Is It Possible to Implement AIDS’ Prevention in Primary School?

In France, health education is included in the primary school science curriculum. A part of this curriculum is called “human body and health education” (MEN, 2002). A quantitative study of teachers’ practices showed that teachers focus mainly on nutrition, hygiene, and dental health (Jourdan, & al., 2002). In the curriculum, the topic “Reproduction of living beings and sexuality education” concerns children aged 9-11 (Key Stage 2). Teachers often acknowledge that teaching about sexuality education and prevention of sexually transmitted diseases is difficult, because they do not feel comfortable with the subject matter. In a previous study (Jourdan et al., 2002), it had been shown that sexuality and AIDS were tackled by only 8 teachers out of 286 that were involved in the study. However, the curriculum guidelines of the French ministry of education (MEN, 2003) and the World Health Organisation (WHO) texts insist on the necessity for implementing early sexuality education and HIV/AIDS prevention programs, particularly in primary schools (WHO, 1999, 2004a). In this context, developing exchanges of experiences and partnership between teachers and health educators (school health services and health education NGOs) seems to be quite relevant. The nature of health education in schools also implies taking ethical considerations into account. The aim is not to promote a new secular morality defining “good” (healthy) and “bad” (risky) behaviours, but to prepare the children for responsible citizenship. Hence teachers in health education should not attempt to impose norms of acceptable behaviours, but should taking into account children’s peculiarities, expectations, needs, and also their representations. Children’s representations are thought to provide coherent models to represent learner reasoning when faced with a problematic situation (Jodelet, 1991; Farr, 1997). The construction of these representations is rather complex as this phenomenon depends on the values and beliefs shared by a social group, and which give rise within a social groupto a common outlook manifested during social interactions. As these representations are linked to an individual’s emotional responses as well as the cultural and social group(s) the individual belongs to, they constitute a decisive element in his/her relationships with the world, and are resistant to change. Representations therefore seem very essential (Fischer, 2001), are closely linked to behaviour (Abric, 1997), and cannot be changes as readily as knowledge. Any programme attempting to change representations should not only take into consideration the relevant knowledge, but also the social and cultural aspects of the


Introduction
In France, health education is included in the primary school science curriculum.A part of this curriculum is called "human body and health education" (MEN, 2002).A quantitative study of teachers' practices showed that teachers focus mainly on nutrition, hygiene, and dental health (Jourdan, & al., 2002).In the curriculum, the topic "Reproduction of living beings and sexuality education" concerns children aged 9-11 (Key Stage 2).Teachers often acknowledge that teaching about sexuality education and prevention of sexually transmitted diseases is difficult, because they do not feel comfortable with the subject matter.In a previous study (Jourdan et al., 2002), it had been shown that sexuality and AIDS were tackled by only 8 teachers out of 286 that were involved in the study.However, the curriculum guidelines of the French ministry of education (MEN, 2003) and the World Health Organisation (WHO) texts insist on the necessity for implementing early sexuality education and HIV/AIDS prevention programs, particularly in primary schools (WHO, 1999(WHO, , 2004a)).In this context, developing exchanges of experiences and partnership between teachers and health educators (school health services and health education NGOs) seems to be quite relevant.The nature of health education in schools also implies taking ethical considerations into account.The aim is not to promote a new secular morality defining "good" (healthy) and "bad" (risky) behaviours, but to prepare the children for responsible citizenship.Hence teachers in health education should not attempt to impose norms of acceptable behaviours, but should taking into account children's peculiarities, expectations, needs, and also their representations.Children's representations are thought to provide coherent models to represent learner reasoning when faced with a problematic situation (Jodelet, 1991;Farr, 1997).The construction of these representations is rather complex as this phenomenon depends on the values and beliefs shared by a social group, and which give rise within a social groupto a common outlook manifested during social interactions.As these representations are linked to an individual's emotional responses as well as the cultural and social group(s) the individual belongs to, they constitute a decisive element in his/her relationships with the world, and are resistant to change.Representations therefore seem very essential (Fischer, 2001), are closely linked to behaviour (Abric, 1997), and cannot be changes as readily as knowledge.Any programme attempting to change representations should not only take into consideration the relevant knowledge, but also the social and cultural aspects of the children's daily environment.(Doise & Mugny, 1997).The interest of taking into account pupils' representations in an HIV/AIDS education programme for children under twelve has been already justified (Fassler, Mc Queen, Ducan & Copeland (1989;Ferron, Feard, Bon, Spyckerelle, & Deschamps,1989;Thomas, 1991;Sly. Eberstein, Quadano, & Kistner, 1992;Schaalma, Kok, & Peters, 1993;Shonfeld et al., 1993;Kelly, 1995;WHO, 1999WHO, , 2004aWHO, , 2004b)).This chapter presents a collaborative research project attempting to identify and study the initial representations of 9 and 10 year-old pupils relating to aids and to examine the impact of an early educational programme on regular teacher's activities and interventions of health educators.Some of the initial results of the study have been already reported in a French journal for teachers (Berger, Collet, Laquet-Riffaud, & Jourdan, 2003).

