Abstract
Research on children’s knowledge of contamination and contagion has been conducted mainly in industrialized settings, and little is known about it in Low- and Middle-Income Countries where contamination and infectious illness are significant health hazards. In this study, 75 children (4–9 years of age) living in Tanzania were interviewed about scenarios showing contaminated and contagious situations. They were also asked how they learned about contamination. All children had a good understanding of water and food contamination and contagion. Younger children were less knowledgeable than older children for scenarios that did not depict contamination, suggesting uncertainty in these situations. Children reported learning about contamination from relatives and teachers. Implications for health interventions, including health-related aspects of climate change, are discussed.
Keywords
Exposure to contaminated water and food and infectious diseases presents serious risks to human health, and children are especially vulnerable (Ferguson et al., 2013; World Health Organization [WHO], 2017). Childhood is a period of rapid growth, and proportionally to body size, children eat more food and drink more water than adults do. Exposure to environmental toxins and infectious agents can overwhelm children physically and, in some cases, lead to irreversible damage or death.
Children engage with the environment differently from adults. Children crawl and play on the ground or floor and are carried about and sit near other people, behaviors that can put children in close contact with tainted substances and infectious agents. Children can encounter contaminated water in household and community resources such as storage containers and local rivers or ponds. Even household water supplies from public utilities can be contaminated, as occurred in Flint, Michigan, in 2014 (Hanna-Attisha et al., 2016). Food can become contaminated by biological sources (e.g., viruses, bacteria), chemicals (e.g., pesticides, toxins), and methods of processing and handling (Rather et al., 2017). Food contamination has increased in recent decades because of changes in agribusiness, food transportation and storage, and the globalized food chain (Robertson et al., 2014). Concerns about contamination and contagion are paramount during public health crises and when environmental conditions drastically change (e.g., after a hurricane). Climate change has increased health risks from contamination and contagion as local sources deplete, pathogens spread beyond customary areas, and animals seek food and water at resources used by people (Intergovernmental Panel on Climate Change [IPCC], 2022).
One way to protect children from these health risks is to give them useful information that matches their understanding and capacities. Such information could enhance awareness of risks, provide guidance about how to avoid them, and increase personal agency. Developmental scientists consider personal agency crucial for helping children cope with the effects of environmental disasters and climate change (Thomaes, 2025).
Researchers can contribute by discovering what children living in different environments know about contamination and contagion (Trentacosta et al., 2016). Children living in Low- and Middle-Income Countries (LMICs) often encounter spoiled water or food during chores and need to evaluate the substance on their own. We report a study involving children 4–9 years of age living near Arusha city, a region of Tanzania with high rates of contamination and infectious illness. As background, we describe the prevalence of water and food-borne diseases in LMICs, including Tanzania, and review research on children’s understanding of contamination and contagion.
Water- and Food-Borne Diseases in LMICs
Diseases resulting from contaminated water and food account for a substantial proportion of childhood illnesses and deaths worldwide. Each year, diarrheal diseases kill close to half a million children; malaria, a vector-borne disease transmitted by mosquitoes that breed in lodged water, causes over half a million childhood deaths; and about 125,000 children die from food contaminated with microorganisms, parasites, or chemicals (WHO, 2019, 2021). Rates of illness and death are especially high among children living in low-income regions of the world. Unsafe water and food are related because contaminated water is often used for cleaning and processing food, creating a vicious cycle of disease and malnutrition (WHO, 2015). Many diseases associated with contaminated food and water can be prevented with better access to a safe water supply and improved water and food practices (Farmer et al., 2013). Involving children in these efforts is critical to their success because children’s responsibilities often include collecting water and preparing food (Nsamenang & Lo-Oh, 2010). Research shows that children can be effective change agents when they are involved in public health efforts (Onyango-Ouma et al., 2005).
Tanzania is among the poor nations in the world, 28% of the population is unable to meet basic consumption needs and 10% live in extreme poverty (World Bank, 2023). Malaria and diarrheal diseases are leading causes of morbidity and mortality in children below age 5. Child mortality has decreased in recent years, but health care still falls short of the United Nations Sustainable Development Goals (United Nations Development Programme [UNDP], 2023). Most Tanzanians live in rural areas, though over the last decade, rural to urban migration has increased. Migrants tend to settle in urban and peri-urban areas unprepared for rapid expansion and with conditions associated with disease transmission, such as inadequate sanitation and water supplies (Keiser et al., 2004).
