Methods Investigating Food-Related Behaviour

10
Methods Investigating Food-Related Behaviour


Moira Dean,1 Monique M Raats2 and Liisa Lähteenmäki3


1 Queen’s University Belfast


2 University of Surrey


3 Aarhus University


10.1 Introduction


In nutrition, the main aim is to measure what and how much people eat and to translate this into information on energy and nutrient intake, which can be used to assess the nutritional quality of the diet. To understand the determinants of food choices and amount eaten, we need to find answers to why people choose what they choose, where they eat, with whom and when. These food-related behaviours are complicated actions governed by a mix of cultural conventions, social interactions, individual perceptions and psychological influences. Food and nutrient systems can be conceptualised as interlinked systems of producer, consumer and nutrition. Further food choices are inter-related and foods are not selected in isolation. For example, when a food is added to a diet it may replace another; omitting or adding a food that is usually eaten in combination with another food may also result in omitting or adding that other food as well. Thus, when dealing with food-related consumer behaviour, there is a need to address not only eating behaviour but purchasing and preparation as well (Figure 10.1).

c10-fig-0001

Figure 10.1 The relationship between food and health.


As food choices and other diet-related behaviours are determined by cultural, social, individual and food-related factors, there are different models to explain these behaviours depending on the viewpoint taken. The choice of approach is dependent on the research questions in which we are interested, but typically all these approaches use methods that have a basis in social sciences. The methods help us understand why people choose what they eat; that is, the reasons for the behaviours over and above the intake methods typically used in nutrition. These methods should be able to elicit the types of explanations given for food choice at the individual, societal and/or cultural levels (Table 10.1). In all cases the reasons for food-choice behaviour are usually elicited at the individual level and the impact of societal and cultural reflections on the individual are subsequently discussed.


Table 10.1 Levels of influence on food-choice behaviour: the Social Ecological Framework.






















Level of influences Examples of influences
Intrapersonal level An individual’s knowledge, skills, attitudes, values, preferences, emotions, values, behaviour
Interpersonal level An individual’s social networks, social supports, families, peers and neighbours
Community level Community resources, neighbourhood organisations, social and health services
Organisational level Businesses, public agencies, churches, service organisations
Public policy level Legislation, policies, taxes, regulatory agencies, health system, social care system, political/geographical environment

The methods introduced here should support nutrition studies by providing tools to measure the reasons behind food choices and intake more effectively. Due to limited space, the methods are described as types of approaches. The aim is to cover the general principles in applying the specific methodologies and to demonstrate what information different methods can give us about the factors influencing behaviour, but active application of the chosen method will require further reading and more in-depth exploration.


When interventions are used to change an individual’s food-related behaviours, the instruments and scales used in this research should be able to measure the changes in behaviour. Therefore, measurements employed to measure behaviour not only have to be good enough to measure food-choice behaviours, but also sensitive enough to pick up behaviour change as well.


10.2 Types of data in food-choice behaviour


There are three main types of research approaches: qualitative, quantitative and mixed methods. Criteria for selecting a research design for a study will depend on the research problem, the personal experience of the researcher and the researcher’s audience. Here qualitative and quantitative designs should not be viewed as polar opposites, but as different ends of a continuum with mixed methods in the middle, as it incorporates elements of both qualitative and quantitative approaches. These approaches differ in:



  • the basic philosophical assumptions the researchers bring to the research;
  • the types of research strategies used (e.g. qualitative case studies or quantitative experiments); and
  • the specific methods employed in conducting these strategies (e.g. collecting data quantitatively on instruments versus collecting data qualitatively through observing a setting).

Qualitative approaches employ emerging methods with open-ended questions and may contain interview data, observation data, document data and audio-visual data that undergo text and image analysis with themes and pattern interpretation. In contrast, quantitative approaches aim to test objective theories by examining the relationships among variables using pre-determined, instrument-based questions that contain performance data, attitude data, observational data and census data. These are statistically analysed and interpreted. A mixed-methods approach uses both pre-determined and emerging methods with open and closed questions to gather multiple forms of data, drawing from all possibilities. Statistical and text analysis are conducted with across-database interpretation.