Methodology
Most evaluations of health education programs are usually quasi-experimental designs, but to study health education other designs seem more appropriate (Victoria, Habicht, & Bryce 2004).In our context, using a controlled randomized study design as a method for assessing the effects of the implementation of a programme would be excessively difficult.The impact of the intervention on the children's social environment means that attempting to use a control group would be delusive, and that attaining true randomisation would be virtually impossible (Tones & Tilford, 2001).This situation results from the complex nature of causal chains in public health interventions.In spite of their limits, several authors have concluded in favour of collaborative research designs aiming at determining exactly what content and what tools would be most suitable for health education (Darroch, & Silverman, 1989;Heymans, 1993).Associating all agents in the design and implementing the programme based on collaborative research design makes it possible to make the interactions between researcher and agents more visible and transparent (Martinand, 2003;Merini, 2005).These would be otherwise masked and confounding factors.The data for the present study concern the two sides of the collaborative research.On one hand, an account of the general course of the study is provided and, on the other hand, the results from two questionnaires (pre-and post-questionnaire) that were used to collect information on pupils' representations are compared and analyzed.

Programme
The model on which this study is based relates to the "allosteric learning model" described by Giordan (1995).This socio-constructivist model assumes that learners build knowledge from their own lives, and learn through their mental representations that depend on their social and biological experiences, and their dispositions.Learning is a highly active mental process that operates in an integrative mode through the conflict between what a learner has in his/her mind and what (s)he can identify and understand from his/her environment.When a learner develops a new model, all his/her mental models must be reorganized based on an interaction between the pre-existing representations and new information from environmental sources (Giordan, 2000).Health education requires the teacher to take the pupils' representations into account and to help them construct new and more relevant ones.Moreover, each child's environment must be taken into account in the programme as children's representations are not only based on what they learned at school, but also on all the other aspects of their lives (Downie, Tannahill, & Tannahill, 1996).The research programme was developed by the Auvergne I.U.F.M. (Teachers' Training Institute), the I.N.R.P. (National Institute for Pedagogical Research) and the School of Medicine at the University of St Etienne, in partnership with the local School Health Services.The research design was regularly approved and evaluated by a pilot committee, which defined its ethical framework on the basis of the texts published by the French Society of Public Health.This pilot committee included representatives of parents' associations, Regional Health Authorities (DDASS), the School of Medicine, the Training Institute, primary teachers, the heads of the schools concerned, and the technical advisers of school health services.The implementation of the project in each school involved its approval by the school council, a meeting with the parents, the training of those involved, and the action in the classroom.4. Working with groups of children of adapted size (no more than 15). 5. Separating children into groups according to gender (separating girls from boys).6.Using a participatory activity design with games and tools that favour high rates of participation.The programme was developed on the basis of previous studies, (see Kirby, 2002 andUNAIDS, 1997).It was first piloted in a school during the school year preceding the study.The team that worked at each site was composed of six people (three per single-sex half group).Two persons from the research team, two representatives of school health services (a nurse and a doctor), and two observers who were to evaluate the teaching project and the way it was implemented.Evaluation of the process was carried out using the following indicators:  For the pilot committee, the number of meetings that were held was compared with the scheduling and the number of participants in each category (parents, teachers, doctors, and nurses).There were three interviews with all the members of the pilot committee, one before the project, one between the two sessions, and one after the results of the project had been made available.


For the school health services, an individual and anonymous questionnaire was used.It dealt with the form of the action, its pedagogical value, and the analysis of the elements benefiting health education in schools.Fourteen school nurses and 14 school doctors were interviewed.


For the school staff, the same type of individual and anonymous questionnaire was used.All the teachers and heads of schools involved in the programme (28) were interviewed.


The participation of the parents was measured for every meeting, and analysed in relation to the age group of the pupils and to the socio-economic status of the schools.Twenty interviews were carried out with parents from 4 categories of schools.


Each session was evaluated by an outside observer, using a grid including items relating to the way the session went, the interactions between adults and children, the involvement of the children, and the amount of time they spoke.