Development of Knowledge of Contamination and Contagion
Knowledge of contamination and contagion includes awareness that an object or organism that is usually innocuous can be tainted by a harmful substance and, if contacted, can cause illness. Research conducted in industrialized settings has found that before age 3 children have a nascent awareness of contamination; for example, 18-month-olds will reject a liked food if it has contact with a disliked food (Brown & Harris, 2012) and 2-year-olds consider a food inedible if it falls on the floor (Toyama, 2000). With increasing age, the ability to detect and avoid contaminated substances improves and how children identify contamination changes (Siegal & Peterson, 1999; Toyama, 2019). Among young children, contamination reactions are based on visible cues, such as pollutants (e.g., dirt, insects), that trigger a disgust response (Curtis, 2013; DeJesus et al., 2015). As children get older, they know other sources of contamination, including invisible agents (e.g., germs, toxins) and environmental conditions (e.g., stagnant water). Children 3 to 4 years of age show some understanding of contamination and contagion from invisible sources. For example, they consider a substance contaminated if it had contact with a potential contaminant (e.g., it fell in the garbage; Kalish, 1996) or after a noxious substance (e.g., a bug) is removed (Siegal & Share, 1990), and 4-year-olds will avoid contact with a person described as sick (Blacker & LoBue, 2016).
As young as age 6 years, children’s understanding of contamination includes biological processes (e.g., having a cold can make someone susceptible to infection; Toyama, 2019). As biological understanding improves, children may overgeneralize the risks of contamination and contagion. In a study with 7- to 16 -year-olds, participants displayed good understanding of contagion, but younger children believed contagion is possible regardless of the type of illness (e.g., a headache, a cold, a stomachache from overeating; Marco et al., 2017).
How do children learn about the biological aspects of contamination and contagion? Some researchers contend that children have an intuitive understanding of biology that becomes more complex with instruction (Inagaki & Hatano, 2002). Personal experience contributes as children make sense of patterns they observe (Siegal, 1995). Cultural beliefs and practices may also play a role. Research with 4- to 9-year-old children in India and the U.S. found much similarity in their understanding of contamination (Hejmadi et al., 2004). However, when asked how to purify a contaminated substance, U.S. children consistently endorsed boiling water, whereas Indian children were skeptical of the effectiveness of boiling water in situations involving personal contact (e.g., someone sipped the water), reflecting Hindu beliefs about purification. Other research found that children’s explanations of illness reflect cultural beliefs. In China, the temperature of food is considered important to health and people avoid extreme food temperatures. Explanations for illness by 4- and 5-year-old children living in Beijing reflected this belief (Legare et al., 2013). In contrast, research involving children and adults (4–60 years of age) living in rural Uganda, where there are high levels of contamination-related illnesses, found explanations of contamination to be factual and to the point (e.g., “animals in water spoil it for drinking”; Gauvain & McLaughlin, 2016). Although the youngest children (4–6 years) exhibited less contamination knowledge overall, no participants cited cultural beliefs as the reason for avoiding contaminated substances.
To summarize, research indicates that children’s contamination knowledge changes from early to middle childhood and may reflect cultural beliefs and practices. There has been little research in LMICs where children have high rates of exposure to water and food contamination, and more research is needed in these contexts (Lansford et al., 2019).
The Present Study
This study investigated knowledge of contamination and contagion among 4- to 9-year-old children living in a peri-urban area in Tanzania, where daily activities put children in contact with contaminated sources and rates of infectious diseases are high. We also asked children how they learned about contamination and contagion. The research was conducted in partnership with Village Network Africa (ViNA), an authorized nongovernmental organization (NGO) that collaborates with The Sisters of Notre Dame, a government-certified NGO in Tanzania. These groups work with local communities to develop safe and sustainable water resources.
Method
Participants
Seventy-five 4- to 9-year-old children attending a government-accredited parochial school in the neighborhood or street (mtaa in Swahili) of Njiro participated. There were three age groups, each with 25 participants. Mean ages were: 4- to 5-year-olds (M = 4.89, SD = .49; 13 boys), 6- to 7-year-olds (M = 6.78, SD = .55; 12 boys), and 8- to 9-year-olds (M = 8.81, SD = .54; 12 boys). Children’s grade levels ranged from Montessori I–III (preschool and kindergarten) to Grade 3. Children spoke Swahili and English, the official languages of Tanzania; almost half (36, 48%) lived in Njiro, 14 (19%) lived in Arusha, and the remainder lived in villages within 15 km (9.3 miles) of the school.
