
At a more abstract level, it is useful to consider the what’s and why’s of research methods and this is done next.
Food and Drink Festivals Research: What’s and Why’s
Broadly, there are two styles of approach to researching food and drink festivals. Either we may pose a research question or define a soluble problem. Each of these approaches will be discussed next.
A research question typically contains a word such as: who, what, where, when, why, how, followed by a ‘?’. Often we have a high-level question such as ‘what brings visitors to our festival?’ which generates a series of more detailed researchable questions which may include:
This is purely an example of the sorts of questions that we may wish to answer. What is also important is that these are reasonable questions to pose and that we have suitable tools to collect data. A common mistake is made where complex motivational issues are explored based on issues that simply occur to the researchers. Good research is grounded in data. Hence attitudinal questions should be grounded in language used by visitors, possibly captured in a focus group and then further explored in a survey.
The research question approach is typical of academic researchers, although it is found in all sorts of studies which include those of commercial market researchers. A more typical consulting approach is however to pose the issue as a problem, such as ‘how can we increase visitor numbers by 50 per cent?’
In practical terms it may be sensible to scope a relevant aspect of a festival using a research question approach in 1 year, and to subsequently develop a problem-solving approach for a further year. This is because the problem solving approach is preceded by a data gathering phase.
If we have answered the research question outlined above, then we can start to identify potential markets that could be tapped into using appropriate advertising promotions. Thus it may be that a geographical region has good transport access to a food festival, but is not in the same administrative region and is therefore not subject to local information about events. It could be that a significant part in the increase of visitors could be generated simply by promoting the festival in the area. In the case of the Ludlow, although located in Shropshire, it is close to the borders of the English counties of Hereford and Worcester, Staffordshire, West Midlands, and to the Welsh county of Powys. As much tourism promotion is developed at the level of administrative counties this means that the Ludlow Festival may miss out on promotion in neighbouring counties. This may be through it being overlooked, but may also be a result of rivalry between destination events and it may overlook the possibilities of lost overnight stays for visitors.
Clearly, the example is overly simplified, but it does make the point that problem-solving is partly a function of scoping and collecting data that is informative for decision-makers.
Qualitative and quantitative data are distinguished in books on research methods. Qualitative data provide richness, but are frequently based on small numbers of respondents collected by interviews, focus groups, and surveys. In contrast quantitative data are perceived by the uninitiated as being statistical and therefore credible. This perception of credibility can be misplaced based on the principle of GIGO – Garbage In Garbage Out. However, complex the statistical analysis, results are only as good as input data quality permits. The abuse of statistical tests when misapplied to inappropriate data is widespread; in fact, many seminal research articles in the social sciences use the wrong statistics. Although good, modern journals no longer publish such work, as a journal reviewer, I am frequently surprised to see how often investigators working today still perpetuate mistakes (such as applying parametric data analysis to non-parametric Likert data). For a fuller explanation see the discussion at the end of this chapter.
Kanji (1999) prefaced a current edition of his textbook with the comment that when he began working as a professional statistician, there were only 12 tests that were required. Now his book focuses on the 100 most important tests, and he invites readers to discuss refining this list. The moral of this is that a prudent quantitative researcher will involve a professional statistician at the design stage of a quantitative survey, and keep them involved throughout. For example, even sample size is a difficult trade-off between accuracy and cost. Accuracy is itself a matter of the consistency of responses. Thus a small survey where people largely agree may well be statistically significant, whereas a large sample with a wide variation of responses may be less statistically significant.
So far the chapter outlines some key issues in food and drink festival research design. In addition, there are several examples of research that people may pursue. Next we have a short overview of some key research studies in the field, some of which are cited elsewhere in this book. This will be followed by the brief annotated reading list of research methods books.
Research Methods used in food and Drink Festivals
The implications of the chapter focus on social sciences research methods, which include both qualitative and quantitative methods including individual and group interviews, focus groups and similar, as well as quantitative surveys. Attitudes and behaviours are both important to social scientists, and can be very different. For example, preferring sausage A to sausage B doesn’t mean that A will be purchased in preference to B. Price may be important, and so may appearance, or the anticipated preferences of other family members. Whereas attitudes are explored by asking questions or setting up experiments to test preferences, etc. behaviours can be explored by asking questions or by observing people. Watching how people shop, explore a festival site, and/or exhibitor’s displays can be often best explored by watching them rather than asking them. Sometimes people tell you what they think that you want to hear, or they may just misunderstand their own behaviours.
