Methodological Implications of Contextual Diversity in Research on Street Children

Abstract

This paper draws on findings from research conducted by the authors in Nairobi, Kenya; Cali, Colombia; and Addis Ababa, Ethiopia to suggest the significance that the diversity of street children has for research. It considers implications of this heterogeneity for sampling and data collection, while advising caution in taking the validity of studies of street children at face value.

Keywords: street children; research methods; diversity

The Diversity of Street Children

Since its inception the term “street children” has been used to refer to children in a variety of circumstances, creating confusion about who street children are and what kinds of experiences brought them to the streets. The terms “on” the street, “of” the street and “on and off” the street are commonly used to classify street children in many countries (Ministry of Labor and Social Affairs 1988; 1992; UNICEF 1993; Heinonen 1996). Children “on” the street are engaged in the street but have regular contacts with their families. Children “of” the street live, work and sleep in the street. This distinction is usually referred to as a “UNICEF definition” (Ennew 1996).

However, this classification is too rigid because it does not correspond to the realities found in most big cities. Street children do not form a homogeneous group, nor do their life circumstances remain constant. Besides the natural transition from childhood to adolescence and adulthood, children's involvement in street life and family contact varies. Furthermore, children may be represented in one or more of the three categories listed above at different times of their lives and careers in the streets. Many alternate between home and street life. UNICEF has become aware of the difficulties with its own definition of children “on” versus children “of” the street. Since the 1990s, it has been grouping all working children, whether working on city streets or elsewhere, as “working children.” It uses the term “street children” to refer to the smaller number of largely abandoned children and youths for whom the city streets are home (Barker and Knaul 1991). Cosgrove (1990) and Lusk (1992) have added deviant behavior such as drug abuse, thieving and violent conduct towards other people to the UNICEF classification of “street children.” It is clear, however, that a variety of factors can interact with these definitional elements to produce very different situations for children. The following sections illustrate this diversity.

Ethnicity

In Ethiopia, Heinonen (2000) found that poverty is a necessary but not sufficient condition that spurs many children to street life, since many poor children in Addis Ababa do not become street children. As a case in point, among the Gurague, ethnicity plays an important part in the lives of street children who have regular contact with their families. They are a singular example of a people whose family ties are not disrupted by migration; they maintain continuity between their rural area of origin and their current urban residence. Urban migrants avoid making a complete break with their rural socio-economic roots and keep the urban-rural link alive (Seifu 1976). Thus the Christian Gurague return to their villages for Meskal (Feast of the Finding the True Cross) every year, just as the Moslem Gurague return home during Arefa after the Ramadan ceremony.

Against this background, many Gurague shoeshine boys and street workers live communally with their peers, older siblings, or relatives. They draw lots to decide who will go home in any given year if they cannot all afford to go back. Those unable to visit their families due to financial constraints make all sorts of sacrifices to send presents to their relatives. Because in a majority of cases Gurague street children are perceived as well-behaved and hard-working, they have a better public image than other street children.

Housing

A unique feature in the situation of street children in Addis Ababa is the historical context of substandard housing. Approximately 75 percent of the sub-standard housing is owned by the state and controlled by Kebeles (local government). If people appear on the official Kebele register, in addition to living in rent-controlled Kebele houses, they will have access to some free education for their children, rudimentary health services, and possibly even jobs and national citizenship papers. Since poverty in Addis Ababa is pervasive and severe, those living in poverty try every means possible to belong to a Kebele. Economic migrants from rural areas and people displaced internally due to civil strife continue to exacerbate the housing crisis in the capital. Kebeles are unable to accommodate those already in the city, let alone house these newcomers. As a result, the children of newer in-migrants inevitably face more difficult circumstances (Tegegne and Daniel 1997).

International and local NGOs are increasingly cooperating with Kebele officials to identify households to include in their projects. Of the 52 families that were part of the Heinonen (2000) sample, 22 had established personal ties with Kebele officials, teachers, health care workers, and neighbors. They were able to secure Kebele houses, which put them in a better position to provide an emotionally and materially supportive environment for their children than the rest of the sample.

Segregation

As in other cities of the developing world, in Addis Ababa the houses from which street children come are made of plastic, mud or cardboard even though they are adjacent to tall buildings and modern villas. They typically consist of a single room no larger than three by five meters where between four and ten people eat, socialize, and sleep. The children commonly sleep on the floor, fully dressed. They cover themselves with skimpy, tattered, filthy blankets and other assorted coverings to ward off the all-year-round cold nights. Many do not have a second outfit. They are obliged, therefore, to sit naked while their clothes are being washed and dried. During the rainy season not only the streets, but also their houses, become flooded.

However, unlike cities such as Nairobi (Kenya) or Cali (Colombia), Addis Ababa is unique in having no squatter settlements. The situation of street children is, therefore, unique. The poor are not concentrated into separate sections of the city. They live scattered about the city among socially diverse communities. It is common to find a family with good shelter and food and medical care next door to poor families among whom the infant mortality rate averages 101/1000 and the under-5 mortality rate is 152/1000 (UNICEF 1993).1 Socio-spatial segregation is not dominant. Street children, prostitutes, beggars, day laborers, street vendors, diplomats, civil servants, rich merchants and academics live side-by-side and rub shoulders as they go about their daily lives.