Tanzania is ethnically and linguistically diverse and the sample children belonged to 16 ethnic groups. Most (60; 80%) were Chaga, or from other Bantu-speaking groups; five children were Maasai, eight were Arab, and two were Pakistani. The research was approved by the Institutional Review Board at the University of California, Riverside (Human Subjects Protocol No. HS-08-087). After formal institutional review, permission to conduct the research at the site was granted by the leadership of ViNA, the Njiro Congregation of the Sisters of Notre Dame, the school principal, and teachers. Consent was obtained from the children’s parents or guardians; children gave verbal assent.
Setting
Njiro, a peri-urban neighborhood of about 10,000 inhabitants, is approximately 10 km (6.2 miles) from Arusha, the third largest city in Tanzania with a population of 600,000. Peri-urban areas mark the transition between urban and rural areas. The community is primarily made up of Bantu-speaking African groups; some Maasai, a Nilotic-speaking people who are traditional pastoralists, live in the area. Diplomatic offices and the East African Community Headquarters are in Arusha and many non-Africans, including South Asians and Arabs live there.
Other than Maasai, people practice monogamy and the family typically consists of a father, mother, and, on average, four children (http://www.ciaworldfactbook.us/africa/tanzania). Other family members such as aunts or grandparents may live in the home, usually a small house or apartment. Houses are made of cement or brick with a small yard used for washing, keeping animals (e.g., chickens, goats), and a garden. Main industries are agriculture and tourism; there is some manufacturing (e.g., tires, pharmaceuticals). There is substantial poverty, and the water supply and sanitation facilities are inadequate.
Materials
Ten stories (one for training purposes) that depicted common situations at home and in the community were told to the children (see Table 1). Each story was illustrated with several grayscale hand-drawn pictures (2–5) printed on 12.7 cm (5 in.) laminated paper. The illustrations were adapted from publicly available materials developed by Africa Ahead (Waterkeyn & Waterkeyn, 2005) and The PHAST Initiative (WHO, Division of Operational Support in Environmental Health, 1996). Children were asked several questions about each story; responses were written verbatim, and interviews were audio recorded.
Story Descriptions and Contamination Status.
Cont. = contagion, F = food, T = training, V = vector, W = water.
C = contaminated, NC = not contaminated.
The training story described a non-contaminated water situation (a boy waters a droopy plant). The nine test stories varied in contamination status; six showed contamination or a contagious situation, and four showed safe (uncontaminated) situations. They covered several topics: cross-contamination (animals grazing near water), food and water safety (washing food, boiling water), vectors (juice with a bug in it), contagion (sneezing near a baby), hygiene (sharing a wash basin), and a modern water source (pump). Two independent judges identified the stories as either about contamination, contagion, or a safe condition; they had 100% agreement.
Procedure
Children were interviewed individually in an empty classroom during school hours. The interviewer, a co-author, provided instructions in English and a translator, a teacher at the school, provided this information in Swahili. The translator escorted the child to the interview room, where the child was seated at a table across from the interviewer, who introduced herself and explained that she would like to tell the child some stories and then ask the child about them. She explained how the translator, seated near the child, would help. The child was asked if they would like to participate and, after replying affirmatively, was informed they could stop the interview at any time. Permission was requested to record the conversation, and then children were asked for their age, grade level, and where they lived.
The 10 stories were presented in the same order for each child. The first (training) story familiarized children with the materials and interview process. The nine test stories described a situation in which a character in the story was or was not exposed to contamination or contagion. The interviewer began each story by placing the first picture on the table in front of the child and explaining it. This procedure was repeated until all the pictures were revealed. The child was then asked several yes-no (e.g., “Should the boy drink the water?”) and open-ended questions (e.g., “What will happen to the boy if he drinks the water?”) (see Appendix A).
To illustrate, Story 6 was about a girl drinking water that had been boiled. The interviewer placed the first of five pictures on the table and said, “The mother is boiling water to use in the house.” When placing the second picture, the interviewer said, “Mother is pouring the water into a clean bucket.” With the third and fourth pictures, the interviewer said, respectively, “The bucket is covered and the water is cooled” and “After a while, a girl is thirsty and she gets a cup from the shelf.” With the last picture, the interviewer said, “The girl brings the cup to the bucket and mother ladles water into the cup.” Then, the interviewer asked these questions, in turn: “Should the girl drink this water? Will anything happen to her if she does? Would you drink this water? Why or why not?”
After completing the stories, children were asked how and what they learned about water and food safety; then the children were thanked and escorted back to the classroom. Interviews averaged 30 min per child.