Population
The study was performed in the south east of France (the regions of the Loire and Haute Loire) in 1998-2000.It concerned pupils in "Cours moyen première année" (CM1) et "Cours moyen deuxième année" (CM2), which correspond to Key Stage 2. The sample was composed of 10 schools and 18 classes.Due to the small size of the sample, its characteristics do not correspond to those of the reference population, that is, it was not a representative sample.Nevertheless, schools corresponding to the main types of school in the country were selected (small size / large size; rural / urban; privileged / under-privileged).The research team asked teachers if they were willing to cooperate in the study.All the teachers that were questioned volunteered to have their class take part in the project.The overall results of the investigation concern 353 children.Among the participating children, 54% were girls and 46% were boys, while 31% and 69% of them came from CM1 and CM2, respectively.The total sample can be divided into 4 sub-groups depending on the social environment of the school.This classification was established using the criteria of the National Institute of Statistics and Economic Studies (INSEE 2003), that is to say, on the basis of the head of the family's profession.Population A (14%) was severely under-privileged (coming from schools classified as "educational priority zones").Population B (31%) was relatively under-privileged.Population C (30%) was quite privileged, and D (25%) was highly privileged.This classification brought out variations in the number of children per family.For Population A, there was an average of more than 4 children, for B and C, there was an average of 1.7, and for D, an average 1.5 of children per family.The children from Population A were the only ones to have parents with a significant age difference.The father was on average 10 years older than the mother, whereas, in the other sub-categories, the father was on average no more than 3.5 years older than the mother.However, the average age of the mothers in the four sub-populations was the same (35 years).The children classified in A were generally older than those in the other sub-populations and faced more difficulties at school.Sixty percent of them repeated a year at least once (16% for the other groups).

Questionnaire
Due to the age of the pupils, it was not possible to use either the same questionnaire for adolescents and adults, or a multiple choice questionnaire to determine, as it was done with adolescents, the way the children represented modes of infection.Indeed, unfamiliar words, coming from adult or adolescent vocabulary about sexuality, inhibited communication with young children (WHO, 1999).However, we designed a new questionnaire based on preexisting ones, but in which the vocabulary had been modified based on the results obtained in the pilot study.Thus, in spite of the fact that it made the questions harder to analyse, we used many open questions, sometimes along with closed questions.Using only closed questions would not have enabled us to grasp the complexity of the representations of AIDS in young children.
The validation of the questionnaire (understanding of the questions, coherence between writing questionnaire, and interview) was carried out at the end of the pilot study with a sample of children, who first filled in the questionnaire and then were interviewed.The questionnaire had 22 questions covering 7 aspects:  Initial representations of the HIV pandemic.


An assessment of communication about AIDS. Knowledge about AIDS  Modes of infection and protection.


Determining how close the subject feels the epidemic to be.


An evaluation of the representations of the possibilities of living with an affected person.


An evaluation of social and individual representations of solidarity towards affected people.The same questionnaire was applied for both the pre-test and the main test following intervention (series 1 and 2).For the analysis of our pilot investigation, we started by devising a thesaurus.Each answer was put in a lexical category and coded.This made it possible to take subtle differences into account.The total number of words was 255, and the number of items we added to the first version of the thesaurus after our first processing was low (< 10%).These precautions were taken in order to standardize the data acquired from the questionnaires and reduce any distortion in interpretation.

Teaching approach
We initially attempted to measure the impact of early preventive action on children's representations.The protocol was composed of two interventions in the course of the school year, one at the beginning and one at the end, at least six months later.Between the two interventions, the regular teachers worked on health education with the pupils ("normal" biology course including sexuality education).The two sessions were designed with the same pedagogical structure, which had two requirements, that is, to collect useful evidence from the questionnaires, and to put the children in a position where they were actors in their own learning process.The two sessions were structured as follows: A short presentation of the team and the framework, a question-writing time, a presentation about HIV/AIDS, work in small single-sex groups on the answers to the questions asked without the teachers, a game (a card-game for the first, and role-playing for the second), and, finally, the collective writing of a text for the teacher and the families.

Presenting the questionnaire
The questionnaire was intended to characterize children's initial representations and it was anonymous.After the pre-test, it appeared to be necessary, in order to attain this goal, to break away from the school environment and the behaviour it induces, especially in relation to writing.So, in the instructions for the procedure, we stressed that neither spelling nor the quality of the writing were important.What we were interested in was what the children thought, and in having them express their ideas in their own words.The intent was not to make things hard for the children by asking them to write, but simply to obtain their answers so we could analyse them and associate them with representations.We also explained that we would not give any further explanations about the meaning of the questions, as, we were afraid that in doing so, we could influence the answers.In order for all the children to be able to fill in the questionnaire as best as they could, we chose a collective approach.Each question was read out aloud and timed.Thus, we were able to include all the questionnaires in the analysis process, even those from children with serious literacy problems.

Information provided
This presentation was intended to provide precise and complex scientific information, and to give unity to sketchy and fragmentary representations, re-situating them in a context, and bringing out the link between the illness, the people, forms of behaviour, and oneself.

