Part II: Conceptualization

Chapter 3: Defining the nutrition problems (Phase 1 - Conceptualization)

Any nutritional intervention must begin with the clear identification of the nutrition problem which is to be addressed.

What is this problem? How is it manifested?

What population is affected?

What is its impact on the social, economic and cultural life of the population concerned?

Is it a priority public health problem?

The extent and the magnitude of a nutritional problem, groups affected, its socio-economic importance and prioritization as public health problems is determined by nutrition workers on the basis of assessment of nutritional status of the population.

The assessment of nutritional status usually involves a single or a combination of direct and/or indirect methods depending on the objective of the assessment.

Direct methods are:

Dietary survey


Clinical signs


Laboratory tests


Biophysical methods


Anthropometric measurements

Indirect methods are:

Assessment of health statistics


age specific mortality rates cause specific mortality rates nutritionally relevant infections


Assessment of the ecological factors socio-economic factors sanitation and water supply coverage of health services geography and climate cultural factors

People dealing with the design and the implementation of the nutrition communication intervention are generally not required to be involved in this exercise and, therefore, not discussed in this book. Only an outline is provided. Interested readers will find detailed discussion in (6).

To recall the principal nutritional problems in developing countries: these are protein-energy deficiency, disorders linked to iodine deficiency, vitamin A deficiency and iron deficiency anaemia. A growing number of countries are confronted with new health risks linked to diet. These were until recently the diseases of affluence, namely cardiovascular diseases, diabetes, obesity and cancers (7). Evidently, the causes of these different nutrition-related problems are not identical in all countries. Consequently, the possible solutions vary especially in that of giving advice to the population. Beyond the universal messages ("Eat, balanced diets! Exercise"), there are characteristics unique to each theme of nutrition education. It is therefore important to define very precisely the nutritional problem at which the intervention is to be directed.

The next step involves an analyses of the causes of the nutritional problem, which should be carried out in an intersectional group setting.


Chapter 4: The causes of the nutritional problems

4.1 How is causal analysis carried out?

Since some time, it is known that causal analysis is absolutely necessary for the successful design and implementation of nutrition education. Why is this so? It is because nutritional problems are the results of an interaction between complex and multiple socioeconomic, biological and environmental factors.

In all nutritional interventions, educational or otherwise, the promoters must first determine the causes of the problem which have been identified.

Beghin et al. (4) have redefined a procedure for the analysis of the factors influencing nutritional status (relative to the specific problem identified). The product of this analysis is a network revealing the most direct and even the indirect factors that affects nutritional status of the population.

How is causal analysis carried out?

The steps in the procedure are as follows:

An example of causal analysis:

In 1990, the Department for prevention of malnutrition in Ministry of Public Health in Morocco, carried out a causal analysis prior to the implementation of a nutrition communication program (6).

This analysis was carried out to determine the intervention which the Department ought to undertake for promotion of breast-feeding and growth monitoring. A three day seminar gathered about twenty participants. The composition of the group was intersectoral although the Ministry of Health was better represented than the other departments.

In the first phase of the analysis, the participants agreed to consider factors which directly influence the nutritional status of children from 0 to 5 years; the food consumption of the child and the biological utilization of foods. Following this decision, the participants proceeded to an analysis of these two factors at the first level.

Figure I shows the branches which influence food consumption.

The group then chose to act on the breast-feeding which intervenes on both branches of the tree. The result of causal analysis on initiation and duration of breastfeeding are given in Figures 2 and 3.

Figure 1: Causal analysis of nutritional status

To facilitate group-work, a few rules for construction of the causal model are given as follows:

4.2 How to identify the behaviour in question?

The stage which is of most interest to communicators is the identification of the behaviour in question. Green et al. (5) call this stage "behavioural diagnosis".

All group members assembled to carry out the causal analysis must pay particular attention to the causes linked to behaviour. Only those factors will be amenable to an educational approach.

