Weights in survey data
Using Weights in the Analysis of Survey Data
What do we mean by Survey weights?
A Value assigned to each case in the data file
Usually it used to make statistics computed from the data more representative of the population benchmark
E.g. A value shall indicate how much each case will count in a statistical procedure
Examples:
-- A weight of 4 means that a case is counted as four in the dataset
-- A weight of 1 means that a case is counted as one in the dataset
-- Weights can (and often are) fractions, but are always positive and non-zero
In stats we call them as pweight
Types of Survey Weights
Two most common type:
-- Design Weights
-- Post-Stratification or Non-responsive weights
Design Weight:
-- Normally used to compensate for over – or under – sampling of specific cases or for disproportionate stratification
--Example: it is common practice to over-sample minority group members or persons living in areas with larger percentage minority. If we doubled the size of our sample from minority areas, then each case in that area would get a design weight of ½ or 0.5
-- The design weight when we want the statistics to be representative of the population
To know more, click here Weights_in_Survey_Data_Final