Personal Context

In Ethiopia, children orphaned by AIDS, war or poverty and children who have disabilities live on the streets alongside children who have been abandoned by, or who have abandoned their families. For this reason, we include orphans and handicapped children into the “on” and “of” street categories. Furthermore, while the main criterion for children “on” and “of” the street has been whether children sleep on the streets or not, later in this discussion we propose to add an additional dimension, “a behavioral indicator,” namely “the level of the children’s engagement in street life,” in accordance with the MOLSA report (1992, 206).

Implications of Diversity for Research

In 1991, UNICEF consultant Peter Taçon estimated that there were about 120,000 street children in Ethiopia, about 20,000 of them “on” the street while the remaining 100,000 were children “of” the street. In 1993, Thomas estimated that one-half million to a million urban poor children were at extremely high risk of becoming street children. In light of their great diversity, it is not surprising that, as in Nairobi and Cali, NGO reports and newspaper articles in Addis Ababa give conflicting numbers, ranging from one hundred thousand to one-half million. It is difficult to carry out a survey that would provide an accurate count, given this variation, the wandering life styles of street children and the changes that occur in their family circumstances with age and work opportunities. However, there are ways of “guesstimating” the number of street children while considering their diversity (Aptekar and Ciano 1999). The rest of this discussion offers suggestions for sampling and data collection procedures.

Sampling

Getting a sample requires knowing the dimensions of the population which, of course, relates to a clear definition of who are being studied.2 Aptekar and Ciano (1999) have reported on establishing a valid procedure for selecting a sample of street children in Nairobi, Kenya. The research team started with a map of the city. Former street children and persons who were working with street children used the map to help identify and mark off areas with high concentrations of street children, and this was then laid over a new map to make numbered grids.

Local people, known by the street children, walked each of these areas every night for two weeks to locate the children’s so-called street households (called choums). These choums were recorded on the city map, which was then divided into equal grids that were labeled according to their high or low concentrations of households. This map was subsequently used for collecting demographic data about the street children.

Working in pairs for greater validity and reliability, researchers collected data on the numbers of street children in each choum in randomly selected areas for three successive nights, and then at weekly intervals for three additional weeks. Each data pair used the same structured interview format to obtain a rough estimate of the numbers of street children, the degree to which they moved from household to household, and the gender ratio (see Aptekar and Ciano 1999 for more information on the questionnaire).

While one researcher asked the child's age, the other made an independent estimate of the child's age. Similar procedures were used for comparing the children's stated tribal affiliation with physical characteristics and language skills. The degree of inter-observer agreement provided an indication of the validity of information the children supplied, and by using this procedure three times, some notion of its reliability.

The resulting census provided the framework for selecting a sample which made it possible to estimate the number of street children and how many of them were using the programs available to them.

The records these programs kept of the children who used their services were used to check which children living in which choums were visiting which programs, and how consistently they visited them. Because their numbers were inflated, there were more than enough programs for the children. The programs grew more by competition than by policy design; as a result of funding being relatively easy to get for food, there were too many feeding programs. In fact, not only were the children not starving, they would discuss which of the several free meals were most appealing.

Demographic Information

Getting information about street children is difficult. This lack of information has adversely affected the formulation and implementation of programs and policies. The most common method of gathering data is to ask the children to complete paper and pencil questionnaires. However, research suggests that it is better not to give questionnaires to street children, and not to trust any answers children give to researcher’s questions, either in ethnographic conversation or in paper and pencil tests that ask them about themselves. As Heinonen has shown (2000), they make their living by manipulating their audiences, of which the data collector and the NGO workers are examples.

It is better to collect both quantitative and qualitative data and to triangulate methods, so that each method offsets some of the inherent problems of the other methods. With qualitative methods, it is possible to listen and allow children to say what is important to them and then to record the information systematically. Using quantitative methods forces the researcher to ask questions in a standardized form.

Perhaps the worst problem that arises in collecting data is caused by not gaining the children’s trust. Aptekar and Giel (2002), for example, reported that data collectors observed by members of the program staff were found to go out in groups of four, carrying with them the papers they needed to administer the protocol. The children knew they were coming and that they would be paid to participate. The researchers did not first become familiar with their living conditions, and because of this they tended to stay by themselves. They all sat together in one tent to collect the data, ate lunch together, and left together at the end of the data-gathering sessions. In fact, the very training they took to prepare themselves for this work made them into a social group which decreased their intermingling with the children. A related problem was that the street children suppressed many types of information, for they often do not want to talk about personal matters when their peers are within earshot.

The street children were paid for the time needed to complete the protocol. Because it took several hours and they were very poor they vied to participate. Since the researchers did not know the children, they were unaware that some children bought the names of others. Mostly children sold their names because they did not trust that they were really going to get the money that the researchers promised them.

Once the children sat for the interview, there was no way to check on the validity of their answers. Many subjects rushed through the questions so as to get paid. They were, after all, much more interested in getting the money than telling some person they did not know the accurate details of their awful experiences. In some cases they just lied, making up complete stories about themselves, including past traumas, because they thought that their fictitious circumstances would bring them more aid.