Whatever the type of intervention decided on, it is unlikely that the group would have included an educational component. This is because human behaviour is often implied in nutrition, whether the behaviour is linked to production, conservation, food consumption or even individual health. The communication thus serves as a basic strategy for integrated programmes (e.g., Joint Program for Applied Nutrition in Iringa, Tanzania).

In certain cases, however, only behavioural factors will be the subject of the nutrition intervention. Certain projects carried out in accordance with the principles of social marketing, as in the Indonesian Programme for Improvement of Nutrition (8), have shown that it is possible to influence favourably the nutritional status of children via educational programs directed at their parents.

In all cases, the group responsible for causal analysis must try to identify the behaviour linked to the factors which form the object of the intervention. In the work carried out in Morocco, the practice of breast-feeding was the object of a complementary analysis carried out by the same working group. As we will see below, this second phase of the analysis allows for the identification of factors which influence breast-feedding in Morocco.

How to identify the behaviour in question?

Breast-feeding consists of two elements: initiation of breast-feeding (indicators: whether breast-feeding has been initiated or not and the time interval between birth and initiation of breast-feeding) and duration of breast-feeding (indicators: duration of exclusive breast-feeding, duration of supplemented breast-feeding).

An example is provided in Figures 2 and 3

Figure 2: Causal analysis of breastfeeding different causes that influence initiation

Figure 3: Causal analysis of breastfeeding different causes that influence duration

Chapter 5: Establishing the educational framework

5.1 What are the issues?

Factors determining behaviour

Once the various behaviours to be changed have been revealed, it is necessary to analyse a little deeper in order to identify the determining factors. External factors may be economic (e.g. revenue), social (e.g. legislation), climatic (e.g. rainy season) or geographic (e. g. the nature of the soil).

Internal factors have been analysed by different authors. According to Leclercq (9) human behaviour is determined by five factors: motivation, knowledge, self confidence, decision and skill (knowledge of what to do and how to do).

Leclercq's (9) model does not really take external factors into account. It is evident that these external factors do influence behaviour patterns.

In nutrition, external factors are of primary importance: food availability, access to health services, and family income are only examples which may prevent or encourage adoption of a behaviour.

Among the factors influencing a course of action, those which depend on the people concerned and those which are, unfortunately, out of their control. must be underlined. The latter may be such that any attempts to alter a course of action without taking them into consideration would certainly fail (example: In the case of a famine due to drought, it would be rather difficult for the people to maintain a balanced diet since the food required would not be available).

In any event, it is crucial to understand why people behave the way they do. In general, people have valid reasons (in their opinion) for behaving in a certain way. It is important to discover what these reasons are before trying to make any changes in their habits.

Channels of Communication

The second issue to be addressed in an educational diagnosis concerns the channels of communication. In nutrition, communication occurs through a multitude of channels, be they interpersonal or mass-media based.

The questions to be raised are as follows:

In the community concerned (which may be very large - the entire country, perhaps or very limited a village) what are the networks of communication? What are the relays, the influential people and institutions? There will be general answers to these questions but those specific to the field of nutrition are particularly important. Various people or institutions may, in fact, play a greater role in communication for nutrition than in other areas (e.g. a mother-in-law often has great influence on infant feeding, but less on food purchase).

5.2 How to conduct an educational diagnosis

There are several methods that can be used to obtain a rapid appreciation of a existing situation. Among those available are rapid participatory rural appraisal or the Rapid Assessment Procedure and the Rapid Ethnographic Assessment. It is a matter of selecting the method most appropriate to the situation being dealt with.

The methods proposed may be used in a variety of situations regardless of the programme. When used together as part of an overall assessment for an educational diagnosis for the population at risk, they can provide a very clear picture of the reality of life of the people.

5.3 Five methods for gathering useful data

How can the factors influencing nutritional habits be analysed? How can the active channels of communication in a given community be identified?