The children's questions
After children had filled in the questionnaire, they were invited to ask any questions they wanted to freely and anonymously, so that the educators could answer them in the second part of the session.Another form had been prepared for this and annexed to the questionnaire.Our aim here was to make the children put their questions in written form before the informational presentation, as well as to give us a representative body of questions, and to define these precisely before providing answers.While the children were at break, their questions were written out again, with no modification whatsoever.After break, the children were put in single-sex groups in separate rooms without their regular teachers so as to make it easier for the children to express themselves more freely on private issues pertaining to genitalia and sexuality.The presence of fellow pupils of the opposite sex and of the regular teacher that pupils will continue to study with could discourage the children from discussing these issues openly.The health educator then read out a question and asked the group to respond, only taking part to give clarification, to substantiate an answer, to get the children talking again, or to regulate the exchanges and make sure that everyone participated.This process was repeated for each question that had been asked by the children prior to the break.Our ethical approach was to use only the vocabulary from the presentation or that was used by the children, excluding any words or expressions coming from adolescent or adult vocabulary, particularly in the field related to the management of sexuality.This was essential as we found that use of unfamiliar sexuality related terms coming from adult or adolescent vocabulary inhibited communication and thwarted our objectives.However, by using in our answers exactly the same expressions and words that the children used to formulate their questions, which were sometimes very direct questions about sexual practices, we could show the children that any subject can be tackled with them.The educator's role was mainly to get the discussion going, to modify, or to substantiate the representations by clarifying points, and, if necessary, to offer extra help in completing fragmentary or sketchy knowledge.

Teaching tools
The card game in the first session: The card game was devised for this experiment and for this particular group.It was based on an approach developed for adolescents (Ricard, 2000) and on the results of the pilot study.It included situations in daily life concerning both close relationships with affected people and more distant situations, so as to enable the children to express their certainties and doubts, and the rumours they had heard.The rules were simple.Each child was given some cards.He read out what was written on the card, showed it to the group, and put it down on one of three cards which indicated no risk, I do not know, or high risk.The child explained his choice and then asked the group to say what they thought.This approach enabled us to involve all the children, even the shyest, and gave them an opportunity to express themselves.Role playing game in the second session: The aim of this activity was to get the pupils to talk about HIV/AIDS while adopting a point of view different from their own.They had to take the role of parents, teachers, and children in concrete situations.This game is intended to put the children in a situation where they could express and become aware of their own representations of the pandemic, the risk of infection, and the ways of protecting themselves.This projected identification had a powerful emotional component.

Final written work:
The children dictated to the educator an account of what they had done, or of the ideas and things which they felt to be important, and which they, therefore, wanted to share with their families and class teacher.The advantage this strategy had over an individual account was that it did not put the children in a difficult school situation by asking them to write.It also made it possible to summarize what was essential.

Statistical analysis
The questionnaires were processed by the statistics department at the St Etienne School of Medicine, according to the thesaurus drawn up during the pre-test, using Epi info 5.01 and SPSS.The level of estimated statistical significance applied for the tests was p. < 0.05.When the size of samples was small, the adjusted Khi2 (Yates method) was used and, if the size of one of the samples was beneath 5, we kept the results given by Fisher's test.The analysis was only univariate.The questions asked by the children were analysed using the method of the "analysis of content" (Bardin, 1993).We therefore put the answers together according to their semantic structure, and observed combined frequency indicators ( co-occurrence analysis), which enabled us to establish links between the data (Microsoft Access).The data described here focus on a comparative study of the results of the two questionnaires.However, the programme was also assessed by the pilot committee, the school medical staff, the teachers and the parents.

