Good day,

I am a final-year Economics student and I'm working on the thesis which involves econometric analysis.

For my thesis, I'm using Understanding Society Survey datasets (https://www.understandingsociety.ac.uk/) and I had to merge 3 waves of the survey to have enough observations. However, some individuals (each individual has his/her specific identifier across waves) now have observations from Wave 2, 4 and 6, while others have observations from 2 or only 1 wave of 3 and each observation from each wave has it's own (and different) weight (b_psnenub_lw and/or d_psnenub_lw and/or f_psnenub_lv), so if an individual has observations from 2 or 3 waves, he/she also have 2 or 3 survey weights. For my regression, I'm using the data of an individual which is the latest (if an individual was surveyed in Waves 2,4 and 6, I'm using Wave 6 data, and if an individual was surveyed only in Waves 2 and 4, I'm using his/her data from Wave 4 and so on). So, I wanted to ask how should I adjust weights or how should I combine weights to construct a single weight for each individual?

Thank you.