Many of these research problems could have been avoided by having the data collectors and the people in charge of the study working more closely with the children, but their paradigm is common. Someone plans research with street children in a far-off location and people who have their own agendas collect the data, without knowing whether the information is ever validated.

The above methodological problems have contributed to widely varying information about street children. For example, it has been reported that street children have little contact with their families (Aptekar 1992; 1997; 2000; Aptekar and Abebe 1997). However, other, more careful studies have shown that this is not the case. Among a sample of street and working children in Addis Ababa, at least 95 percent of the total had regular contacts with their families. The remaining five percent were groups of homeless boys and a few girls sleeping rough in roadside ditches, pavements, shop verandas or outside church walls (Ministry of Labor and Social Affairs 1988; 1992; 1993; Wondimu 1996). This demographic information reflects the reality found by the long-term study in Addis Ababa undertaken by Heinonen (2000) who also found that a majority of street children were born and reared in the city, although increasing numbers of migrant children are brought to the city by internally displaced parents due to wars, social conflict and famine situations. A significant number of women migrate to the city to escape either rural poverty or abusive marital/familial situations; many bring their children with them. None of the rural-born mothers interviewed by Heinonen throughout her five years of field study kept contact with family back home.

Of the 32 homeless boys and girls in the Heinonen study (2000), 26 had run away from home in order to escape the cruel treatment they received from their stepmothers, step-sisters, step-brothers or older siblings, or the neglect and abuse they had suffered from their fathers. In short, the simple dichotomy of “of” and “on” the streets no longer applies. In order to understand how the children live it is necessary to know more about their family circumstances and social world than simply establishing where they sleep.

A study of children in Kaliti, a refugee camp for the displaced outside of Addis Ababa, further illustrates the much greater complexity behind the simplistic dichotomy “on” and “of” the streets. Many of the Kaliti children have become street children, yet their histories are very different from those of many street children in Cali or Nairobi (Aptekar 2003; Aptekar, Paardekooper and Kuebeli 2000).

The saga of the children of Kaliti began in 1991, when the 30-year civil war between the forces of Ethiopia and Eritrea temporarily ended (it has since resumed, ended, and resumed again). Eritrea then gained independence from Ethiopia, and the Ethiopians living in the new territory were forced to leave. Because the government considered women married to Ethiopians still Eritreans, these families faced a terrible dilemma: they could stay in Eritrea with their families of origin, which meant saying good-bye to their husbands and children who were considered enemies, or they could join their husbands and children on the long uncharted march to Ethiopia through the Danakil Depression. In the Danakil Depression there is no vegetation or water, and temperatures reach 140 degrees Fahrenheit in the shade. The journey was worse than everyone feared. For those who walked, the important thing was to make progress, no matter how slow. Before the middle of the day the marchers shuffled and suffered in silence. They tried not to look at the old, the infirm, or the small children whose parents could no longer carry them, because they knew they could not help them, nor watch their agony, nor participate in their death. Many of the survivors, who were young children at the time, are now living on the streets in Addis Ababa. Are they victims of war, or street children, or both? Their problems highlight the shortcomings of UNICEF's definition of street children, as explained below.

At the end of their desert ordeal, 160 males in Kaliti were categorized as war orphans and placed under a special program for unaccompanied children. The aid agencies classified young men without parents as unaccompanied minors and orphans, and as such a special category for assistance. However, there was no easily discernable category for young females without parents. They were seen to belong to a larger category of single women in charge of families, even if they had no children.
While the benefit scheme seemed to favor the boys, their being considered “orphans” actually pushed them towards the streets. The assistance they received resulted in encouraging the boys to stay apart from the community because they needed their special orphan status to survive. As they grew from teenagers to young men, they hesitated to start their own families because an adult status would have forced them to relinquish the financial resources they received as unaccompanied children. Humanitarian assistance had been relatively easy for them to obtain but when it ended, developmental assistance was harder to secure. Many of them were forced to migrate yet again. In this case, many entered the streets. Their path of development brings out the complex realities surrounding attempts to arrive at an agreed-upon definition of street children.

Historical Changes and the Emergence of New Categories

The reasons why children go to the streets have changed over time, but historical development has also changed the conditions of life on the streets, even in the cases where the same reasons for going to the streets prevail.

The situation of street children studied in Colombia in 1983-4 has undergone significant changes (Aptekar 1988; 2000). The most obvious of these changes is that the children have become engaged in battles between the narcotraficantes, the government, and the private paramilitary organizations. How they cope with these competing demands is not well understood, but it is clear they are not the same circumstances that Colombian street children faced nearly two decades ago. At that time they were much more like the abandoning or abandoned children referred to as “in” and “on” the streets.

To address some of the many contextual problems that have arisen in defining street children, Heinonen (2000) presents a more detailed classification for the street children in Addis Ababa. It includes differences between school attendance, street related activities such as begging or working, age and gender differences, and family dynamics. Her classification falls into three broad categories: street working children, working children, and street children.

Street working children are children who live at home, attend school part time, and work or trade in the street the rest of the time. These children have extensive economic and affective ties with their families, especially their mothers who help them with shelter, education and health.