In a study design the researcher not only selects a qualitative, quantitative or mixed-methods approach, but also has to decide on a specific strategy of inquiry within each of the approaches.


10.3 Qualitative approaches


When to choose a qualitative approach


Qualitative research explores and understands the meaning that individuals or groups ascribe to a social or human problem. The process of research involves emerging questions, procedures and data collected in the participant’s setting. Data are analysed inductively, building from particular to general themes, and the researcher interprets the meaning of the data. Researchers support an inductive style, focusing on individual meaning and the importance of interpreting the complexity of a situation. Qualitative research is thus well placed to answer complex questions about food-related behaviour, as it can investigate how and why individuals act in certain ways. Qualitative approaches are typically applied in studying how consumers relate to novel types of foods, such as functional foods, or to gain a better understanding of cultural practices around food provisioning. Qualitative strategies include:



  • Ethnography, where the researcher studies an intact cultural group in a natural setting over a prolonged period of time by collecting, primarily, observational and interview data.
  • Grounded theory, where the researcher derives a general, abstract theory of a process, action or interaction grounded in the views of the participants. This involves using multiple stages of data collection and the refinement and inter-relationship of categories of information.
  • Case studies, where a programme, event, activity, process or one or more individuals are investigated in depth. Case studies are bounded by time or activity and researchers collect detailed information using a variety of data-collection procedures over a sustained period of time.
  • Phenomenological research, where the researcher identifies the essence of human experience about a phenomenon described by the participants, by studying a small number of subjects through extensive and prolonged engagement, to develop patterns and relationships of meaning.
  • Narrative research, where the researcher studies the lives of individuals by asking one or more people to provide stories about their lives. These stories are then re-told and re-storied by the researcher onto a narrative chronology.

Sampling and recruitment in qualitative research


Qualitative research uses some type of purposive sampling to recruit a relatively small number of participants (Table 10.2). These people are deliberately selected to address the research aim, as they are considered to be a rich source of data in relation to that aim. They could be selected on personal characteristics such as gender, socio-economic status, nutritional status, health status or their experience of a specific event (e.g. regular shoppers), their behaviour (e.g. organic shoppers) or their attitudes and beliefs (for or against genetic modification). While there are different types of purposive sampling, theoretical sampling is commonly used, originating from grounded theory where sampling continues until no new data or properties relating to the construct(s) of interest emerge, a point known as theoretical saturation. Convenience sampling (where a group is easily accessible, but not biased) and snowball sampling (asking recruits to suggest new participants from among their friends and acquaintances) are also employed in qualitative research; these may be the only practical options when working with vulnerable and hard-to-reach populations.


Table 10.2 Sampling methods used in food-choice research.






































Sampling approach Sample selection strategy
Probability sampling methods where random selection is used to ensure that all members of the group of interest have an equal chance of being selected to participate in the study
Simple random sampling Every member of the population being studied has an equal chance of being selected. All of the population needs to be available.
Systematic selection (interval sampling) Used when a stream of representative people are available, e.g. shoppers in a particular store.
Stratified sampling (proportional and disproportional) The population is divided into non-overlapping groups (‘strata’) and samples are taken from within these groups.
Clustered sampling Used when the population of interest is large and widely geographically dispersed. Clusters within the population are randomly selected, e.g. cities.
Purposive (non-probability) sampling, i.e. sampling with a purpose in mind, usually an interest in particular groups
Convenience sampling Participants will be those to whom the researcher has relatively ‘easy’ access, e.g. students.
Snowball sampling Participants meeting the study requirements will recommend others with the same characteristics, e.g. members of a club. This method is used when trying to access difficult-to-reach populations.
Quota sampling Participants are non-randomly selected according to a pre-defined fixed quota. Proportional quota sampling aims to match the proportions of a particular characteristic found in the population as a whole; in no-proportion quota sampling one is less restrictive about matching the population-level figures.
Typical case sampling Participants are selected to be typical, normal or average for a particular phenomenon.
Theoretical sampling Participants are selected on the basis of the results of the data collected to date. The goal is develop a deeper understanding of the topic and to develop theory.

