Study Design


While not initially designed for academic purposes, the Making Connections Survey was conducted in accordance with scientific standards and addresses topics of great interest to researchers in the fields of economics, psychology, public health, public policy, social work, and sociology. These topics include mobility, social capital and networks, neighborhood processes, resident perceptions and participation, economic hardship, the availability and utilization of services and resources, and child and adolescent well-being.

The survey was designed in order to comply with academic standards of design, sampling, and content. The design team included methodologists and researchers from NORC at the University of Chicago (NORC), the Urban Institute, Case Western University, University of North Carolina Chapel Hill and the University of Chicago1. NORC ( conducted the survey. NORC/University of Chicago statisticians designed the sampling and calculated the weights for analytic purposes2. University researchers from Case Western University, University of North Carolina Chapel Hill, and University of Chicago oversaw all aspects of the survey3.

Description of survey

The survey was conducted in Denver, Des Moines, Indianapolis, San Antonio, Seattle (White Center), Hartford, Milwaukee, Oakland, Providence, and Louisville. The baseline survey was fielded in each of the ten Making Connections neighborhoods, and in each county that contained each Making Connections neighborhood. The Wave 2 survey was fielded in the neighborhoods only. Baseline data were gathered between 2002 and 2004. Wave 2 was completed between 2005 and 2007 in the same ten sites. The Wave 3 cycle, scheduled between 2008 and 2011, was conducted in seven of the ten sites.

To view the boundaries of the neighborhoods included in each site, click on the files listed below.

The county-wide telephone survey was conducted for nearly 700 households in each county at baseline in order to understand the context relative to the neighborhoods that were chosen. Each wave of the neighborhood survey consists of roughly 800 completed interviews divided between families with and without children. The instrument was translated into the languages spoken by residents that comprise at least 10% of the population in a particular neighborhood; in addition to English, the survey was conducted in Spanish, Vietnamese, Cantonese, and Hmong. Response rates for the neighborhood samples in Wave 1 varied from 63 to 78%, in Wave 2 from 74 to 83%, and in Wave 3 from 77% to 87%.

The neighborhood survey was designed so that a longitudinal sample was maintained over the three waves, and a representative point-in-time sample was also collected. Families with children that left the neighborhoods were followed between waves in order to achieve an understanding of who moved and why.

Additional details about the sample design are provided in this document:
MkCon Sample Design.pdf.

Policy Relevance

The diversity of the ten sites means they offer good examples of the wide range of challenges being faced by local leaders and other policymakers as they try to make headway in improving today’s economically disadvantaged communities. The stereotypical declining neighborhoods of the survey’s older industrial cities (e.g., Louisville, Milwaukee, Indianapolis) remain among the most critical, but they can no longer be said to fully represent America’s “urban problem.” There are other poor neighborhoods in the East and Midwest that have similar challenges but where, in addition, expanding immigrant populations are shifting the traditional dynamic (e.g., Des Moines, Hartford, Providence). And yet other troubled neighborhoods in other regions operate differently, ranging from fairly stable Hispanic communities with severe persistent poverty (e.g., San Antonio) to rapidly growing, racially diverse neighborhoods where extraordinary housing affordability pressures are overlaid on the more traditional barriers to family stability (e.g., Denver, Oakland, Seattle).

Research using the survey data should offer valuable guidance, not only for community development practitioners and neighborhood groups in the field, but also for state and federal officials whose job it is to define and develop guidelines about best practices for community policy.

Unique features of the data

In addition to the topics and characteristics of the survey listed above, some features make the data different from many other surveys of this size:

  • During the initial neighborhood waves, respondents were asked to draw their neighborhood on a map.
  • While only one child was the topic of the first survey, data was collected on all children in the household during the second and third waves. This allows for greater understanding of the characteristics of entire families.
  • Data was collected about the schools children attended, including the name of the school.
  • The questionnaire includes school readiness items for each young child.
  • Data regarding youth transition into adulthood (aging out) outcomes were collected.
  • In wave 3, respondents were asked to give permission to allow the potential linkage of their survey data to federal unemployment insurance wage data.

1 Claudia Coulton, Bob Goerge, Catherine Haggerty, Lisa Lee, Tom Kingsley, Kenneth Rasinski, Marge Turner, Colm O’Muircheartaigh, Lynn Usher, Mary Winkler
Colm O’Muircheartaigh, Ned English
Claudia Colton, Lynn Usher, Bob Goerge