Technical files 1-5 provide a number of approaches which will enable the planning committee to obtain a realistic view of the population group under consideration.

In addition to the methods of data collection described above, important data can be collected during implementation of the intervention. The method involves integrating the opinions and suggestions gathered during group discussion in order of perceived importance. This method will be examined later in discussion of slides, videos and rural radio. The data collected then takes on an educative role because it is used to stimulate discussion on the subject.

It is possible and, indeed, strongly advisable, to establish two-way communication in any group discussion. It is even more crucial to try to understand the system of communication used in the target population even before studying nutritional habits.

TECHNICAL FILE N°1: A literature review

What is its purpose?

Issues related to communication in the community, individual health, production and food consumption have already been addressed in the literature. Before embarking on any new, costly studies, is advisable to review the literature or unpublished reports and accessible files.

How to do it?

First, it is necessary to define the subject area for review. The subject area must not be too broad (e.g. "The development of agriculture in the country" is a subject on which there is already a vast amount of material, much of which may be quite useless to the particular study on communication). The area of study should focus on three main points: the field of study, the social and/or the geographic context, and the period of interest (e.g. "Family production of food rich in Vitamin A in the region of Maradi, Niger since 1980").

How to do it?

Second, decide from where the information can be obtained. Libraries and documentation centres have a two-fold advantage in that the material is already classified and there is usually one, or sometimes even several, specialists who can assist in getting the required documentation.

The documentation centres in international organizations are a valuable resource that should not be forgotten. Although much of the material typically available has not been officially published, it is nevertheless a very useful source of information. Technical sections of institutes and ministries have often valuable information.

How should the research be carried out?

There are basically two ways of conducting the review. One method includes searching the bibliographic indexes for recent articles and studies on the particular subject and noting the appropriate sections. The other involves going through the library files in a systematic way (author, title, subject area, key words). This is easier if the system is computerized.

Some data bases are accessible through modem. They allow the user quick access to references which are not available in the country. Other documentation centres will, on request, provide information by the mail. Examples include Centre International de l'Enfance in France (Chateau de Longchamp, Bois de Boulogne (75016 Paris) and the documentation unit A.P.H.A. in the United States (APHA, 101515th Street NW, Washington D.C. 20005, USA)

How should the research be carried out?


TECHNICAL FILE N°2: Interviews in a central location

What purpose is served?

These improvised interviews permit rapid collection of information among a significant number of people gathered in a specific place for any reason other than that of the survey.

How to carry it out?

A location must first be selected where one is likely to meet members of the target group (e.g. patients in an hospital, mothers of young children at MCH Centres, housewives in the supermarket, etc).

A questionnaire is then formulated and pre-tested. It should not be too long because people would not have been notified in advance and thus do not like to spend much time in answering.

How to carry it out?

The interviewers should go to the place identified and carry out the investigation with people meeting the required profile. As there are usually many people present, a large number of persons can be contacted in a short space of time. This is, above all, the primary advantage of this method.

Some people will be willing to answer questions while others may refuse to do so. Therein lies a sample bias. In any case, the people interviewed in a central meeting place are not truly representative of the target population. This sample type is never representative whatever the number of people interviewed. However, this kind of survey provides some indication of the people who will be studied.

How to use the findings?

The findings of the survey may be presented in a table as follows.

Example 1: The description of the sample




e.g. women 20 to 40 years



  women 40 years and over






Example 2: The responses

There are differences between closed questions (the person interviewed must choose one of several responses proposed by the interviewer) and open questions (the interviewee responds freely). It is very easy to present responses to closed questions in a table since all that is required is a listing of the proposed responses and the frequency for each of these.

e.g. Did you buy avocados at the market today?










Answers to open questions are analyzed in a different fashion Here all the responses given by people interviewed must be taken into account

e.g. What foods did you buy at the market today?







e.g. Why do you eat avocados? (Several answers possible)

Because I like it



Because it's good for my health



Because it's cheap



Because it's easy to prepare





How to draw conclusions?