Coding
There were 18 yes–no questions (2 training, 16 test), and responses were scored as correct or incorrect. For contamination stories, the correct response is to reject the substance or behavior; for stories without contamination, the correct response is to accept the substance or behavior. Following correct responses, children were asked why they answered as they did. There were 16 such questions and each response was coded as follows:
Does not know or no relevant knowledge: child does not know or says something unrelated, for example, I like juice.
Phenomenism: phenomenological or circular response, may comment on the story, but no other information is provided, for example, It’s bad for you.
Presence of external agent: identifies a visible condition that could be hazardous, but no other information is provided, for example, He has dirty hands.
Health concern: identifies a health benefit or risk, for example, It makes you feel good; You can get sick.
Die: says the story character or the child themselves would die.
Health-related process: explains how the situation protects or threatens health, for example, Boiling water makes it safe; Dirty hands pass on sickness.
Two independent coders coded 100% of the responses. Reliability was very good (Kappa = .72–1.00, M = .88), with disagreements resolved by discussion.
Three questions probed other health-related issues, and codes were created for them. For the contagion story (Story 5), children were asked “What will happen if the girl (coughing in the picture) holds the baby?” Responses were coded as mentioning a physical or health risk (e.g., baby will fall or get sick), or a physical or health benefit (e.g., sister cares for baby; baby will feel better). For Story 4 about cross-contamination (a water hole with animals nearby), the follow-up question asked about water treatment (“How could you make this water better to drink?”). Responses were coded if they mentioned boiling, filtering, or chemical treatments. For Story 9, the child was asked if they would eat a mango lying on the ground, and why. Responses were coded as about taste (e.g., tastes bad) or health (e.g., mango on the ground is rotten). The reliability of two independent coders, based on 100% overlap, was excellent (κ = 1.00).
Learning About Contamination
Children were asked how they learned about contamination and what they learned. The identity of the instructor was coded as a relative, teacher, or health worker. Responses about what was learned were coded as phenomenism, a circular response (e.g., food is good to eat); environmental concern, the need to be careful about something in the environment (e.g., watch out for insects); or health-related information about avoiding possible contaminants (e.g., boil water; wash hands before eating). Two independent coders coded all the responses; interrater reliability was perfect (κ = 1.00).
Results
Preliminary Analyses
We examined if there were gender differences in any responses, and none was found (X2 values ranged from .073 to 3.21, ns); the data were collapsed by gender for all subsequent analyses. Most children (n = 72, 96%) responded correctly to the training questions, saying the boy should water the plant, which indicated they understood the task. The age groups did not differ (see Table 2). When asked why, 50 children (69%) said it would make the plant grow, 22 (31%) said it would make the food good, and 2 (2.8%) said it was too late, with no age group difference, H(2) = 2.06, p = .39, ns. When asked if they would water the plant and why, 66 (88%) children said they would and most (n = 36; 54.5%) said it would help the plant grow; some children said it would make healthy food (n = 28; 42.4%); with no age group difference, H(2) = 3.85, p = .15, ns.
Number (Percentage) of Children Who Answered Yes/No Questions Correctly by Age Group a .
The correct response is in parentheses following the question.
Response rate was not above change; responses not marked were above chance.
Contamination and Contagion Knowledge
A Chi-square goodness-of-fit test was used to determine if responses to the yes–no questions (2 training, 16 test) differed from chance, and most did, χ2(1) values ranged from 5.88 to 71.05, p < .05. For 15 questions, correct responses were greater than chance. For three questions, correct responses were below chance; two were about cross-contamination; one was about getting water from a pump (see Table 2).
A series of 3 (Age Group) between-subjects factorial ANOVAs examined the total number of correct responses to the 16 yes-no test questions and, separately, for the number of correct responses for stories about contaminated and not contaminated situations. We also examined rates of correct responding in the four content areas: contaminated water, contaminated food, vector, and contagion. Pairwise comparisons were conducted for significant results to determine which age groups differed.
Total Correct
Children answered an average of 11.97 (75%) yes-no questions correctly (see Table 3). There was an overall age group effect, F(2,72) = 3.25, p = .04, η2 = .08, and a medium size linear effect, F(1,72) = 6.50, p = .01, reffect = .28. Planned contrasts showed that 8- to 9-year-old children answered more questions correctly than 4- to 5-year-old children, t(48) = 2.28, p = .03, d = .64, a medium size effect; 6- to 7-year-olds did not differ from 4- to 5-year-olds, t(48) = 1.27, p = .21, or 8- to 9-year-olds, t(48) = 1.48, p = .15.
