Our Sample and Survey Methods
Young Lives is investigating the changing nature of children’s lives over 15 years from 2000 to 2015, which is the timeframe set by the UN to assess progress towards the Millennium Development Goals (MDGs).
We are following two groups of children in each of the four countries: 2,000 children who were born in 2001-02 and 1,000 children who were born in 1994-95. These groups provide insights into every phase of childhood. The younger children are being tracked from infancy to their mid-teens, and the older children through into adulthood, when some will become parents themselves.
When this is matched with information gathered about their parents, we will be able to reveal much about the intergenerational transfer of poverty, how families on the margins move in and out of poverty, and the policies that can make a concrete difference to their lives.
Our research is designed to cover three broad areas:
- What are the factors that shape children’s lives to either increase or reduce poverty and its effects?
- What effects does poverty have on children’s lives and experiences during childhood and into adulthood?
- To what extent are current international and national policies effective in reducing childhood poverty in the study countries?
Young Lives uses a ‘sentinel site’ methodology, borrowed from health surveillance studies, which is a form of purposeful sampling where the sentinel site is deemed to represent a certain type of population. This produces a broadly (but not strictly statistically) representative picture. The sentinel methodology is an important way to ensure a study is able to keep in contact with people over long periods of time (the core benefit of a cohort study).
When Young Lives was set up in 2001, Andhra Pradesh was chosen because it had been the role model for several new government initiatives to eliminate poverty in the 1990s and there were concerns about the levels of child work within the state. With the bifurcation of the state in June 2014, we now also have study sites in the newly formed state of Telangana.
A sentinel site was defined as a mandal. At the time, Andhra Pradesh was divided into 23 administrative districts, each sub-divided into a number of mandals depending on district size. There were 1,125 mandals and about 27,000 villages in Andhra Pradesh.
For the purposes of our long-term panel survey, districts were ranked according to their relative levels of development based on economic, human development and infrastructure indicators. A mixture of poor and non-poor, rural and urban sites were selected across Andhra Pradesh’s three distinct geographical regions – Coastal Andhra, Rayalseema and Telangana, plus the city of Hyderabad. Together, these areas offer examples of different geographical regions, levels of development, urban-rural balance and population characteristics (including ethnicity) across Andhra Pradesh and Telangana. We carry out our research at different sites in the districts of Anantapur, YSR Kadapa, Srikakulam and West Godavari in Andhra Pradesh. In Telangana our study sites are in Karimnagar and Mahbubnagar. We also have a study site in Hyderabad.
When we compared our sample to the Demographic and Health Survey (DHS) (1998/9) to assess its representativeness, we found that while our sample typically includes households slightly wealthier than the average Andhra Pradesh household, and with better access to basic services and more ownership of assets. Differences could be partly accounted for by the earlier collection year of the DHS. Despite this, our sample covers the diversity of children in Andhra Pradesh and Telangana and our findings represent a wide variety of children in terms of wealth, consumption, health, nutrition and education similar to national datasets.
We are carrying out multidisciplinary and interdisciplinary research using a mix of qualitative and quantitative methods, including a detailed survey of all the children, group work and participatory activities with a smaller group of some children, and policy and budget monitoring and analysis.
Household and child survey
We carry out a detailed, comprehensive survey of all the Young Lives children (younger and older cohort) and their primary caregiver every three years. The first survey round was conducted in 2002, and the second round in late 2006/early 2007. The third round was conducted from late 2009 to early 2010. The next rounds will take place in 2013 and 2016.
The surveys use a questionnaire for each child and his or her primary caregiver. We also talk to community leaders to understand more about local resources and services and changes happening in the local area.
The questions cover material well-being, physical health, education and cognitive development, perceptions of wealth and general well-being, psychosocial health, time use and activities, risk and vulnerability, and social and political capital.
In-depth sub-sample research
We carry out more in-depth work on particular aspects of children’s lives with a smaller sample of 48 children in four sites (one urban, two rural and one tribal). The sample also reflected equal number of older and younger age groups (cohorts) and boys and girls. This involves a mixture of group-based activities, individual interviews and observational work exploring their local environment. Three rounds of data were collected in 2007, 2008 and 2011 as part of the qualitative research.
We talk to the important adults in each child’s life, including their main carer, teachers and local community figures, both individually and in groups. So far the major focus of our qualitative research has been on early childhood transitions, varying trajectories into school, managing transitions in early childhood education, and children’s perspectives on well-being.