Data

Starting in 2003, the CDC began conducting the National Survey of Children’s Health (NSCH). In 2011 (the most recent year available), they completed a total of 95,677 child-level interviews. The survey is a cross-sectional telephone survey of households with at least one child age 0-17. They stratify the samples by state and type (landline vs. cell phone) and cluster them by household (multiple children in the same house are clustered together). In English, they do a bunch of fancy statistics that allow us to use this dataset as a nationally representative sample.

A complete list of available variables can be found here. There are two variables related to the position of a child relative to his/her siblings (AGEPOS4 and TOTKIDS4). I created a new variable based on these two variables that indicates whether a child is an only child, youngest, middle, or oldest. I used the survey package provided by Dr. Thomas Lumley to analyze the data in R.

Methods

The effect of confounding variables cannot be ignored in this analysis. For example, let’s say I find that children without siblings are 50% more likely to be covered by health insurance. What if low income families tend to have more children than high income families? That would mean that part of that 50% difference could be explained by the the income of the child’s family, rather than whether or not they have siblings.

I use a logistic regression to control for these confounding variables:

  • Age (in years)
  • Sex
  • Race (black, white, and other)
  • Poverty Ratio (the ratio of a family’s income to the poverty line for that family)
  • Parents’ Education (less than high school, high school, and more than high school)

In short, when I say:

Middle children are 189% more likely to eat ice cream.

what I actually mean is:

Holding age, sex, race, poverty ratio, and parents’ education constant, the odds of a middle child eating ice cream divided by the odds of an oldest, youngest, or only child eating ice cream is 2.89.

Results

Health

Only children are 31% less likely to have health insurance, while oldest children are 34% more likely to have health insurance. Only children are 32% more likely to be clinically obese. Only children are 17% more likely to have asthma, while oldest children are 14% less likely to have asthma. Only children are 50% less likely to have gone to the dentist in the last year. Oldest, middle, and youngest children are 25%, 68%, and 13% more likely to have gone to the dentist in the last year, respectively. Only children are 12% more likely to have missed at least one day of school in the last year due to illness or injury. Oldest children are 36% more likely to have been breast fed, while youngest children are 16% less likely to have been breast fed.

Behavior

Oldest children are 21% less likely to be affectionate and tender with their parents, while youngest children 17% more likely to do the same. Only children are 29% more likely to show interest and curiosity in learning new things, while middle children are 32% less likely to do the same. Middle children are 39% more likely to argue with their parents too much, while only children are 31% less likely to do the same. Oldest children are 30% more likely to bully or be cruel/mean to others, while youngest children are 18% less likely to do the same. Only children are 62% more likely to be unhappy, sad, or depressed, while middle and youngest children are 37% and 15% less likely to by unhappy, sad, or depressed, respectively. Oldest children are 23% more likely to always do their homework, while middle and only children are 17% and 9% less likely to do the same.

Activities

Only children and oldest children are 18% and 45% more likely to have their parents read to them every day, respectively. Youngest children are 24% less likely to have their parents read to them every day. Only children are 13% less likely to have played with other children on 4 or more days in the last week, while middle children are 39% more likely to do the same. Youngest children were 20% less likely to have been on a sports team in the last year, while oldest and middle children are 9% and 16% more likely to do the same, respectively. Only children are 10% less likely to have exercised on at least 5 days in the last week. Oldest children are 12% more likely to spend at least 1 hour per day reading for pleasure, while middle children are 15% less likely to do the same. Only children and youngest children are 29% and 12% more likely to spend at least 2 hours per day in front of a TV, respectively. Oldest children are 21% less likely to do the same.