Means (and SD) for Correct Responses to Yes/No Questions for Contamination Status and Type of Contamination Variables by Age Group.
Seven questions pertained to water and three pertained to juice; the latter were also considered separately with the presence of a vector.
Contaminated and Not Contaminated Situations
For the 10 questions about contaminated situations, children answered an average of 8.27 (83%) correctly, with no age group effect, F(2,72) = 1.05, p = .36, nor linear trend, F(1,72) = .90, p = .35. For the five questions about situations that were not contaminated, children averaged 3.71 (74%) correct. There was an effect for age group, F(2,72) = 3.41, p = .04, η2 = .09, and the linear contrast was trending, F(1,72) = 3.46, p = .07. Planned contrasts revealed a difference between 6- to 7- and 8- to 9-year-old children, t(48) = 2.73, p = .009, d = .77, and a marginal difference between 4- to 5- and 8- to 9-year-old children, t(48) = 1.90, p = .06, d = .54. Four- to five- and 6- to 7-year-olds did not differ, t(48) = .60, p = .55.
Water, Food, Vector, and Contagion Questions
Children answered an average of 8.25 (82%) of the 10 questions about water contamination correctly. Although the overall effect for age was marginally significant, F(2,72) = 2.64, p = .08, η2 = .07, there was a linear age effect, F(1,72) = 5.16, p = .03, reffect = .25. Planned contrasts indicated that 8- to 9-year-olds performed better than 4- to 5-year-olds, t(48) = 2.17, p = .04, d = .62, a medium size effect. Six- to seven-year-olds did not differ from 4- to 5-year-olds, t(48) = 1.36, p = .18, or 8- to 9-year-olds, t(48) = .94, p = .35.
For the five questions about food contamination, children answered 3.01 (60%) correctly with an age group effect, F(2,72) = 3.89, p = .03, η2 = .10, and a medium size linear effect, F(1,72) = 4.89, p = .03, reffect = .30. Planned contrasts revealed that 8- to 9-year-old children performed better than 4- to 5-year-olds, t(48) = 2.13, p = .04, d = .61, and 6- to 7-year-olds, t(48) = 2.51, p = .02, d = .71, both medium size effects. Four- to five-year-olds and 6- to 7-year-olds did not differ, t(48) = .39, p = .69.
For the three questions about a vector, children averaged 2.75 (92%) correct, with no age group difference, F(2,72) = .42, p = .66, nor linear contrast, F(1,72) = .36, p = .55. For the contagion question, 71% of children answered correctly, with no effect for age group, F(2,72) = 1.61, p = .21, nor linear contrast, F(1,72) = .00, p = 1.00.
Children answered three questions incorrectly. Two were about cross-contamination; 56 children (75%) said the boy should eat the food after sharing a wash basin, and that they would eat food if the boy handed it to them. There was no age group effect for either question, F(2,72) = .068, p = .93. The other question pertained to drinking water from a pump, 48 children (64%) said it would be bad to drink this water, with no age group difference, F(2,72) = 2.12, p = .13. Children’s responses to these questions are taken up in the “Discussion” section.
Explanations for Accepting or Rejecting Substances or Behaviors in the Stories
Figure 1 shows the proportions of explanation types by age group for stories about contamination. When children correctly rejected contaminated situations, the most frequent explanation, provided by 51% of the children, pertained to health, with no age group difference, F(2,72) = 1.86, p = .16. The next most frequent explanation was based on phenomenism, with an age group effect, F(2,72) = 4.18, p = .019, η2 = .104. Follow-up t-tests revealed that 4- to 5-year-olds gave more phenomenism explanations than 8- to 9-year-olds, t(48) = 2.68, p = .01; 6- to 7-year-olds did not differ from 4- to 5-year-olds, t(48) = 1.71, p = .093, or 8- to 9-year-olds, t(48) = 1.20, p = .23. Less than 10% of explanations included biological information (treatment or process) and there was an age effect F(2,72) = 10.13, p = .001, η2 = .22. Follow-up t-tests showed that 8- to 9-year-old children gave more biological explanations than 4- to 5-year-olds, t(48) = 3.18, p = .003, and 6- to 7-year-olds, t(48) = 1.78, p = .001. Explanations based on visible information or that someone would die were infrequent, with no age group differences (Visible: F(2,72) = .60, p = .55; Die: F(2,72) = 1.29, p = .28).