The adequacy of sample size is determined by the breadth of the phenomenon under study: the broader the phenomenon, the larger the number. Rough indicators of what might be adequate could be obtained from similar published studies in high-quality journals. It has been shown that theoretical saturation could be achieved with six interviews.


Data collection and analysis of qualitative methods


Individual interviews


Interviews are a commonly used technique for data collection in qualitative research and can be described as a ‘construction site of knowledge where two individuals discuss a theme of mutual interest’. Interviews can be categorised into four types: the informal, conversational interview; the guided interview or topic approach; the standard open-ended interview; and the co-constructed dialogue interview. The four types vary in the degree to which they are structured, with one end of the continuum being unstructured, in-depth interviews where the researcher introduces a topic and then gives the interviewee the freedom to talk, with only occasional prompts and probes. This type of interview allows participants’ experiences, meanings, values and priorities to emerge with minimal interference on the part of the researcher, but it may be less useful for answering specific research questions. At the other end of the continuum, structured interviews comprising pre-defined questions are administered in a pre-set order.


Semi-structured interviews seek to obtain descriptions of the life of the interviewee with respect to interpreting the meaning of the phenomenon being described. They consist of a sequence of themes to be covered, as well as some suggested questions, with an understanding that changes to the sequence and form of questions are possible in order to follow up the specific answers given and the stories told by the participants.


Preparation for the interview stage usually involves a script called an interview guide, which structures the course of the interview more or less tightly. In a semi-structured interview the guide will include an outline of topics to be covered with suggested questions. Depending on the particular study, the sequence of the questions will be pre-determined and binding, or it can depend on the judgement and tact of the interviewer, who can follow up the interviewee’s answers and any new directions that open up.


The interview questions must be brief and simple. Different types of questions may be used. To allow for an easy start and to build rapport, introductory questions such as ‘Can you tell me about…?’ or ‘Could you describe in detail…?’ can be used, followed by questions to which the participants’ answers are extended through a curious, persistent and critical attitude on the part of the interviewer. This can be done through direct questioning of what has just been said or by a mere nod or ‘mm’. The interviewer can use probing questions such as ‘Could you say something more about…?’ or ‘Can you give a more detailed description?’ to pursue the answers without stating what dimensions are to be taken into account. The interviewer may also follow up with more operationalising questions, such as ‘What did you actually do?’ or ‘How did you react?’ The interviewer can introduce topics directly, preferably in the later part of the interview. Furthermore, interpretive questions can take the form of rephrasing the question or clarification.


Depending on the participants, some interviews need more careful consideration. For example, when interviewing people from a different culture, specific factors such as asking questions as a means of obtaining information, replying directly, referring to taboo matters, making eye contact when speaking or sending a man to interview a woman and vice versa need to be thought through in detail. Difficulties in recognising disparities in language use, gestures and cultural norms may also arise within the researcher’s own culture when interviewing across gender, generation, social class or religion. Although differences between subcultures may not be as pronounced as those between different cultures, if the researcher makes the implicit assumption that everyone belongs to a common culture, intercultural variations may be harder to detect.


When interviewing children, it is important to use age-appropriate questions and to be aware of several difficulties of interviews with adults that may be aggravated, such as the interviewer asking long and complex questions or posing more than one question at a time. In some instances, it may help to interview children while they are carrying out some other task such as drawing, reading a story, or watching TV or a video. See Box 10.1 for an example of a study using interviews.