As the sample of people interviewed is not representative of the target audience, all generalisations must be avoided However, the result of the survey can enable formulation of hypotheses about certain behaviour and their causes These hypotheses can be tested by carrying out a KAP survey (Knowledge, Attitude and Practices,) using a sample which is representative of the population studied.


TECHNICAL FILE N°3: In-depth individual interviews


Certain persons, by virtue of their professional activities or their position in society, may possess a wealth of knowledge on a particular subject. In-depth interviews are held with one or more of these resource persons for about one to two hours.

How to conduct the interview?

Individual interviews must be conducted by well-trained persons and if possible by members of

the planning committee. This is to eliminate the risk of losing important information.

The interviewer must be capable of gaining the confidence and trust of the person being interviewed so that the latter will feel free to answer questions asked. The place where the interview is to be conducted should be quite comfortable and private and allow for frank continuous dialogue without disturbance.

These interviews may be recorded if the respondent agrees. In this way, the interviewer can concentrate on the interview rather than on note-taking.

How to draw up the interview guide

The interviewer uses an interview guide to keep the session focussed. The guide incorporates mainly those issues facing the planning committee and makes use of open ended questions. However, consideration should also be given to the respondent's competency in the subject.

Some examples:

What have been your experiences with programs or activities for preventing malnutrition in this region?

In your opinion, what are the main causes of malnutrition in children under five years old in this region (or whatever the specific problems are)?

Are there any habits or practices which you believe contribute to malnutrition in this region?

Can we change any of these? Why do you say so?

The information gathered is qualitative in nature. The opinions expressed by the people interviewed are generally subjective. As such, the committee will give credence to the findings depending on their assessment regarding the credibility of the particular resource person.

How to conduct the interview?

How to draw up the interview guide


TECHNICAL FILE N°4: Focus groups

Purpose served

Focus groups are used to obtain qualitative data on the opinions, beliefs, attitudes and values regarding a specific subject.

How to conduct the sessions?

Under the guidance of a well trained facilitator, groups of 8 to 12 persons are convened for discussion of a specific theme: food consumption, feeding of infants 6 to 12 months, etc..

For focus groups to be productive, a homogeneous group is needed. To determine what important characteristics have to be taken into account to ensure that the group is homogeneous, it is necessary to consider the particular subject to be discussed. Also, it is crucial that focus groups do not include persons with widely varying social status, educational levels, age or gender. Issues concerning maternity should be discussed in different groups of women: those who have children and those who do not have children. The same applies for issues concerning farmers: it is necessary to differentiate between small farmers and large farm proprietors.

How to conduct the sessions?

The groups should reflect the diversity of the society being studied but does not constitute a representative sample. As many focus groups are convened as can provide new information on the subject being studied.


In a study of the reluctance of women to use contraceptive methods in rural Rwanda, 8 focus groups were formed as indicated in the table below:


Using Contraceptives

Not Using Contraceptives

1. Literate women without children


2. Illiterate women without children


3. Literate women without children


4. Illiterate women without children


5. Literate women with children


6. Illiterate women with children


For criteria relevant to the use of contraceptives whether or not women have children and whether or not they are literate were considered. On this basis, eight types of focus groups were convened. All the possible combinations do not necessarily have be formed. If there are only a few illiterate women with no children, this group does not have to be formed. During the course of the investigation one may decide to form three groups of the same type in view of the liveliness of the discussion. Similarly a group not indicated in the table may be formed, for example a group of men, or a group of pregnant teenagers.

How to formulate the guide?

There are three steps involved. First, clearly define the aim of the investigation. Then, make a list of the subject areas to increase the likelihood of collecting the important data required. Finally, design the guide itemizing the subjects to be studied.