Evaluation of the process
The pilot committee: The committee supervised the research activities all the way throughout the entire project.They met before the sessions to validate the protocol and also defined an ethical framework based on respecting people, and respecting the convictions of the children and their families.After the first session, the results of the first set of data were presented, as well as a report written by observers from outside the team about the way the ethical framework had been respected, and how the sessions had gone and been managed.Once the whole protocol had been applied, the different results and analyses were presented and discussed.All members of the committee attended regularly, including parents' associations.In the interviews at the end of the project, the committee members declared that their opinions had been taken into account.Medical staff: The evaluation of the schools' medical staff (school doctors and nurses) was carried out through an anonymous individual questionnaire.The entire data set obtained by this questionnaire cannot be analysed here.The results show that the medical staff found the organization relevant.After the experiment, they admitted that they felt more comfortable about tackling the issues of AIDS and sexuality in a comprehensive approach to health education for young pupils.They expressed their need for training, to update their knowledge about HIV, to learn how to teach health education, and to develop their theoretical and pedagogical background.Teachers: For the teachers, an anonymous individual questionnaire was also used.Teachers said they were in favour of this kind of intervention in schools, insisting on how advantageous it was to build up partnerships with competent professionals who have been trained for such actions with children, not with a view of making up for insufficiencies or to replace the class teacher, but to working with the teacher on a common project that is part of their syllabus.Before the intervention, most teachers found it hard or even impossible to talk about such matters with their pupils, although they were well aware of the need for it.The reasons they put forward for this were: (a) They did not have enough knowledge about the disease, the way it is caught, and what protection can be used.(They considered that the only information they had was from the media, and deemed this to be inadequate for giving precise information to children).(b) They were afraid of how the parents might react as they considered this topic to a delicate or sensitive subject.(c) They found it hard to tackle questions about sexuality with children.(d) They were worried that they might be asked questions that they could not answer.Moreover, they all stated that they had changed the way they considered having HIV positive children in their school, and felt better prepared to tackle the issue with parents and colleagues.The parents: Parental attendance at meetings organized in each school before the interventions varied enormously in relation to the social category involved.Families from the most under-privileged social categories attended less than the others.The aims of the meetings were to present the collaborative research project, answer any questions, and give an account of the results.Right from the start, we noted that it was not really possible to get parents from the most underprivileged schools involved, and the number of parents present was always very low.However, there was a high attendance rate for parents from more privileged schools.As a result of these meetings, it was obvious that very few parents were against early AIDS prevention, and there was not any obvious and definite opposition.The observations made by parents mainly concerned their desire that family religious and philosophical beliefs be respected.

Analysis of the questions asked by the children
We were able to study 350 forms.The variables which we used were gender (190 girls and 160 boys), the class at school (114 CM1 and 236 CM2), and the social class (highly privileged 88, quite privileged 103, quite underprivileged 109, and seriously underprivileged 50).Only the first ten questions asked by each child were taken into account and analysed.During the first session, the children asked a total of 1267 questions, and during the secon 759.Thus, there was a drop of 40% (p.<10-3).The average number of questions asked in the first series was 3.62 per child, and for the second 2.16.In the first session, 95.7% of the children asked at least one question, and 73.7% in the second.The number of questions asked per pupil goes down significantly faster in the second series than in the first (p.<10-3).Between the two sessions, there was a significant increase (p <10-3) of the number of children not asking any questions, rising from 19 in the first series to 96 in the second.The analysis of the questions showed that the changes varied according to the item concerned.There was little or no change for the questions about the disease, "love and sexuality," anxiety, the fight against AIDS, and living with the virus.There was a significant decrease in the number of questions concerning the modes of infection (p.<10-3) and protection behaviours (p.=0.025).The questions on protection, anxiety, attempts to understand, and even the questions on modes of infection, go down much more for the boys than for the girls.The children coming from severely underprivileged families still asked a lot of questions (p.=0.03), as did those from a highly privileged background (p =0.04).

Analysis of the questionnaires
The pupils were required to complete the questionnaire before session 1 and before session 2. The results are shown in Table 1.For the closed questions, the results are expressed as percentages of the total number of questionnaires taken into account in the analysis.For openended questions, the responses have been grouped into different items and are expressed as percentages of the total number of questionnaires including an answer to the concerned question (data are given in Table 1 only if the items are cited in more than 5 % of the cases in session 1 or 2).For multiple choice questions, the total percentage could exceed 100, because children were allowed to give more than one answer.When a significant impact of gender, age, or social status on the responses was observed, it is indicated.When a significant difference was observed between second and first session, the data are in bold print.The analysis of the first questionnaire gives an overview of the initial representations of the pupils.The results are shown in Table 1.The comparison between pre-and postquestionnaires guided us to identify where a modification of representations was observed.The analysis was performed taking into account five points: (a) knowledge about AIDS, (b) communication about AIDS, (c) knowledge about the disease, (d) knowledge about modes of infection and protection, and (e) relationships with affected people (the analysis of the other parts of the questionnaire are not shown in this article).

Knowledge about AIDS
The analysis of the first questionnaire (pre-test) indicated that more than 92% of the children had information about AIDS, while, six months later, this percentage increased to 98% for the second questionnaire.The main source of information was television (88%) followed by the family (25%).However, these results (Question 2) were inconsistent with the results from another question (Question 4), where more than 65% of the children stated that they had talked about AIDS with their families.The only source of information which changed significantly between the two questionnaires was the school (p.<10-3).Children mainly associated AIDS with words suggesting, Illness, Death, and Sexuality.They also mentioned, to a lesser extent, condoms, blood as a vector for infection, taking drugs, and finally prevention, and solidarity.The intervention did not trigger any substantial change in initial associations with Illness/Death/Sex, but it nevertheless allowed most children, who had not ever discussed the subject, to be involved in discussions about AIDS.Three-quarters of those who did not mention anything initially, did contribute after the intervention.Thus, the highly privileged group D referred initially to sex and sexuality more than the severely underprivileged group (p.<10-3).But, this difference was much smaller at the end of the session (p.=0.05).