Group discussions


A commonly used form of group discussion is a focus group, which is defined as a carefully planned discussion designed to obtain perceptions on a defined area of interest in a permissive, non-threatening environment. The focus group presents a natural environment in which participants are influencing and influenced by others – just as they are in real life. The researcher serves several functions in the focus group: moderating, listening, observing and eventually analysing using an inductive process.


The topics of discussion in a focus group are carefully pre-determined and sequenced based on analysis of the situation. This analysis includes an in-depth study of the event, experience or topic in order to describe the context, ingredients and components of the experience. The questions are placed in an understandable and logical order.


Focus groups produce qualitative data that provide insights into participants’ attitudes, perceptions and opinions. These results are solicited through open-ended questions, to which respondents are able to choose the manner in which they respond, and from observations of those respondents in a group discussion. The discussions are audio-taped and transcribed. Careful and systematic analysis of the discussions provides clues and insights into how a product, service or opportunity is perceived. The researcher derives understanding based on the discussion as opposed to testing or confirming a pre-conceived hypothesis or theory. See Box 10.2 for an overview of the strengths and weaknesses of group discussions and Box 10.3 for an example of a study using group discussions.


Laddering


Laddering refers to an in-depth, one-to-one interviewing technique used to develop understanding of how consumers translate attributes of products into meaningful associations, following means–end theory. Salient attributes of product choices are often elicited by asking why one product is chosen over another, followed by asking why these attributes are important in the choice. The second ‘why’ question elicits what consequences the attributes have for the person; further ‘why’ questions about the importance of consequences lead to values. With the laddering technique we can find out what product-related attributes signify to a person. Its advantage over other qualitative approaches is that the meanings in the means–ends chain are personally relevant. Therefore, laddering can provide results that are more closely related to preference and choice behaviour.


As described above, laddering uses a tailored interview format employing a series of directed probes, typified by the ‘Why is that important to you?’ question, with the aim of determining connections between attributes (A), their consequences (C) and values (V). These association networks, or ladders, represent combinations of elements that distinguish between the products in a given product class. The higher-order knowledge structures represented by the A-C-V ladders provide a perspective on how the product information is processed from what could be called a motivational perspective, in that the underlying reasons for an attribute or a consequence being important can be uncovered. If product-related attributes are linked to values via consequences, they are likely to be more central in making decisions over alternatives than those attributes that are not linked to values or even consequences: it is not about what the product is, but what the product can do for you. Laddering interviews have been used extensively in the food domain, although sometimes the level of elicitation has not reached the value level.


The analysis of laddering data involves coding the attributes, consequences and values by a standard content analysis procedure. The summary table constructed represents the number of connections between the elements. From this the dominant connections can be selected and graphically represented in a tree diagram, called a hierarchical value map. Interpretation of this type of qualitative, in-depth information permits an understanding of a consumer’s underlying motivations with respect to a given product class. Each unique pathway from attribute to value represents an opportunity to understand the values driving food-choice decisions. See Box 10.4 for an example of attributes, consequences and values generated in a laddering study and Box 10.5 for a further example of this kind of research.


Behavioural observations


As opposed to self-reported behaviour, observational techniques allow behaviour to be observed directly. The data obtained is very rich because it includes non-verbal and physical behaviour. Measures of what, when, where and how much is being selected, purchased, prepared, cooked, shared or eaten can be collected by observers by photographing or filming behaviour. Data can be collected in unobtrusive naturalistic or controlled settings, such as observation labs where participants shop, cook or eat in a controlled environment that uses cameras and/or sensors to record behaviour. See Box 10.6 for the strengths and weaknesses of behavioural observations and Box 10.7 for an example of a mixed-methods study that included behavioural observation.

Jun 13, 2016 | Posted by in NUTRITION | Comments Off on Methods Investigating Food-Related Behaviour

Full access? Get Clinical Tree

Get Clinical Tree app for offline access