Proportion of explanation types for contaminated stories by age group.
Figure 2 shows the proportions of explanation types by age group following correct answers for non-contaminated situations. Health concerns, stated by 30% of children, were the most frequent and there was an age effect, F(2,72) = 7.47, p = .001, h2 = .17. Follow-up t-tests indicated that 4- to 5-year-old children offered more health explanations for these items than 6- to 7-, t(48) = 3.44, p = .001, and 8- to 9-year-old children, t(48) = 2.90, p = .006; 6- to 7- and 8- to 9-year-old children did not differ, t(48) = .75, p = .46. Twenty-two percent of children gave explanations based on phenomenism and there was an age effect, F(2,72) = 4.02, p = .02, h2 = .10. Follow-up t-tests indicated that 4- to 5-year-olds gave fewer phenomenism explanations than 8- to 9-year-olds, t(48) = 3.11, p = .003; 6- to 7-year-olds did not differ from 4- to 5-year-olds, t(48) = 1.35, p = .18, or 8- to 9-year-olds, t(48) = 1.37, p = .18. Process explanations that included biological knowledge were provided by 20% of children, with no age group difference, F(2,72) = .51, p = .60. Few children (< 10%) gave explanations based on visible information, with no age group difference, F(2,72) = 1.18, p = .31. No children said the person would die.

Proportion of explanation types for not contaminated stories by age group.
Learning About Contamination
Most children (92%) said they learned about contamination from another person; 64% said relatives, 28% said teachers (see Table 4). There was an age group difference, χ2(4, N = 58) = 27.37, p = .001, φ = .36; 8- to 9-year-old children said teachers more than 4- to 5-year-old, H(1) = 22.86, p = .001, and 6- to 7-year-old children, H(1) = 20.88, p = .001. Most 4- to 5- and 6- to 7-year-old children said relatives. Few children (8%) said no one taught them, sometimes adding that they just know it.
Responses to Questions Regarding Learning About Contamination by Age Group a .
The full sample (N = 75) replied to the question “Who taught you?”, with 25 children per age group. Only 58 children replied to the question “What did you learn?”, broken down by age as follows: 4-5 years, n = 20; 6-7 years, n = 16; 8-9 years, n = 22.
When asked what they learned, 58 (77%) children responded and the age groups did not differ, χ2(6, N = 58) = 7.49, p = .28. Twenty-four children (41%) said they learned about environmental conditions (e.g., don’t drink dirty water), which was more frequent than phenomenological, χ2(1) = 8.00, p = .005, or health information, χ2(1) = 10.80, p = .001. Twenty children (34%) said they learned about behaviors that would protect them from contamination (e.g., boil water before drinking it), which was more frequent than phenomenological, χ2(1) = 5.14, p = .02, and health information, χ2(1) = 7.54, p = .006.
Finally, we examined whether responses differed by ethnic groups. When we omitted the 10 Arab and Pakistani children from the data set and reran the analyses, there were no differences in the findings. Also, close examination of the means (and SDs) of all variables by ethnic group indicated no difference in performance by child ethnicity.
Discussion
Children in this study displayed high levels of contamination knowledge for situations in which water or food was contaminated. Explanations for rejecting substances mostly centered on health concerns or environmental hazards. Inclusion of biological content varied by situation and child age and was rare when a contaminant was visible. For situations without a visible contaminant, most explanations from children 6 years and older included biological content. Some younger children gave explanations that hinted at biological (or biochemical) causality, such as this 4-year-old who said, “insect has gone in and is removed, but it’s still not good.” Older children’s explanations were more substantive, as this 6-year-old explained, “insect already spit poison in the juice.” The oldest group, 8- to 9-year-olds, had a clear biological grasp, and provided explanations like these 9-year-olds, “the insect put germs in the juice” and “something will remain there, dirt or bacteria, depends on the kind of insect.” When conditions were not contaminated, older age groups performed well, but many 4- to 6-year-olds responded incorrectly or stated health concerns, indicating uncertainty. This finding suggests that instructions about contamination for young children should include how to determine if a potentially hazardous situation is safe.
Overall, children’s explanations were factual and did not reflect cultural beliefs (e.g., “the food was washed in dirty water; it is not clean and can make you sick; the insect left germs in the juice”), consistent with some earlier results (Gauvain & McLaughlin, 2016). By including personal questions (i.e., would the food or water be safe for the respondent to consume), disgust reactions may have been activated. Disgust is a biologically based repulsion to substances unfit for consumption or contact (e.g., bad-smelling food, feces; Curtis, 2013). Although cultural beliefs may augment disgust reactions, disgust does not depend on cultural beliefs (Rozin et al., 2016).