Communication about AIDS
Figure 2 shows the differential influence of socioeconomic status on the impact of communication about AIDS with adults (Have you ever talked about AIDS with adults?) and in the family (Have you talked about AIDS in your family?).While an increase incommunication with adults was observed for all 4 groups, it was limited to the  underprivileged, privileged and highly privileged groups for communication inside the family.Results are expressed as percentages of the total number of questionnaires including an answer to the question.Pupils also exchanged on the topic of AIDS with adults, with friends and at school.Fifty-one percent of the children had talked about AIDS with adults before the intervention.At the end, 76 % of them have talked about the subject with adults, either before the first session or between the two sessions.The intervention did not bring on a significant increase in discussion of AIDS within the family in the severely underprivileged group (p.=0.3 49%), unlike in the other groups, where there was a significant increase of 74%, 79% and 85%, for groups B, C and D with p=0.01, p.=0.01, and p=.001, respectively, indicating that communication between the pupils was also enhanced.(clarify the meaning)

Knowledge about the disease
Before the intervention, more than half of the children associated AIDS with a fatal illness.On a scale ranging from 0 to 10, the children rated the dangerousness of AIDS at more than 8. Population A alone stands out by assessing its gravity at less than 8 (p.=0.007).The illness which is symbolically associated with AIDS is cancer.Infectious illnesses are not often quoted, and only 5% of the children mention Hepatitis B. After the intervention, we found that references to infectious diseases dropped considerably, and associations with childhood illnesses disappeared.Two-thirds of the children stated that they knew what a virus is, and were able to give a relevant explanation, with a definition based on one of three 'concepts,': a microbe, an illness, or a vector of an illness.However, only one-third knew what HIV positive means.

Modes of infection and protection
Before the intervention, 88% of the children associated AIDS with a transmissible disease and 97% after the intervention.The change was slight but significant.In the pre-test, 74% of the children correctly answered the question "What gives you AIDS ?" and in the post-test 89%.For the children, AIDS is transmitted by vectors: secretions (sperm), sex, drugs, and the HIV virus; and by behaviour: sexuality, drug addiction, and medical practices related to the handling of blood, such as, transfusion and giving blood.Drug addiction was scarcely mentioned, and references to syringes or exchanging syringes were very uncommon.Similarly, references to materno-foetal transmission, and to incorrect vectors, such as, saliva, mosquitoes, daily actions, morality, or God, were almost non-existent.The lexical field used was fairly limited, but it was wider in the second session.The question was put in such a way as to give the children the possibility of replying by designating supposedly high-risk groups (homosexuals, prostitutes, drug-addicts, dirty people, and others).The pupils did not consider that people identified as 'deviant' were responsible for beginning the infection.As far as modes of infection are concerned, after the intervention there was a modification concerning the answers about vectors of infection, and those about behaviour.Representations definitely became clearer.Before the sessions, more than half the children explained that contamination came from vectors: sex (1/2) and drugs (3/4), but after the session, they referred to "dangerous" behaviour (90% sexuality and 50% also mentioned drug addiction).Preventive action modified representations concerning modes of infection (p.=0.001).However, this reversal was less obvious for the very underprivileged social categories (p.=0.03).
In order to know whether an individual may have been infected, more than half the children suggested active solutions, such as, having a test, or going to see a doctor.Fifteen percent suggested passive solutions, waiting for the symptoms to appear, or waiting till you feel ill.The girls suggested fewer active solutions than the boys (p.=0.013), and the severely underprivileged children fewer than the highly privileged (p.=0.049).After the intervention, reference to detection increased considerably (p =0.002), and there was less mention of adopting a passive stance or waiting for symptoms to appear (p =0.016).Prior to the intervention, 68 % of the children suggested the condom as a way to be protected, and this percentage increased to 91% afterwards.The intervention mainly gave rise to a considerable increase in references to condoms, protection, and avoidance.There were no statistically significant difference related to age, sex, or social status in this increase.

Relationship with affected people
One out of two children had heard of someone who had or had had AIDS, both before and after the intervention.Only one in ten had heard of it through a channel other than television.Before the intervention, 64% of the children thought it was dangerous to live with an HIV positive person.Twenty-nine percent continued to think so, even after the intervention, but there was a significant change in the way infected people are seen and in the perception of the absence of risk of infection in everyday life.

