Skip to main content

How To Test for Significant Differences Using the SHADAC Data Center: New Brief

SHADAC Staff
December 08, 2016

The SHADAC Data Center not only provides state-level estimates for health care access, health care utilization, and coverage indicators for several federal surveys, but it also provides information about the margins of error (MOEs) for these estimates. 

A new technical brief from SHADAC explains how to use these MOEs to calculate whether there is a statistically significant difference between two estimates from state to state, from year to year, or between demographic groups. The Data Center facilitates two approaches to this calculation:

  1. To do a quick, approximate assessment of whether the differences between two estimates are significant, users can visually compare the MOEs of the estimates on the SHADAC Data Center’s state rank charts.  
  2. Users who want to get a more robust test can also use the MOEs to conduct a hypothesis test called a t-test.

Using an indicator from the National Health Interview Survey (“Had General Doctor or Provider Visit in Past Year by Age”) as an example, this brief provides step-by-step instructions for conducting a quick visual assessment for significance and for conducting the more robust t-test to see if the difference between two independent estimates is significant at the 95% confidence level. 

Access the brief.