Most children were knowledgeable about contagion and said that the child who sneezed should not hold the baby. Their reasons, even among the youngest children, indicated some understanding of the spread of disease from one person to another, as one 4-year-old put it, “the baby could get the disease from the girl.” Although there were only two questions about contagion, we feel confident about our interpretation. Furthermore, our data were collected shortly before the COVID-19 pandemic, and children’s knowledge of contagion has improved since that time (Leotti et al., 2021).
Children responded incorrectly to three questions. One was about the safety of water from a pump, and we are unsure why children responded incorrectly. The drawing showed a typical modern pump in the area, and people generally believe that water from these pumps is safe from contamination. It may be that children have little personal experience with these pumps, or they were uncertain and responded cautiously. The other questions that were answered incorrectly were about cross-contamination. The images showed children washing their hands in the same basin and eating a meal with their hands from a common container. Children’s answers may reflect the cultural value of sharing, an important norm in Tanzania, especially in rural and peri-urban areas. The value of sharing goes beyond food and goods; it includes sharing spaces (e.g., rooms at home), washing (e.g., a communal basin), and eating together (e.g., munching on the same bun or banana). In some communities, public-health information aimed at protecting children from cross-contamination in certain shared activities may violate cultural norms and should be approached respectfully.
Cross-contamination when sharing an activity is not unique to this region or to children. The COVID-19 pandemic revealed many such vulnerabilities in industrialized societies, especially when sharing food (Saulnier et al., 2023). When food is shared, infectious viruses can be transmitted in many ways: in speech droplets when talking or eating; on utensils, plates, and glassware when serving and eating food; and in food presentation (e.g., a common serving bowl from which food, such as bread or chips, is grabbed by hand). During the pandemic, people needed to curtail these behaviors to avoid cross-contamination. However, getting people to adopt new behaviors was difficult because it required altering well-established routines of preparing, handling, and serving food (Saulnier et al., 2023). In LMICs, suffering during the pandemic was largely due to poor infrastructure, fragmented health care, and lack of safe and reliable food and water supplies that protect against contamination and contagion (Sachs et al., 2022). Yet even when supplies exist, methods of storing and handling food and water can cause disease. For instance, handwashing with soap is effective, but soap may be in short supply, and maintaining this practice requires vigilance and funds (Curtis, 2013). Even in industrialized societies where soap is plentiful, adults do not wash their hands as often as needed to curb contamination (Mieth et al., 2021).
Children’s responses in this study were on par with those of children living in industrialized settings. How can this be explained? Some researchers contend that knowledge of contamination and contagion is an adaptive human characteristic with an evolutionary basis (Apicella et al., 2018). This view may have some validity, but it fails to recognize all the learning that occurs – learning contingent on environmental conditions that are themselves mutable. It is true that avoiding contaminated substances is sometimes triggered by a disgust reaction (Curtis, 2013), which may indeed have an evolutionary basis with little need for learning. But what about instances when a toxin or infectious agent does not arouse disgust? Here, again, we turn to lessons learned during the COVID-19 pandemic. In 2019, the SARS-CoV-2 virus was a new pathogen that emerged quickly and was dangerous to human health. It was unclear what was safe, and many protective measures were adopted, including distancing from others, wearing masks, and sanitizing groceries before bringing them into the house. Some measures were effective, and, in time, the reasons why became clear. That is, it was not until the pathogen was better understood that useful protective measures made sense. People living in industrialized settings, many with scientific knowledge of germs, did not do well initially in this unfamiliar and frightening context. Early recommendations for avoiding the virus overgeneralized the likelihood of contagion, the very kind of reaction we see in young children when they first learn to avoid toxic and infectious situations.
Learning plays a powerful role in helping children develop knowledge of contamination and contagion. People want to protect themselves and their community from contamination and infectious diseases, and much of this effort is directed toward children who are particularly vulnerable to negative health effects. Our findings indicate that even young children in LMICs are willing and able to learn about contamination and contagion, and that community members are instrumental in this learning. Children living in LMICs may be especially primed to learn this information because they may have witnessed the ravaging effects of contamination and infectious diseases in their community. Most children in our study, even the youngest age group, reported that family members and teachers taught them about contamination and contagion. In communities such as this, where illness from contamination and contagion is a major concern, older children and adults are proactive in helping children learn about these health hazards. This knowledge helps protect children, and public health workers can build on this foundation.