Discussion
The aim of our study was to identify the initial representations of pupils on AIDS/HIV and to analyse the impact of an educational programme based on regular teacher's activities and interventions of health educators on these representations, on communication about AIDS/HIV, and on the way in which infected people are seen.The main novel features of our study were its target (young pupils aged 9 and 10), the close partnership between teachers and health educators, the involvement of parents, and the fact that it was based on a learner-centred model (the allosteric model as described by Giordan, 1995).First, we are going to discuss the relevance of such a research design and, secondly, we will analyse the pupils' initial representations on AIDS/HIV and the impact of the program.Finally, the issue of communication about AIDS in the family and with peers will be addressed.The main characteristic of collaborative research is the close involvement of the target population in the development and management of the program, or, in other words, the proximity between researchers and actors (Martinand, 2003;Merini, 2005).It also aims at an improvement of practices here and now.Our study shows the interest of such a design in AIDS/HIV prevention.Indeed, the actors (teachers, parents, doctors, nurses etc.) were highly involved in the programme throughout the two years it took place.The intervention was conducted in a coherent manner in relateion to the educational environment of the pupils.In addition, the design lead us to take into account the ethical issues linked to preventive intervention (respect for people, cultures, family upbringing etc.) Nevertheless, we must also underline the limits of such a design.It was time consuming and the involvement of the under-privileged group was lower than that of the other groups.As described in previous studies (e.g., Anochie & Ikpeme, 2003), the analysis of the initial questionnaires indicated that 9-and 10-year-old children did have representations of the HIV pandemic, the people affected, and the modes of infection and protection, but they had incomplete information on the subject.More than half of the pupils associated AIDS with a fatal illness as serious as, or more serious than cancer, transmitted by 'sex,' and 'caught' especially by adolescents and adults.They thought the illness could be avoided by putting on a condom (68 %), and detected by 'tests' or going to 'see a doctor' (80%).The content of their scientific statements was still at times completely or partially incomprehensible, as they could not fit them into a more general conceptual framework of knowledge, which would allow overall understanding (Kirb, Short, Collins, Rugg, Kolbe, Howard, 1994;Kirby, 1995;UNAIDS 1997).It can be noted that the highly privileged group D refered to sex and sexuality more than the severely underprivileged group A. It was also evident that the severely underprivileged children generally used a much more limited lexical field than the others.This observation was evident in the questionnaire as well as in the analysis of the transcripts of work in sub-groups.This lexical limitation seemed to have interfered with establishing complex representations, and these pupils were not able to avoid reductive over-simplification.At the end of the session, more children answered most of the open questions, and did so using more words.The lexical field concerning biomedical knowledge was of higher qualityRegarding modes of infection, we found the focus on vectors of infection decreased whereas attention tobehaviour increased.Before the sessions, more than half the children explained that contamination came from vectors, such as sex or drugs, but, after the session, they mainly referred to dangerous behaviour (sexuality, drug-addiction).Regarding protection, the study showed the interventions had had considerable impact.At the end of the sessions, only 8 children answered that you cannot avoid catching AIDS.There was a 150 % increase in the number of children stating that "the condom protects you from HIV infection" and three times more children spoke about protective behaviour.These data have to be interpreted with precaution, because it is well known that there is no direct link between knowledge and behaviours (e.g., UNAIDS, 1997).In addition to the influence of socioeconomic status on children's representations, we observed an influence of age and gender.The representations of the 10-year-olds were more relevant than those of the 9-year-olds, who are still quite childish,.Researchers working on representations in children of different ages have made similar observations (BMA, 1997;UNAIDS 1997;Brown 1990).However, most authors found little difference between girls and boys (du Guerny &Sjoberg, 1993;Guthrie, Wallace, Doerr, Janz, Schottenfeld, Selig, 1996;Prah Rugger, 2004;UNAIDS, 2004).The study also investigated communication about AIDS.People with whom pupils speak about AIDS were mainly their families and peers.Nevertheless, in the second questionnaire, only 1 % to 4 % of them stated that they had never heard their friends talking about AIDS.In a study performed with primary school children (11-yearsold), Anochie and Ikpeme (2003) found that friends were not an important source of information for pupils (4 %).It is not easy to interpret this statement, as, in question 3, 44% of the same children stated that they have talked about AIDS with other children.Perhaps this contrast indicates that other pupils are not considered to be a worthy source of information, bu comparison to other sources, which they see as more knowledgeable.It is highly likely that the children hear more about AIDS through the media than from their friends, which caused them to underestimate the importance of the information they got from their peers.Moreover, the children appear to have discounted this information as not being serious and, therefore, not worth mentioning, in comparison with information given by experts on the TV, 'which tells the truth.'This interpretation also proved to be valid with the analysis of the work done in the sub-groups of the study.About 62% of the children have talked about AIDS in their families before the intervention, whatever their age, sex, or social origin.In the second series, 76 % of them have talked about the subject with their families, either before the first session or between the two.But our intervention did not bring any significant increase in communication within the family in the severely underprivileged group.These data show how hard it is to get a family to talk about AIDS, particularly for the severely underprivileged, and raises the question of family communication in the field of health education.It is likely that the intervention triggered discussion in families where there was a readiness for this.Our analysis shows that more than 90% of the families of the underprivileged group were of foreign origin (North African and Turkish).Talking about sexuality, especially with boys, in a cultural framework that was profoundly steeped in tradition, meant adopting a new Western-style cultural position.Thus it was difficult to talk about such a private subject in the family.Their priority was apparently to not deny their origins, and to preserve their identity, so as not to be swallowed up by integration, which was experienced as culturally destructive.As a result, no standard model of intervention could be put forward because the cultural dimension was a significant variable in actions and their impact (Rosenthal, 1990;Tones & Tilford, 2001;WHO 1997WHO , 2004aWHO , 2004b)).The whole community must really be involved when the intervention occurs in a multicultural environment.The analysis of the interviews indicated that communication about AIDS in families and between friends was related to an external stimulus, generally the media (but sometimes school).Television news and special programmes made families react.Families who tackled the issue without any direct link with the media were only few and far between.When they it was more frequently to warn children about the risks of sex and drugs than to incorporate this into a more general discussion about exclusion, life, its risks and the management of these risk, or about sexuality and pleasure.The cross analysis of the questions showed that when the question of the integrating an HIV positive person in different situations was raised, the attitude of children from families where AIDS was discussed was no different from that of children from backgrounds where it was not.So, it would seem that the family message did not focus on the integration of infected persons.Nor was it a message of exclusion.It was likely that the parents' message did not concern infected people.The reality of the infected person remained largely virtual.Information mainly came from the mass media and television, and contact with sufferers in their daily lives was rare.