The guide should be pretested and modified, if needed, before using it in the focus group sessions. The interview is semi-directive and typically lasts for one to one and a half hours. The facilitator should be flexible adapting quickly to the flow of the discussion and altering the content or sequence of questions originally planned, if it becomes necessary. The facilitator must lead the group into in-depth discussion of the subject, listen carefully and try to determine reasons for the ideas expressed.

The facilitator is assisted by a person acting as a recorder. The recorder is concerned with taking note of ideas and opinions voiced and the manner in which they are expressed. The use of a tape-recorder will greatly facilitate the analyses of information collected.

How to analyse the information?

It must be recalled that the data collected is qualitative. A common error is the generalization of information in such statements as "the majority of persons think that..." or even "X percent of people think that...". The information collected from focus group sessions are only opinions, useful in understanding the behaviour of the population concerned or for designing effective messages addressing these behaviours. The data can certainly not be used to generalize on the characteristics of the population being studied.

In some cases, one opinion voiced by only one individual in the group becomes the most useful bit of information for the program. In other groups, the comments (whether verbal or non-verbal) on an opinion expressed by a group member may become the most pertinent.

The report following a focus group session need not necessarily be very long. It should help those concerned with communication to understand why people behave in a certain way and what are the underlying reasons for the behaviour. It should add a new dimension to the understanding of these phenomena.

TECHNICAL FILE N°5 - Observation techniques

It has been said that during the course of an interview conducted with a questionnaire, the interviewees give the responses which they believe the investigator expects to hear. This phenomenon which is related to "desired social prestige" is not a major one in in-depth individual interviews or focus groups. Nevertheless when people speak about their practices, beliefs and values they inevitably distort the truth a little so as to present themselves in a favourable light.

This limitation of interviews is partially overcome by observing people in their normal environments. This method claims greater objectivity in describing and understanding human behaviour.

Observation techniques


Two methods of observation must be clearly distinguished:

Observation through participation and systematic observation of certain practices. Observation through participation is used in anthropology. The technique was first presented by MALINOWSKI in 1922. Since then, there have been many followers. Among them are several noted anthropologists. It is impossible to describe in a few lines a technique which requires extensive theoretical and practical training. The investigator becomes a part of the life of the community members whose practices he intends to study. While living in the community he observes daily life and makes notes in private. He continuously interprets the findings of his observations which serve as a guide for future investigations.

The success of this anthropological approach is explained by its ethnographical aspect entailing a complete description of the community, its social systems, its beliefs and ways of life. To develop this profile anthropologists base their study not only on observation but also on in-depth interviews with 'informants' from the community being studied.

Systematic observation of practices is more of an ethnological and anthropological technique. It involves making a detailed report of specific behaviours. As an example: for oral rehydration detailed description were made comparing the different aggregate behaviours (36). Observation techniques are very objective; one has to count recurring behaviour, note the order in which it appears, the duration of each behaviour, etc.

In both techniques, the investigators must remember that their presence may affect the behaviour of the people being observed. However, by using some discretion, the investigators can be successful in obtaining a more or less realistic description of daily life.

The picture of reality revealed by these techniques is as true as can be expected, more so than with any interview conducted with a questionnaire.

How to analyse the results

Both methods can furnish information useful in the preparation of an educational intervention in nutrition. Observation through participation helps to place the practices in their cultural context and so obtain a better understanding of their underlying causes. It also enables the planners to better predict the effects of changing these practices. Systematic observation will help to provide a detailed explanation of behaviours. It will reveal the advantages and disadvantages of each behaviour. The analysis of even the most insignificant behaviour facilitates the definition of the appropriate message to encourage a specific behaviour.

5.4 KAP studies

A KAP survey seeks to determine the knowledge, (K) attitudes (A) and practices (P) of a population. It is based on a questionnaire used in a sample that is representative of the population being studied.

The KAP survey can also be used in the evaluation of a programme. In this case, a survey is conducted before and after the intervention.

