WERT Learning Community
WFI-EZ National Means in WrapTrack.net
May 24, 2016 | firstname.lastname@example.org
WERT has moved away from calculating combined scores across respondents for the WFI-EZ tool. We encourage sites not to calculate those kinds of scores either. We think that the experiences of youth, caregivers, and team members, and the perception of facilitators are all distinct components of wraparound fidelity, and that when you combine scores you wash out important differences.
If you would like to calculate a “total respondent” national mean, you could calculate the average of the four respondent-level means. However, this brings with it its own challenges. To reflect an accurate mean you would have to weight each average by the number of respondents nationally whose scores created the respondent-level means for each category (roughly those numbers are as follows: n=1071 facilitators, n= 1121 caregivers, n=443 youth, and n=557 team members). Without this step, creating a simple average of the 4 respondent-level means would skew the total average toward the averages of the categories with the most respondents, and therefore would not represent an accurate total-respondent mean.
There are several ways someone might calculate a “total” score in a situation like this: averaging each of the respondent-level averages, as suggested, weighing the scores by the number of respondents of each type (essentially putting everyone into a single pool), or first creating team-level averages and calculating means from there.
Each of these options has its own disadvantages, and we have decided that the most useful approach is to simply look at each of the scores separately. Please be reminded to view the Data Interpretation document on our website under the WrapTrack tab for Current Collaborators for more information on how to select respondent-level national means.