Just as the design of the data collection instruments affects data quality and representativeness, the way you go about collecting data is also important. Here are some tips to help you increase participation and ensure that you are gathering high-quality, representative data.
What States Are Doing
Almost all states use self-administered paper or online surveys to collect data on parent involvement.
Two elements of timing are important for data collection: scheduling and the field period.
Data collection mode or delivery method can affect both data quality and response rate. Common issues associated with different survey modes include
Having complete and accurate contact information is very important. The better the quality of parents’ contact information, the better the chance of achieving a higher response rate.
OSEP requires states to report the number of parents to whom they distributed surveys and the number of parents who responded. States also must compare response rates from year to year for specific groups of parents and describe strategies they will implement to increase the response rate, particularly for those groups that are underrepresented. It is important that states have a system for monitoring response rates and representativeness. (See the section on creating a data file for more information about how to set up your tracking file to ensure that the data you enter are accurate.)
Nonresponse bias: The state must analyze the response rate to identify potential nonresponse bias and take steps to reduce any identified bias and promote response from a broad cross-section of parents of children with disabilities. Nonresponse bias can occur when the individuals who complete the survey differ in meaningful ways from those who do not. Nonresponse bias arises when two conditions occur: (1) certain subgroups are less likely to respond to a survey, resulting in their systematic underrepresentation in the survey data, and (2) the underrepresented subgroups differ from other subgroups in what the survey is trying to measure. Plan to monitor data as they arrive and conduct nonresponse follow-up during data collection to minimize the possibility of bias as well as check for nonresponse bias after data are collected.
What States Are Doing
Connecticut modified its survey distribution process in an effort to increase response rates. The state sent districts an Excel template with the state-assigned student identifiers for the special education students in their district and asked districts to enter the most current mailing address for each student. This process was a change from prior years when the state sent districts an Excel spreadsheet with students’ mailing addresses prepopulated (as extracted from the state’s special education data system) to confirm the mailing addresses. This revised process helped to ensure that each district reviewed and edited the Excel spreadsheet with the most current addresses. In addition, districts were able to indicate if a child had moved out of the district or if any additional students had been identified. The state also asked districts to provide parents’ email addresses, when available, to allow for direct parent access to the survey through a personalized link. Nearly all districts were able to provide emails for some or all parents of students with disabilities.