I have seen work by Robert Renssen about 20 years ago on the file matching problem using three sets of weights. My conclusion at that time was that it would become complicated to generalize that to item nonresponse. Stef
Verstuurd vanaf mijn iBrain Op 17 jun. 2020 om 18:20 heeft Paul von Hippel <[email protected]> het volgende geschreven: Surveys often calculate weights to compensate for unit nonresponse. But what about item nonresponse? If different variables are missing from different cases, it seems to me the optimal set of weights would depend on the particular combination of variables being analyzed. Is there any systematic way to deal with this? The surveyor can offer different weights for different analyses, but the number of possibilities quickly become unwieldy. Best wishes, Paul von Hippel LBJ School of Public Affairs University of Texas, Austin (512) 537-8112 PaulvonHippel.com<https://urldefense.com/v3/__http://paulvonhippel.com__;!!Dq0X2DkFhyF93HkjWTBQKhk!Cz_puWUJHw4mOdYcsAUpSdPCuFf_D2UwNTMv-ePW3KrGhVTbCRqaO4rz71vJGVxE96Dvhr_i3CpvLw$> @PaulvonHippel<https://urldefense.com/v3/__https://twitter.com/paulvonhippel__;!!Dq0X2DkFhyF93HkjWTBQKhk!Cz_puWUJHw4mOdYcsAUpSdPCuFf_D2UwNTMv-ePW3KrGhVTbCRqaO4rz71vJGVxE96Dvhr_dVimQgw$> This message may contain information that is not intended for you. If you are not the addressee or if this message was sent to you by mistake, you are requested to inform the sender and delete the message. TNO accepts no liability for the content of this e-mail, for the manner in which you use it and for damage of any kind resulting from the risks inherent to the electronic transmission of messages.
