Wyoming Prevention Needs Assessment Survey

Effects of COVID-19 Mitigation Efforts in 2020

In the middle of March 2020, schools throughout Wyoming ended their in-person instruction as they implemented COVID-19 mitigation efforts. Because the PNA is conducted in the schools, these events shortened the 2020 PNA data collection from the planned four months to only a month and a half. The shortened data collection window resulted in lower response rates and less participation than typical survey years.

While the survey was open, from February through Mid-March 2020, we collected data from 114 schools of the 185 eligible schools (62%). We usually have about 17,500 valid surveys at the student level when the large school districts elect to participate. In 2020, the number of valid surveys was reduced to 11,046, 63% of what the research team expected.

Lower Response and Participation Numbers

From 2001 to 2018, the PNA results were weighted to account for grade-level response rates in each school district. These weights balance different amounts of participation in each grade level. They also ensure that when multiple school districts’ data are combined for county and state-level estimates, each school district’s size is correctly accounted for. An example of weighting on the PNA is provided below.

In 2020 with over 11,000 valid surveys, the research team felt it was appropriate to present as many results as possible for the state, counties, and school districts. Because of the interrupted data collection, the missing data patterns resulted in more large weights than typical, and many school districts were missing data from one or more grade levels. Based on these concerns, the research team decided it was better to present the unweighted results for 2020.

Comparisons of weighted data results (PNA 2001 to 2018) and unweighted data results in 2020 should be made with caution. Changes or differences from past years to 2020 may be partially because the 2020 results are unweighted.

Additionally, for some counties and school districts, the research team did not present 2020 results because the available data may not be representative. For example, in a county that has multiple high schools, only one participated. The data needed to come from more than one of the schools to be representative. Thus grade level results for 10th and 12th grades are not provided for that county. Decisions to not present specific results are listed on each graph’s data notes.

WEIGHTING EXAMPLE:

In a school district, if 80% of the 6th grade participated in the survey that year, then all the 6th-grade students would have a weight of Wgrade 6 = 1/0.80 = 1.250. This means that each 6th-grade student’s survey represents one and a quarter person. In contrast, if the 12th grade had a 60% response rate in the same school district, they would have a weight of Wgrade 12 = 1/0.60 = 1.667. The 12th-grade student’s survey represents one and two-thirds person. The lower the response rate, the higher the weight. When survey results are unweighted, these calculations are not performed, and each valid survey only represents one student (Wgrade i = 1).