Working children include children living at home, aged eight and older, who do not beg, do not attend school and work full time in the street. Both working children and street working children play a vital role in the economic survival of their families. They nevertheless depend heavily on the presence of adults (usually their mothers) for their sustenance at home and to some extent for their career in the street.

Street children are aged five to eighteen, do not attend school, and beg full time in the street. Most of those under eight years of age alternate between home and street life. Once they reach ten, most join loosely knit social groups of same-sex or mixed-gender gangs. The criterion differentiating them from home-based working children is their lack of family ties and therefore support in the form of nurture, shelter, health care and education. In the Heinonen (2000) study, all the homeless street children knew the whereabouts of their families, but had deliberately chosen either not to contact them or to regulate the type and amount of contact they had with their relatives.

In contrast to Gurague street working children, the general public in Addis Ababa perceives other street children and street working children collecting alms as psychopathological, drug-dependent, criminal, or inclined to deviant behaviors such as theft and vandalism. Street working children, whether Gurague or non-Gurague, engage in a wide variety of income-generating activities such as shoe shining, working with taxi drivers to help passengers being collected and dropped off, and selling lottery numbers, cigarettes, candies, ground nuts, etc. Street children are mostly engaged in begging, to which they refer as work.

Working children are defined by their behavioral activities and are perceived as being good children who work in order to help their families. Street children, on the other hand, are usually defined by negative descriptors. Swart (1990) and Glauser (1990) both found that, without any evidence, street children are commonly considered to be a danger to the community. The core concept of street children incorporates public fear, which is fuelled by the media’s self-serving exaggeration and selective reporting. It is also biased by international organizations that compete for resources as well as by the educated urban populace who have adopted a global concept of childhood based on western standards. This has led to claims that there is a separate sub-culture of street children, although this perception is more likely to be a creation of the public’s fears than to reflect the true direction and context the children’s lives. (See also Beazley 2003).

When the public uses the term “street children” it does not refer only to the relatively small number of children and youths for whom the city streets are home but to the combination of children found on the streets. It is colored by a perception (accurate or not) that street children are behaviorally deviant with regard to drug abuse, thieving and violent conduct towards other people.

The term used for street children in Addis Ababa reinforces a negative image. They are collectively referred to as Borco (an adulteration of the Italian words sporco, meaning filthy/dirty or porco, meaning pig). An Ethiopian anthropologist has translated Borco to mean: “One who lives in the streets and is unable to return to a normal way of life” (Tessema 1998, v). This interpretation encapsulates a generalized perspective. Most people also use the words Godana (an abbreviation of Godana Tedadari, meaning street children) or Dura (which is short for Duriye, meaning vagabond). Other poor children share the community’s perception of street children as being bad.

Unlike home-based working street children, homeless street boys and girls in Addis Ababa, particularly after the age of ten, live in loose-knit groups, often, but not always, of the same gender. Despite this, the children do not, on the whole, derive emotional or social support from belonging to a group. Heinonen (2000) found that the children in her study expressed a desire for autonomy coupled with a need to feel part of the group. These conflicting desires, exacerbated by the constant violence between and within groups, precluded the formation of enduring affective emotional attachments and therefore a sense of responsibility for one another.

The Borcos, much like the Malunde in South Africa (Swart 1990) shared the values and abided by the norms sanctioned by mainstream society. In many other aspects, Heinonen’s findings were different from many other studies of street children, which have found that street children gain certain benefits from living together. For the most part, group life for the Borcos did not provide individual children with emotional support and affective relationships. Exceptions occurred during illness or bereavement. Situations of unconditional and altruistic support by group members of injured comrades also occurred among the Malunde (Swart 1990). Unlike the Borcos, the strollers of Cape Town in South Africa “continually reinforced each other in their survival activities, in their joys and fears, and in the majority of cases, they shared their earnings equally among themselves” (Scharf et al. 1986). Similarly, relationships between Colombian street children living in a group were characterized by a sense of responsibility towards one another and lasting emotional bonds (Tyler 1997).

Research on street children in Addis Ababa also shows how the context of childhood there stresses dependence rather than independence and compliance instead of arbitration (Heinonen 1996; 2000). Dependency, blind obedience and total deference to parents and all adults are all associated with the ideal child and childhood in Ethiopia. Any infringement of the rules of appropriate social behavior is met with a severe beating or verbal abuse, known as tefat ena ketat (misdemeanor and punishment). However, cultural and ecological context influence how children are socialized into acquiring social roles considered appropriate within the societies to which they belong. The squalid and overcrowded urban environment homogenizes the manner in which destitute mothers socialize their children. Family setting has a strong influence on the type and amount of domestic violence Ethiopian children suffer at the hand of their parents (Heinonen 1996).

The comparative and contextual model suggested in this paper highlights an important fact: once one gets close enough to street children, superficial notions of how they relate to each other often break down. This fact should inform policy and programmatic approaches. It has also important methodological implications.