Strengths and Limitations
The effectiveness of our method in assessing children’s knowledge of contamination and contagion may reflect its significance in this community. Adapting the tasks to the setting, inviting a community member to translate, and using a familiar room in the school may have helped. Although the experimenter was not from the community, she was living and working there as a volunteer with a well-known and respected NGO team, which also may have helped.
There are several limitations. Reliance on verbal reports to assess children’s knowledge may underestimate what children know. Images of food and water are limited in what they convey about contamination; odor, temperature, and the reactions of others undoubtedly affect decisions about contamination in real-world settings. Naturalistic observations of children making decisions about water and food safety would have been useful. Finally, the sample size is small, and the focus on children living in a peri-urban area may limit generalization to urban and rural areas. Also, the climate and resources of an area, including health care and water supplies, may affect what children know about contamination and contagion.
Conclusions and Implications
Unsafe water and food supplies, sanitation, and hygiene account for most of the life-threatening diseases for children living in LMICs (Prüss-Ustün et al., 2014). These conditions are of central concern in community development projects, whose success relies on behavioral change (Farmer et al., 2013). Many projects are directed at adults, but children should be included (Jukes et al., 2007). Children spend much time away from adult supervision and engage in domestic chores and subsistence activities that involve food and water (e.g., preparing food, feeding young children, collecting and storing water; Nsamenang & Lo-Oh, 2010). Our findings show that even young children living in LMICs are interested in the causal factors involved in the transmission of illness from water and food, and they possess some awareness of the biological processes involved. The findings also suggest that these children are open to learning more about how to reduce contamination and contagion risk.
Children’s knowledge of contamination and contagion is an important aspect of cognitive development that integrates knowledge of biology, the environment, and human health. Because environmental conditions are a critical component of this knowledge, understanding this development across settings is essential. Environments vary in contamination risk, resources to inform and support children, and treatment when illness occurs. Improved understanding of children’s knowledge of contamination and contagion can inform parents, teachers, and public-health workers about the capability of children from early to middle childhood in understanding and dealing with these hazards. This information will be increasingly useful as the climate changes and exposure to infectious agents and contamination increases and takes on new forms (IPCC, 2022). Providing all community members, including children, with information about contamination and contagion may help reduce health risks as well as provide a greater sense of control in the widespread and potentially frightening conditions resulting from climate change (Sanson & Masten, 2024).
Footnotes
Appendix
|
|
|
|---|---|
| 1. Training | Should the child pour water on the plant? |
| What will happen to the plant if the child pours water on it? | |
| Would you pour water on the plant? Why or why not? | |
| 2. Boy playing soccer | Should the boy drink this water? |
| Would you drink the water after he does? Why or why not? | |
| 3. Dinner preparation | Should the children eat this food? |
| What will happen to them if they eat this food? | |
| Would you eat this food? Why or why not? | |
| 4. Sister gets water | Should the girl drink this water? |
| What will happen to her if she does? | |
| Would you drink this water? Why or why not? | |
| How could you make this water better to drink? | |
| 5. Girl and bab |
Should the girl hold the baby? |
| What will happen to the baby if the girl holds it? | |
| 6. Girl thirst |
Should the girl drink this water? |
| What will happen to her if she does? | |
| Would you drink this water? Why or why not? | |
| 7. Juice and bug | Should the boy drink the juice? |
| What will happen if he does? | |
| Would you drink the juice? Why or why not? | |
| If the boy took the bug out of the juice, would it be ok for him to drink it? Why or why not? | |
| 8. Washing for dinner | Should the boy eat the food? |
| What will happen if he does? | |
| Would you eat food if he handed it to you? Why or why not? | |
| 9. Mango on ground | Should the girl eat the mango? |
| What will happen if she does? | |
| Which mango would you eat? Why? | |
| 10. Water pump | Would it be ok to drink this water? |
| What would happen to you if you drank this water? |
Acknowledgements
We appreciate the help of Anita Boling, Sister Mary Rashmi, Sister Mary Shobana, The Notre Dame School in Njiro, TZ, the translator, and the children who participated in this research.
Ethical Considerations
The research was approved by the Institutional Review Board at UC Riverside and by the ethical review boards at Village Network Africa (ViNA) and the Notre Dame School and The Society of the Congregation of the Sisters of Notre Dame, Tanzania.
Consent to Participate
Informed consent was obtained from all participants and their parents or guardians.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We received intramural funding from the University of California, Riverside Office of Research.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
