Sampling
● In years past , survey participation rates have been markedly lower among non-white residents , as compared to white residents .
● This year , in an effort to increase participation among non-white residents , a portion of the postcard invitations were sent only to zip codes containing higher proportions of non-white residents .
● As shown on the next slide , of the 12,000 postcards sent :
• 7,000 of the postcards were sent as usual – to randomly selected residential addresses proportionate to Arlington ’ s zip code distribution . This continuity in methodology is important in ensuring that any year-over-year changes in survey results are due to shifts in resident perceptions , attitudes , etc ., and not due to a change in sampling methodology .
• 5,000 of the postcards targeted zip codes containing higher proportions of non-white residents . Survey completes from this group of invitations were marked as an augment so that any differences in results caused by this change in methodology could be identified .
● Ultimately , this limited augment did not meaningfully increase survey participation among non-white residents and had little impact on survey results . Throughout this report , data shown in total includes the augment group described above .
Copyright © 2024 Decision Analyst
6