Gender Differences

The street and working children of Addis Ababa and elsewhere cannot be understood without considering gender differences (see Aptekar and Ciano 1999). The gender ratio for working street children in Ethiopia is approximately four boys to one girl, especially among those aged nine years and over (Wondimu 1996). In Addis Ababa (as in most cultures) most (but not all) street boys are taught by their mothers to cope with the necessity of having to make do in a very limited economic environment by becoming independent at a far earlier age than the dominant society deems appropriate. When compared to other poor boys and to the other boys in the same family, Kenyan street boys were more resilient (Aptekar and Ciano 1999). Less resilient boys were unable to leave home and help their families who were living in extreme poverty. The opposite situation held true for Nairobi street girls. Poor mothers taught their girls how to cope with the vagaries of poverty by staying at home and off the streets. Thus, street girls, for the most part, were found to be more psychopathological than their sisters who stayed at home.

Heinonen (1996; 2000) examined the gendered aspect of child rearing among street children’s families. She established how many boys and girls are socialized only by their mothers, thus revealing an adult-on-child as well as a child-on-adult aspect in socialization. All the street children and parents who were interviewed maintained that raising and socializing children is a woman’s job. A father’s role in bringing up children was usually restricted to guiding and disciplining troublesome children, especially boys, when mothers were unable to cope. Girls were supposed to be guided, disciplined and trained into womanhood by their mothers or other female members of their families. Yebetmoya is the term for the domestic skills imparted to children (usually girls) by female members of the household so as to equip them for adulthood and motherhood. The notions of yebetmoya and of tefat ena ketat embody the two main aspects of socialization regarding child rearing in the domestic sphere.

All the street boys in Heinonen study (1996) had relatively independent financial arrangements with their parents, especially their mothers. The type of work opportunities available to them meant that they were able to operate more or less independently from their parents. Their mothers had no knowledge of the amount of money they earned or how they spent it but most boys gave their mothers a portion of their earnings. As the boys grew older, most mothers accepted their much-reduced parental authority or risked being abandoned by their sons.

Except in rare cases, a street girl’s financial arrangements in Addis Ababa were inextricably tied up with that of her mother. Girls worked for their mothers who expected them to hand over the entire proceeds of the day. All income from daughters was considered “family income” to be disposed off by the mother as she deemed fit. There was often a conflict of interest and friction between mothers and daughters, resulting in extensive abuse of girls by their mothers. Boys, on the other hand, faced a higher degree of violence in the street by the police, other street children and the public than did the girls.

Because boys are able to control their own earnings and girls are not, this introduces a further gendered aspect of street life. Boys purchase a great variety of nutritious foods such as peanuts, bread, bananas, boiled eggs or potatoes for as little as ten to 25 cents. They are also considered creditworthy by most street vendors and teahouse owners. They are thus better fed and in better health than generally expected (Yemane and Yemane 1998).

Aptekar and Ciano (1999) illustrate the above point through the case study of Pleasant, a Kenyan mother of a street child. Pleasant was 28 years of age and had been married by common law to a night watchman for five years. Together they had four children (three boys and a girl) whom they supported until about two years ago, when he, like many other men in his culture, began to drift away from his family. To Pleasant this did not come as a surprise: “This is what most men do.” By the time he left completely, she had developed strong ties to other women in her neighborhood whose husbands had also left them: “This is what women do.” These women helped each other with food, access to medical care, and other necessities.

By the time Pleasant’s eldest boy, Mbisa, had his sixth birthday, he was accustomed to playing in the streets with older boys in the neighborhood. Mbisa’s mother rarely supervised him and he learned early to fend for himself. After his father left and the household income dropped, Mbisa began to drift further from his home and go into other neighborhoods to park cars, clean windows, and find income which he brought home, to his mother’s great delight. Pleasant worked on and off as a domestic worker, and showed her oldest daughter, Dominion, how to take care of household chores. By the time Dominion was seven years of age she would fetch water, make fires, and cook most meals.

While Pleasant and Mbisa were bringing home income, it was possible to occasionally pay the school fees for the two younger boys. Then Dmisa, a man whom Pleasant had known previously, moved into the home. Their combined incomes kept the two younger boys in school for longer periods of time and even allowed Mbisa to return to school.

Pleasant understood the vicissitudes of a woman’s economic and romantic situation. She was aware that eventually her boyfriend would move out, or that she would kick him out and she knew what the economic implications of these changes would be. Without additional family income the two younger boys would have to leave school and go to the streets like Mbisa to find income. As she said, only complete financial destitution or the utter demise of her mental health would lead her to send Dominion to the streets. (Some women did have daughters begin work in the streets, but an older child supervised the daughters and the mother would make every effort to see that her daughter was not abused.) By understanding how women like Pleasant, in the context of their positions as the heads of impoverished households, cope with poverty and with the men they live with, it is possible to understand the mental health of street boys and girls.

The family dynamics of the genders are substantially different. At 12, 13, or 14 years of age, a boy's body image changes to that of an adult. The public then no longer perceives boys as cute and worthy of pity, but instead they are looked upon as dangerous. This public perception forces the boys who beg, clean car windows, help to park cars, or look after parked cars, into the same kind of menial work as other poor adult males. This includes work for trade, piece-meal work, or intermittent salaried employment.