Some of the more useful observational techniques to consider include:
Throughout the other chapters of the book, the major source of detailed data has been obtained through the use of social science research. The presentation of these data in the book takes several forms, ranging from case studies and shorter scenarios through to statistical reportage.
The research methods used to generate economic impact studies is typically generated through social science research methods, but also applies economic models that apply appropriate methods of factoring multipliers into leverage up the impact of direct expenditure on a food festival within a local economy. What is most important about this approach is that data generated by social science research methods are used as inputs to other processes and do not themselves answer questions such as ‘what is the economic value of a food festival to the host community?’
However, not all research on food festivals reported in this book or undertaken elsewhere fits the social science research mould. Some work is concerned with exploring pre-existing documents which are increasingly electronic, but maybe paper based. It is easy for social scientists to define such work as secondary research as it draws in various ways on the endeavours of others. However, much is actually primary research in so far as it provides new understanding and uses research methods more typically previously used in the humanities, rather than the social sciences.
Exploring archives and records that maybe contemporary with an event or written at a later date (such as the New Testament Apostles) is all relevant work in the humanities. However, it shares much with those using contemporary or on-line sources of data. In particular, establishing the authenticity of a record is of absolute importance. Most readers are aware of the computer scams which involve tricksters setting up web sites to look like legitimate banks. However, this extends beyond financial fraudsters to critics of organizations or ideas and to simple tricksters. Meticulous study of records to establish their provenance is always critical. How this may be done varies between subjects and may include some science as in carbon dating artefacts, some statistics such as in seeing if the style of writing of one author is matched in another document such as a play, through to expert testimony from experts who have long worked in a field.
At its simplest writing a case study on a festival based on data published on line or existing in paper or electronic archives is actually primary research and the case study is the output of the research. We must avoid the conclusion of some social scientists who may seek to diminish such publishable output as secondary research, as their view of research is discipline based and fails to appreciate the research tradition of the humanities. Other studies using similar datasets might simply quantify the frequency of occurrence of key terms associated with a festival in various databanks, and report the result as a statistic.
Down to Detail
So far we have looked at the broad issues associated with researching food and drink festivals. Now we will consider the reality of what the reader needs to appreciate.
Firstly, there is one very important distinction to be made between the prospective researcher, who is seeking to pass an assessment on an academic course of study, and a person seeking to change and improve a festival.
For the former, the ‘truth’ of their findings is of passing interest. What is important is to pass the various assessment criteria used on their course. Thus reliance on statistics reported in the trade press is frowned upon in an academic project, whereas as statistics collected by trade bodies and government agencies is approved of, as are data collected by other academics and published in properly blind refereed academic journals. In reality, the officially collected and published data may be of dubious validity, as may the output of the academics. However, the academic researcher may rely on them, whereas a person investing money would be well advised to be more circumspect.
The use of web sites to collect data highlights the problem. Firstly, it can be difficult to even establish the provenance of a web site, and if it genuinely represents the alleged owner – witness many banking scams where a credible web site is created to look just like a proper bank.
Secondly, data published by an interested party is always likely to be biased. Sometimes this means presenting data truthfully, but in a biased form, but it may also include simply lying. Checking up on web sites or other published documents is always important and rarely simple. Even historians often make the point that history is written by the winners, so how can we really know what happened?
Moving on to a practical social science issue, students almost always ask how big their sample should be. Some supervisors provide an answer, a sample size such as (say) 100. Others give the more correct answer that ‘it all depends’. Looking at this more closely, the issue unpacks as follows: firstly there is a question as to the statistical significance being sought in an answer. The larger the sample the more significant the answer is likely to be. However, there is a further issue to consider, which is the level of agreement or otherwise between respondents. So if everyone gives the same answers, a very small sample will be statistically highly significant, whereas if answers differ greatly a large sample will be needed. For a student it is often only necessary to prove that you appreciate the correct use of statistics when studying a festival, whereas a person investing money should need greater certainty.
