class 2- undercoverage Flashcards
types of frames
area frame- counties
list frame- list of hospitals, students
telephone frame
frames
list used for sampling
implies finite population
coverage error
mismatch between the frame and the target population
target population
finite population that the survey is attempting to characterize
inferential population
potentially infinite population conclusions are supposed to apply to
frame population
population accessible via the frame- all people living in households accessible by landlines
covered
in frame
in target pop
overcoverage
Ineligible in frame out target pop \+blanks, empty listings \+boundary problems
(undercoverage)
omissions
exclusions
part of target population but not linked to frame
multiplicity problems
person linked to multiple Housing units in area survye
duplicate address for same residence
multiple phone lines in same residence
improper use of clustered frames
failing ot adjust-e.g.,people in certain size hh have diff selection rate
sampling within household withought adjusting-using weights
incorrect-missing information
incorrect information for contact
access to contact info
e.g., email addresses for web based survey
consequences of exclusion-undercoverage
difference in means for people covered in the frame compared to the mean of the target population
consequences of overcoverage
increase costs
List frame problems
duplicates
omissions, outdated
ineligibles