Girls begin street life much later than boys, usually after they are ten years of age. Even though they may appear to be alone, an older sibling often supervises them. As girls became pubescent they are perceived and evaluated in sexual terms. By the time they are young women, they often follow in their mother's footsteps by having children, often many and by different men, who as a rule do not view them as legitimate wives, and thus not worthy of continued financial support.

Because boys are expected to bring income home, and thus go to the streets, while girls are expected to stay at home and help out with the household chores, the street boys and street girls relate to their families of origin differently. It is common for street boys to remain connected to their mothers; indeed they often contribute part of their incomes to them. However when girls are on the streets and not in the home, they often have more difficult and distant relationships with their families of origin.

Street boys are commonly on the street because they have been brought up to be independent while street girls are on the streets because they are fleeing a very difficult situation. Their mental health is therefore frequently considerably worse than that of the boys. Many have developed adequate coping strategies, which allow them to function at least as well as their poor counterparts who pass less time in public view. These coping strategies include finding a niche in the economic market, which gives them sufficient income to eat and clothe themselves. They are also able to find and take advantage of programs that serve them, become sufficiently informed about their physical health to stay reasonably healthy, form close friendships with peers, and maintain some form of connection to their family of origin.

When researchers refer to “street children” without gender differences, important distinctions between boys and girls get blurred.

Overcoming Biases in Data Collection

Data collectors often over-romanticize the children which creates a bias toward resilience. Others over-dramatize the children thus making them appear worse off than they really are. Based on our past research (Aptekar 1988; 2000; Heinonen 2000), we believe that developing a good rapport and collecting long-term data from street children will reduce the distortion of facts about their lives and social world. Sometimes the relationship leads to researchers becoming involved in the plight of the children as well as their day-to-day relationships with peers and relatives.

In circumstances where it is impossible to establish long-term rapport with the children, in order to avoid distortions, two or more persons can collect the same or similar data. This makes it possible to compare the children’s answers to establish a degree of internal validity through inter-observer reliability measures.

Another possible bias derives from the context of the data, which might be collected on the streets, in the child’s home, with or without parents present, etc., all of which can influence the data. It is sound practice to ask the same questions more than once in more than one location to check for variation in answers. It is also important to know whether or not a child is being paid to give answers, or being forced to give answers in return for placement in a program or to obtain services.

Reducing Children's Distortions

By increasing the amount of data that is performance-related or projective, many of the problems of children's distortion of information can be circumvented. Examples include children making drawings of their families or themselves or of their daily routes through the city, or children reproducing figures taken from standardized tests like the Bender-Gestalt. Projective techniques, such as incomplete sentences and the three wishes test can also be used. Other performance data like sociograms of the children’s social networks, drawings of their family trees or of the important people in their lives, or of their daily life routines are also helpful. Children can depict their daily life through mural type drawings. The researcher can arrange cards that force the children to make Q sorts showing their preferences or their fears in ordinal amounts. They can do sentence completion quizzes orally.

Additionally, photos can be taken from magazines and used in projective techniques. For instance, the children can use the photos to explain relationships between the people they are seeing, or the wishes people have for their future. They can be asked to build their own life stories by cutting out pictures from magazines.

Scorers can be trained until they have developed an appropriate level of internal reliability. Methods can be used more than once with the same child or comparable children and in these ways it is possible to obtain numerical data.

It is possible to gather quantitative data by recording how many children visit a certain program or how many pass a certain point in a given period of time. Using random time sampling when collecting ethnographic data- collecting data both during the day and at night, during the week and on week ends, and in all kinds of weather- helps to reduce the bias of otherwise time-skewed data. Random medical examinations of the children can be used to determine their nutritional level as well as their abuse of drugs.

The Value of Assessing Mental Status

A much-overlooked and extremely important method of garnering accurate psychological data is by means of the mental status exam (MSE). Essentially it is the psychological equivalent of a physician's physical examination. The MSE includes an assessment of the child's current functioning including memory, thought processes, language abilities, motor behavior, feelings, and judgment. It also provides information about the child's attitudes toward other people and their appearance. Importantly, the skills needed to administer a MSE are quickly taught to non-professionals.

As soon as a researcher makes contact with a street child, the MSE should begin. The researcher can observe children's appearances: Are they clean or dirty? Are there any wounds; are they cared for or untreated; are there ticks or tremors? Are there any unusual facial distortions or movements, or any stereotypical or repetitive movements or mannerisms? Do the children pace or show other signs of agitation, or are their movements very slow and labored? Is the amount of eye contact appropriate? What is their attitude toward strangers and people they know? The answers to each of these questions help to determine how children are taking care of themselves.

The same can be said of questions that are more related to cognitive skills. What is the child's speech like; is the flow of speech slow or pressured; is the tone quality unusual; is the volume appropriate? What is the level of vocabulary; does the child have a problem finding words, or in misusing words, or finding the names of well-known objects?

There are other things to observe such as the child's mood and, over time, the range of moods. The researcher can ask about personal goals to assess the level of realism children have about themselves as well as the degree to which they are insightful about their problems, and what kind of judgments they are making about their lives. Certain games can be played that will help to ascertain their level of simple arithmetic, or their memory.