Conclusion
This study shows an evolution in the representations of pupils about HIV/AIDS.The intervention led them to build new representations that take more objective facts into account.These results are interesting but have to be discussed, as it is well known that there is no one to one link between knowledge and behaviour.The mere provision of knowledge is not enough if the aim is a relevant scientific education, but the educational process here includes helping children "to clarify their values in relation with themselves, health, healthinfluencing behaviours" (Downie et al., 1996).In addition, such an intervention makes it possible to talk much more about a much broader spectrum of themes related to health.In working on HIV/AIDS prevention and sexuality education, numerous other aspects of science education are tackled, and mainly the status of science in relation to everyday life (nature of science and scientific knowledge, application of science concepts, values that www.intechopen.comunderlie science etc.)By providing an HIV/AIDS education programme, it is only possible to promote a comprehensive health approach (St Leger & Nutbeam, 1999), if the whole educational environment is involved, if the intervention is really learner-centred, if the programme is sufficiently open and does not aim at enforcing some form of behaviour, and if the ethical framework is clearly defined.Such an approach, to be effective, must take into account the complexity of health, and the factors which influence it, but also actual science education theory and practice.This last point is decisive as one of the most important difficulties in implementing relevant programs is, in addition to taking into account cultural and social diversity, the involvement of teachers and school staff (Ayo-Yusuf, 2001;Han & Weiss, 2005).1. Analysis of the responses to the questionnaire.The pupils had to fill it in before session 1 and before session 2. The results are shown as follows.Closed questions: results are expressed as percentages of the total number of questionnaires taken into account in the analysis.Open questions: the responses are put together in different items; results are expressed as percentages of the total number of questionnaires including a response to the concerned question (data are given in the table only if the items are cited in more than 5 % of the cases in session 1 or 2).For multiple choice questions, the total percentage could exceed 100 because children were allowed to give more than one answer.When a significant impact of gender, age or social status on the responses is observed, it is indicated in the table.When a significant difference was observed between second and first session, the data are in bold print.CM1: young group (age 9), CM2 old group (age 10), A : severely under-privileged B : relatively under-privileged C : quite privileged, and D : highly privileged.Statistical significance : Impact of sex, age or social status on responses in session 1 or session 2 : * p < 0.05.difference between session 2 and session 1 : + p < 0.05.

Fig. 1 .
Fig. 1.Research's design Figure 1 presents the collaborative research design founded on six principles: 1. Insure complementarily between regular teacher's activities and interventions of health educators.2. Thoroughly preparing the context of the project by involving the families, teachers, and school health services in the comprehensive approach.These partners actively participated in the design of the study (questionnaire, interventions in the classrooms, relationship with the population).3. Inclusion of all the classes at each school level investigated.

Fig. 2 .
Fig. 2. Influence of Socioeconomic Status on the Impact of the Intervention on Communication about AIDS.