This unit will discuss the kinds of research questions that require linked micro and macro data, methods of assessing micro data for variation between macro units prior to linking the data, how to link micro and macro data, weighting linked micro and macro data for particular kinds of comparative analysis, checking the quality of micro data with macro data and adjusting weights to account for discrepancies, and basic approaches to the analysis of linked micro and macro data.
Most of the activities in this unit are based on the analysis of a subset of the European Social Survey which can be found at the LIMMD web site. The data are in Stata format and the activities use Stata code.
Those who have access to Stata but have never used the programme before should find that for most of the activities it is simple to issue the commands in the Command window and view the results. For those without access to either Stata or the data (or simply those without the time to do the activities), even if you are not familiar with Stata, you should find that the Stata command language is sufficiently straightforward that it is possible to understand the commands just by inspection and comparison with the example output. Similarly, those who use other statistical analysis software should find it relatively easy to identify how to change the Stata code into some other language that they are familiar with. For instance, SPSS users should notice that the Stata command tabulate is similar to the 'crosstab' command in SPSS. The main purpose of the activities in this unit is not to provide instruction on the use of Stata, but to add clarity and aid learning of the statistical analysis and data manipulation procedures with practical examples.