I'm not sure how to use the weighting columns in the NTS. Is NTS representative sample of population?, and if not, are the weights used to adjust for that?
1.12 There are some households where all the interviews were completed, but not all travel diaries. In this instance, the interview data is still useful, but if the diary data were to be used for the remaining individuals, it would give an under-estimation of the travel behaviour for those households. These households are known as partially productive households.
1.13 A household where everyone completes the interview as well as everyone completing the travel diary, is called a fully productive household. For these households, weighting W1 is equal to one (please see chapter 3 – Data Structure – for a discussion of weighting)
1.16 Because of this distinction between fully- and partially-productive households, as well as different weights applied for data obtained from the interviews and the travel diaries, it is important to understand what sample you are using.
Data gathered from the interview is called the Interview Sample. If you are considering a problem which only requires information from the interview (Household, Individual, Ticket, Vehicle tables only), then you are using the interview sample. This sample also includes partially-productive households.
Data gathered from the travel diaries is called the Diary Sample. If you are considering a problem which requires information from the travel diary (Trip, Stage, LDJ tables), then you are using the diary sample, no matter how little data you use for these tables. This sample only includes fully-productive households
I'm not sure how to use the weighting columns in the NTS. Is NTS representative sample of population?, and if not, are the weights used to adjust for that?
Notes from NTS technical reports
from the Data Extract User Guide:
Further explanation in NTS technical report - Chapter 6 (Weighting)
Using weights in matching