,
11 tweets,
5 min read
Read on Twitter
#ESRA19 Todd Hughes up next talking about the California Health Interview Survey redesign. Goal of CHIS to provide substate and city level estimates and subpopulations. 
#ESRA19 Hughes - CHIS traditional methods = RDD to interview adult, adolescent, and child. But.... the known problems have occurred. Including this response rate decline.
#ESRA19 Hughes talking about statewide pilot for web with telephone follow-up. Web instrument was available in English and Spanish. Push to web had 3 mailings with next birthday method. Could call in if wanted. Then followed up with phone for matched HHS
#ESRA19 Hughes Mailed materials differed slightly for anticipated Spanish speakers
#ESRA19 Hughes web + phone doubled RR over RDD CATI. Lots of web completes. About 50% of CATI completes were inbound calls!
#ESRA19 Hughes - did more web completes in urban areas, so had better response rates and less variation across state in response rates. Wildfire in northern CA may have depressed RR in those areas.
#ESRA19 Hughes - not as successful at getting Rs in languages other than English. Maybe the higher rates of cell only in Spanish speaking populations at play.
#ESRA19 Hughes- key estimates - increase in middle aged Rs, increase in college graduates, more high income groups in web (CATI missing these folks, apparently). Lower Hispanic representation in web.
#esra19 Hughes Most estimates of health data not different, but some important estimates were different in web vs phone
#ESRA19 Hughes Costs for web + phone substantially less than CATI
#ESRA19 Hughes - report available here!