The MSE, when used properly, gives the researcher important information about children's abuse of drugs, as well as the degree, and to some extent the type of psychopathology. It also provides important information about intellectual and cognitive capabilities.

It is less useful to probe for factual information about personal histories such as their school attendance than it is to use questions designed to establish their intellectual capacities, for instance by asking them to interpret local proverbs, or to name the President. During conversations it is also possible to note their use of logic, their clarity of thought and their ability to articulate their opinions.

Data from Other Sources

This paper has focused on children, but there exist other useful data sources. It is particularly important to understand society's attitude toward street children and their historical context. Relevant information can be gleaned from archival data such as press reports and other written documents such as government and NGO sources, and from civil and criminal laws. It is valuable to see the children in the context of how they are presented by the media. It is also important to follow the various events for street children as well as to gather data during events that the children themselves organize.

One way to obtain data from many people is through focus groups. These might encompass parents, or those who worked with the children, or those whose only knowledge of the children is gleaned from the popular press. Themes for focus groups can be chosen from concerns the participants voice when they meet or from the issues brought up by street children or the research questions. If street children are the participants, nonverbal data can be obtained from drawings or by theatrical presentations. (See the Field Report by Salami 2003).

Caution in Reporting Research

Two important problems are widespread. First, authors frequently over-generalize their findings by moving from data collected on a small and often non-random sample to conclusions that suggest a much larger population.
Secondly, there are problems with reifying numbers presented in studies of street children. No matter how much authors qualify the validity of their findings, readers tend to take them as incontrovertible facts and perpetuate them even if only by using them in literature reviews. Quoting numbers gives an impression of validating research and makes readers feel that research is “scientific.” However, if numbers are inaccurate they may serve the opposite of scientific inquiry, and promote prejudice over truth.

Conclusion

Almost two decades of research and work with street children have shown that there are a number of common factors that can be identified when one speaks of street children. All street children regard their form of obtaining income as “work”. Most defend their right to work. Secondly, the range of work that is possible on the streets is limited. There does not seem to be much variation from one city to another. Finally, most street children have made a conscious decision to be on the streets, whether as working, home-living children, or as working, street-living children. They have chosen, therefore, to be agents of their own destinies.

Just as importantly, research and work with a wide range of street children shows that the situational context in which the children are found definitely affects their experiences on the streets. Although, as described earlier in this article, almost all children found on the streets are treated with contempt, the Brazilian extermination squads described by Rizzini and Butler (elsewhere in this issue) are not found in places like Addis Ababa. Research not only into the children’s backgrounds and into how they experience street life is important but it is perhaps even more important to understand how each society treats their street children. All are necessary for program design.

There are many reasons for benefactors to offer assistance, but many do not understand what the street children need most. The result of this means that while many people and programs are well meaning they are not always helpful.

Reliable data collection with street children is difficult because of their fluid lifestyle, the difficulty of defining a research population and the problem of retaining contact with individual respondents. Payment for information—whether in food or money—tends to generate false data. Information is best gathered over a period of time, and replicated by independent observers to assure accuracy.

Endnotes
1. In fact low-income households constitute, by far, the bulk of Addis Ababa’s population. According to a 1996 study, only 3.9 percent of the housing units in Addis Ababa have flush toilets, 16.7 percent have shared pit latrines, 20 percent have private dry pit latrines, and 59.3 percent have no facility whatsoever (Tegegne and Daniel 1997).
2. The absence of a clear definition is only one problem. For several reasons there is a bias towards inflating the numbers of street children. The higher the number, the larger the problem appears; the larger the problem can be shown to be, the more donors can be pressed to contribute funds. If funds are flowing there are more programs for street children, and each program employs people who have an obvious bias toward inflating the numbers. Finally, the press is biased towards increasing the numbers, in addition to reporting other worst-case scenarios- identifying the youngest child on the street, the most drug dependent, the gravest delinquent acts, etc.- because this is what sells newspapers. There is a field of research here that falls under the rubric of the sociology of information or the sociology of knowledge.

Lewis Aptekar received his Ph.D. in Educational Psychology from the University of Michigan. He is a licensed clinical psychologist and a professor of Counselor Education at San Jose State University. His cross-cultural research has been supported by various awards, including two Fulbright scholarships, and has led to numerous publications on children in difficult circumstances around the world.

Paula Heinonen is of Italian/ Ethiopian parentage and was born and brought up in Addis Ababa. From 1994 to 2001 she was a lecturer in the Department of Sociology and Anthropology as well as Head of Research at the Centre for Research and Training in Women and Development. She is currently the Development Officer at the International Gender Studies Centre of the University of Oxford.

References

Aptekar, Lewis (1988). Street Children of Cali. Durham: Duke University Press.

Aptekar, Lewis (1992). "Are Colombian Street Children Neglected? The Contributions of Ethnographic and Ethnohistorical Approaches to the Study of Children." Anthropology and Education Quarterly 22(4): 326-349.

Aptekar, Lewis (1997). "Street Children in Nairobi, Kenya: Gender Differences and Mental Health." Journal of Psychology in Africa 2: 34-53.

Aptekar, Lewis (2000). A World View of Street Children in the Year 2000. Program for Street Children Symposium, Jyvaskyla, Finland.

