Understanding Society User Support

Hi Liam,
the variables starting with the ff prefix are not technically derived variables. They report the value of that variable that has been recorded during the last interview with that person. This last interview may have taken place in the previous wave but it could also be from before then, e.g. a proxy interview [_ivfio=2] may not count as the last interview. Generally it should be noted that the feed forward file is typically compiled at a time when the data processing has not concluded. That is, it uses unedited data that has not yet undergone the plausibility testing we apply before we release the data. We provide ff variables to allow users to replicate the routing used in the interview.

An implication of all this is that it is possible that a fed forward variable indicates that there is, e.g., a valid job description b_ff_jbsoc00>=0 in the data file b_indresp but the person's value in the a_indresp data file indicates that the description is missing or wild, i.e., a_jbsoc00==-9.

We hope this helps,
Gundi

Hi Gundi,

That is helpful, thank you.

Am I right in saying the variables are calculated as I thought above (data cleaning aside)?

w_ff_ivlolw=w-1_ivfio; w_ff_jbstat=w-1_jbstat; w_ff_emplw=(w-1_jbhas==1 | w-1_jboff==1)

Also, I have another question about the routing in the Annual History Data. From what I can tell, it is possible that participants can be asked both w_notempchk and w_empchk, which means w_cjob can refer to (and be routed from) two different jobs given by participants where w_nxtst==1 and (w_jbsamr==2 | w_samejob==2). Do you know what happens in the module where participants are asked both w_notempchk and w_empchk? (I have attached a diagram of the routing for Wave 2, as I understand it.)

Thanks,

Liam

This issue continues in #957.