Requirements / Responsibilities of EDI users

Users must agree to follow specific implementation rules in order to obtain accurate, reliable data from the EDI and maintain confidence for comparative purposes in future EDI data.

  1. Prospective users must purchase a license from the Offord Centre for Child Studies (OCCS), through the EDI author, and sign a user’s agreement. Researchers should provide a brief project description including specific timing of data collection and data collection area.
  2. Users agree to share an electronic copy of raw data collected using the EDI with the OCCS, following a signed data-sharing agreement.
  3. Data collection must take place in the second half of the kindergarten year, no earlier than five months from the start of the school year and no later than eight months after the start of the school year, giving teachers sufficient time to become familiar with their students’ abilities. For example, if the school year starts in September, the EDI should be completed no earlier than January and no later than April.
  4. Data in analyses and reports must be reported only at aggregated levels without any information that could be used to identify specific individuals and EDI data must never be used at an individual level for diagnostic purposes.
  5. EDI data can be used at an individual level during analysis for matching data from other sources (e.g., results from other school-directed testing), but only if the resulting analyses are based on group comparisons.
  6. The database must not include students’ names. The database with the EDI data should include: child’s date of birth, gender, and postal code, unique school and teacher identification numbers, and time of class.
  7. Confidentiality of raw data must be maintained at all stages of the data collection and analysis.
  8. The OCCS may be requested to analyze data and provide analytical results for report preparation, with payment for such services.
  9. If users conduct analysis and reporting, their guidelines must be made available to the OCCS.
  10. If the EDI data are linked with any other data for research purposes, the OCCS must be informed in advance of such a project, and a copy of the report outlining the results sent to the OCCS.
  11. A copy of any research paper or any other publication including data collected with the EDI should be sent to the OCCS.

Additional requirements for community-wide implementation

  1. For implementations directed from the OCCS, local project directors must contact the OCCS team no later than September 30th of the school year in which implementation is going to take place.
  2. Any reports sent to schools and communities must be accompanied by interpretation guidelines and copies sent to the OCCS by the project directors.
  3. Supporting materials available from the OCCS must be used to inform teachers and community representatives. The updated version of the EDI Guide will be provided to teachers.
  4. The groups for which the EDI is implemented must be based on geographical (e.g., city, census tract) or administrative boundaries (e.g., school board, network of schools).
  5. Extreme caution should be given to interpretation of data for aggregations of 10 or fewer children. The smallest level of data aggregation recommended is either the school or census tract.
  6. For population-wide reporting purposes, the designation “vulnerable” should only be applied to children who score in the lowest 10th percentile on one or more scales.

Additional requirements for research projects

  1. If the EDI is used for a research project, a list of all other measures used that may potentially be linked with the EDI will be provided to the OCCS.
  2. Major analyses must always entail the five predefined domains. If any other subscales are reported, they must be scientifically justified, and their internal reliability must be examined and included.
  3. If users require data for research purposes only, the time of implementation can be adjusted; however, the timing of data collection in relation to the school year must be clearly stated in the brief research description sent to the OCCS.