Aptekar, Lewis (2003). "Cultural Problems for Western Counselors Working with Ethiopian Refugees." In Bemak, F. et al., eds. Counseling Refugees: A Psychosocial Approach to Innovative Multicultural Innovations. Westport: Greenwood Press, 208-225.

Aptekar, Lewis and B. Abebe (1997). "Conflict in the Neighborhood: Street Children and the Public Space." Childhood 4(4): 477-490.

Aptekar, Lewis and L. Ciano (1999). "Street Children in Nairobi, Kenya: Gender Differences and Mental Health." In Rafaelli, M. and R. Larson, eds. Developmental Issues among Homeless and Working Street Youth: New Directions in Childhood Development. San Francisco: Jossey Bass, 35-46.

Aptekar, Lewis and R. Giel (2002). "Walks in Kaliti Life in a Destitute Shelter for the Displaced." In Jong, Joop de, ed. Trauma, War, and Violence: Public Mental Health in Socio-Cultural Context. New York: Kluwer Academic/Plenum Pub, 337-366.

Aptekar, Lewis, B. Paardekooper and J. Kuebli (2000). "Adolescence and Youth among Displaced Ethiopians: A Case Study in Kaliti Camp." International Journal of Group Tensions 29(1-2): 101-135.

Barker, R. and J. Knaul (1991). Exploited Entrepreneurs: Street and Working Children in Developing Countries. 1. Childhope.

Beazley, Harriot (2003). "The Construction and Protection of Individual and Collective Identities by Street Children and Youth in Indonesia." Children, Youth and Environments 13(1). Retrieved [date] from http://colorado.edu/journals/cye.

Cosgrove, J. G. (1990). "Towards a Working Definition of Street Children." International Social Work 33: 185-192.

Ennew, Judith (1996). "Difficult Circumstances: Some Reflections on Street Children in Africa." Africa Insight 26(3).

Glauser, B. (1990). "Street Children: Deconstructing a Construct." In James, Allison and Alan Prout, eds. Constructing and Reconstructing Childhood: Contemporary Issues in the Sociological Study of Childhood. London: Falmer Press.

Heinonen, Paola (1996). Some Aspects of Child Rearing Practices in the Urban Setting of Addis Ababa (with Special Reference to Street Children). Addis Ababa.

Heinonen, Paola (2000). Anthropology of Street Children in Addis Ababa, Ethiopia. Unpublished Ph.D. Dissertation. University of Durham, Durham, U. K.

Lusk, M. (1992). "Street Children of Rio De Janeiro." International Social Work 35: 293-305.

Ministry of Labour and Social Affairs (1988). Survey on Street Children in Addis Ababa. Radda Barnen (Sweden).

Ministry of Labour and Social Affairs (1993). Proceedings of the National Seminar on the Problem of Begging in Addis Ababa.

Ministry of Labour and Social Affairs (1993a). Study on Street Children in Four Selected Towns in Ethiopia. Cork: UNICEF.

Ministry of Works and Urban Development (1996). Addis Ababa. Addis Ababa, Ethiopia.

Salami, Irene and Henk van Beers (2003). "Nigerian 'Shade Tree Theatre' with Street Children." Children, Youth and Environments 13(1): Retrieved [date] from : http://colorado.edu/journals/cye.

Scharf, W., M. Powel and E. Thomas (1986). "Stroller-Street Children of Cape Town." Growing up in a Divided Society: Northwestern University Press.

Seifu, Ruga (1976). "Molding Traditional Self-Help Associations to Meet Modern Demands: A Case Study." In Ottoway, M., ed. Urbanization in Ethiopia. Addis Ababa: A. A. University.

Swart, Jill (1989). Community and Self Perception of the Black South African Street Child. Symposium on Theory and Practice: Street Children in the Third World, April 20-21, University of Amsterdam.

Swart, Jill (1990). Malunde: The Street Children of Hillbrow: Witwatersrand University Press.

Taçon, P. (1991). Protection, Respect and Opportunity for the Street Children of Ethiopia. Unpublished Report. New York: UNICEF.

Tegegne, G. and S. Daniel (1997). Proceedings of the National Conference on Urban and Regional Development Planning and Implementation in Ethiopia. Addis Ababa: National Planning Institute.

Tessema, E. (1998). Urban Adaptation and Survival Strategies. The Case of Internally Displaced Groups in the Arada Area of Addis Ababa. M. A. Thesis in Anthropology. Addis Ababa University.

Thomas, M. (1993). Children and Women in Ethiopia. Addis Ababa: TGE & UNICEF.

Tyler, F. B. (1997). Urban Settings, Youth Violence and Prosocial Communities. Urban Childhood Conference, June 9-11, Trondheim, Norway.

UNICEF (1993). Children and Women in Ethiopia: A Situation Report. Addis Ababa: Transitional Government of Ethiopia.

Wondimu, H., ed. (1996). Research Papers on the Situation of Children and Adolescents in Ethiopia. Addis Ababa: Addis Ababa University.

Yemane, Beyene and Berhane Yemane (1988). "Health and Social Problems of Street Children." Ethiopian Journal of Health Development 12(1, April).