In order to distinguish effects of the intervention, it is often necessary to carry out the same survey in a control group, that is, in a community not exposed to the intervention.

There have been several criticisms levelled at KAP studies:

Nevertheless, it is relevant to recall the essential elements of a KAP survey.

1. The questionnaire must be carefully formulated. The questionnaire is comprised of mainly closed questions to facilitate analysis of the data. It involves extensive preparation. Pertinent questions need to be identified. The findings of qualitative methods of investigation described above can prove useful here. Questions need to be formulated, translated and pretested before the final questionnaire is ready. The length of the questionnaire must be adapted to the time respondents can devote to the investigation. The questionnaire must be based on the objectives of the study. Questions unrelated to the objective of the study should be scrupulously avoided.

2. The sample must be representative of the population under study. To extrapolate conclusions to the entire population, the margin of error used in determining the sample must be small. The sample size must be sufficiently large and the sample must be random. Some statistical training is needed or advice should be obtained in order to observe these two conditions.

3. Rigorous organization is vital. Interviewers must be well trained and the interview standardized otherwise sources of interviewer bias will significantly affect the results of the study.

4. An efficient and rapid treatment of data collected is required. Computer based data management and analysis is preferred over manual treatment. The former is generally more efficient. Consider for example the months of hard work required to analyze 20,000 questions (40 questions x 500 persons) manually!

5.5 Selection of a method

There is no single method that is best for data collection. Each method has its advantages and limitations. Most often it is a combination of several methods which provides the best results. Table 2 is a comparison of seven methods of data collection. The selection of a particular combination of methods depends on the objectives of the study and the resources available.

Selection of a method

Table 1: A comparison of six methods of data collection for educational diagnosis


Type of data collected

Level or competence required

Time necessary

1. Literature review

Data available in the literature varies from country to country but can generally yield much more information than expected.

Good knowledge of the subject area in which the review is being conducted

A few days or weeks depending on the subject area.

2. Interviews carried out in a central location

Data can be collected from a significant number of persons in a location very heavily frequented by the target population

Interviewers must have received some basic training. Supervisors capable of advising them and analysing the data collected interviewers must be very knowledgeable

A few days on a full-time basis

3. In-depth individual interviews

Data collected from a limited number of resource persons

on the subject area and capable of conducting in-depth interviews, and analysing the data collected.

A few days or weeks depending on the number of persons to be interviewed.

4. Focus groups

Data collected from groups of people from the target population.

Facilitators and recorders trained in moderating the discussion and in observation and note taking as well as analyzing the information obtained.

A few weeks

5. Observation techniques

Data obtained from observation of the community health resources or selected households will and interpreted together with information obtained during interviews

Researchers must be well trained; social anthropologists be needed if an ethnographic profile is required

Several weeks, even months

6. Field survey

Data obtained from a statistically representative sample of the target population

Skills required for formulation of questionnaires, deriving the sample size and composition, conducting the survey and for organization and statistical analyses of data

Several months



Possible difficulties


1. Literature review


Information is sometimes not readily accessible especially in areas outside of the capital city

Reliability and validity of data obtained cannot always be determined

2. Interviews in a central location


Unwillingness of the people to participate in the interview

Not statistically representative of the population in spite of the valuable information that can be obtained. Data can be superficial

3. In-depth individual interviews

Quite expensive

Identifying "good" interviewers

Subjective views on the situation

4 Focus groups

Economical if one does not convene too many group sessions

Determining the number and composition of the groups Recruiting the participants. Striking a balance between free open expression and focused discussion. Interpretation of data.

Not statistically representative

5. Observation techniques

Very expensive because of the level of researcher required and the duration of their work

Focusing on the necessary observation elements of the observation grid

Perception bias on the part of the "observer on the observed"

6. Field survey

Very expensive

Terminating the process without omission of necessary steps

Bias due to desirability for social prestige. Information