https://iserredex.essex.ac.uk/support/https://iserredex.essex.ac.uk/support/support/favicon.ico?15995719382013-06-07T14:15:22ZUnderstanding Society User SupportUnderstanding Society User Support - Support #153: weighting BHPS sample correctly in longitudinal analysishttps://iserredex.essex.ac.uk/support/issues/153?journal_id=4322013-06-07T14:15:22ZRedmine Adminjpeterb@essex.ac.uk
<ul><li><strong>% Done</strong> changed from <i>0</i> to <i>50</i></li></ul><p>Dear Robert,</p>
<p>If you are interested in descriptive statistics and your analysis<br />sample is UKHLS Wave 2 BHPS sample, then using the weights<br />b_indin01_lw, produces unbiased estimates for the 2001 UK (and<br />England) population that has survived until 2010. These weights<br />correct for unequal selection probability (particularly because of the<br />regional boost samples) and attrition from 2001 to 2010. As you are<br />using only the England sample, unequal selection probabilities is not<br />a problem, but attrition is. As you rightly pointed out these weights<br />do not correct for the additional self-completion non-response. One<br />option is to estimate a self-completion non-response model yourself<br />and then multiply the existing weight with the inverse of the response<br />probability.</p>
<p>But in case of multivariate analysis, there is another problem. The<br />answer to the question, whether using weights in multivariate analysis<br />produces consistent estimates of the coefficients is not quite<br />straight-forward. Here is a good reference. "What are we weighting<br />for?" Solon, Haider and Wooldridge, NBER working paper 18859.</p>
<p>Best wishes,<br />Alita</p> Understanding Society User Support - Support #153: weighting BHPS sample correctly in longitudinal analysishttps://iserredex.essex.ac.uk/support/issues/153?journal_id=4542013-06-27T14:08:01ZRedmine Adminjpeterb@essex.ac.uk
<ul><li><strong>Status</strong> changed from <i>New</i> to <i>Closed</i></li><li><strong>% Done</strong> changed from <i>50</i> to <i>100</i></li></ul>