We must not lose sight of how data will be used. So if we want to know if visitors differ over the various days on which a festival happens the question becomes one of defining the appropriate sample size for each day. In the consumer work that I have done with Liz Sharples at the Ludlow Food and Drink Festival, we have normally had samples typically of 120 groups of people on each day surveyed, representing around 1,000 individual visitors for the three days of the festival (as each respondent was typically representing the views of their party of three or so members). However, we had much smaller samples of traders and volunteers when these groups were interviewed, as there are many fewer people in total in each group.
Statistical analysis of data has been mentioned consistently in this chapter. Using an appropriate statistical analysis is of absolute importance. For example, simply saying that 45 per cent of people preferred sausage A (as above) and 55 per cent preferred B does not necessarily mean that B is preferred to A. The percentages are a statistic and not a number, so what is important is if the result is statistically significant? I would not spend my own money on the percentage statistic without the significance test being properly applied. As a general principle always seek expert advice on collecting and then using statistics, and that means talking to a practicing statistician and not relying on someone who passed an exam in statistics or research methods 10 or 20 years ago!
Selecting Books to Support your Research
I had initially proposed to review a selection of good texts for readers to use if they wish to undertake research into food and drink festivals. However, I have been doing some research into the subject which has alerted me to the variety of users that I may be addressing, and the wide range of contexts in which you may be working.
For those in large universities working in research you will, like me, enjoy a breathtaking range of formal and informal support in the form of computing packages, and more specifically expert colleagues who usually generously share their time and experience.
Other readers will be entry level students who are being assessed on their ability to design work unaided. Whilst some readers will be festival organizers or volunteers working without free access to a resource rich university environment.
I have therefore decided to produce a checklist for readers to check the suitability of books for themselves. There are two over-riding questions to answer in selecting a book:
You will have to decide your own answer to the first question, but assistance can be provided with the second. In that context remember that research like everything else is developing rapidly and ideas are changing.
Is a Book up to Date and Technically Competent: some Tests
You do not need to understand the meaning of these detailed points to apply these tests, any more than you need to know how a car works to drive it. However, reading the research methods book should enlighten you.
Test 1
Firstly look to see if the book is clear on there no longer being a competition between qualitative and quantitative research in the social sciences research community. Excellent projects typically start with a qualitative study such as focus groups to define what to survey, continue with a quantitative study such as a survey, and finish with further qualitative study such as in-depth interviews to probe the meaning of survey results.
Test 2
Check the clarity of the book when describing the nature and use of quantitative data. The different types of data that you can collect need different forms of statistical analysis. A book should define these data in a simple and understandable way. Here is the contents of a teaching slide that I use for this; see if the book is clear on the points.
Types of quantitative data
– named categories, e.g. favourite TV channel
– ranked categories, e.g. social class
– score categories, e.g. IQ, temperature with no real zero
– similar to interval, but with a real zero e.g. length
Test 3
Check up that the book is correct in its recommendations on the nature of data. If the book mentions a form of question using something called a ‘Likert scale’, check up that the analysis of the results include the calculation of the median or mode as acceptable ‘averages’, not the mean. Likert data are ordinal (as in test 2 above) but often inappropriately treated by some authors as interval.
Test 4
If the book suggests the use of a computer in analysing quantitative data, for the non-expert user, it is probably important that you are advised that a modern spreadsheet program like Microsoft Excel can be used. If the use of expensive specialist programs like SPSS is required then the book will not be of great use to the inexpert or person working outside of a research organization for a food festival.
References
Business Dictionary (2007) at http://www.businessdictionary.com/definition/Values-Attitudes-and-Lifestyles-SystemVALS-2.html accessed 26.11.2007.
Emery, F.E. and Trist, E.L. (1965). The causal texture of organizational environments. Human Relations.
Kanji, G. (1999). 100 Statistical Tests. Sage Publications, London.
Lyons, H. (2005). Food industry case studies: A suitable medium for publication. British Food Journal, 107(9), 702–713.
Michels, C. at http://extranet.mapinfo.com/common/
library/geodemographic_fundamentals_august_2003.pdf, accessed Wednesday, 07.03.2007.
The Times (2007). Bee, P. Talking Dirty, August 28
Zeithaml, V. A. and Bitner, M. J. (2002). Services Marketing, 3rd edition. McGraw-Hill, New York.