The data you are about to download is from the SALDRU household survey. It contains detailed household and individual level data from South Africa collected in 1993. Due to the size of the SALDRU data set, the data files have been compressed in the ".zip" format. You can uncompress/unzip the data using a number of freely available utilities. (Click here to obtain a zip utility to uncompress the file if such a utility is not available on your personal computer.)
The data in zipped format will use up approximately 2 megabytes of hard drive space. When you unzip it, it will use up approximately 20 megabytes.
We will be using this data in the statistical software program STATA. For older computers, the entire data set may not fit into memory when you run STATA. For this reason, we provide smaller sub-samples of the full data set. These files are randomly selected sub-samples.
You will want to download only the dataset that you expect to use. We provide some guidelines to help you guess which sub-sample to download. As a first pass, assuming you are using Microsoft Windows, click on the "Start" button, then select the "Control Panel" from "Settings." In the window that will open, click on the "System" icon. The system window that opens should provide information related to your personal computer. In the bottom right corner, the total available RAM should be listed. This is the total available memory on your computer, please make a note of it.
Remember, download the data set that most closely fits your computer's memory capability. To download the data set, click on the link below using the right mouse button. A target menu will pop up with several options, select the "Save Target As.." option.
You will have the opportunity to save the (zipped) data file to your hard drive. We recommend saving the data file in the folder C:\MellonCourse. Regardless of where you choose to save the data please make a note of which of the data sets below you download and where it is located on your hard drive.
** IMPORTANT NOTE -- We should note that throughout the lesson modules we will assume users are using the smallest of the sub-sample data files Saldru12. For those users who downloaded a larger data file (Saldru40, Saldru60, Saldru100), it may be useful to download the Saldru12 data file as well. By using the Saldru12 data file when working through the lesson modules, you will produce the same results as those given in the lesson modules (if the steps you take in STATA are correct that is). If you use one of the larger data files to work through the lesson modules your answers may be slightly different than those presented. Now certainly when you go to work on your own project, or want to explorer the data on your own you will want to use the largest data file possible.
SALDRU Data for approximately 16mb of total memoryThis data file contains about 5,300 Individual Level observations and 1,062 Household Level Observations.
SALDRU Data for approximately 24mb of total memory
This data file contains about 17,460 Individual Level observations and 3,542 Household Level Observations.
SALDRU Data for approximately 32mb of total memory
This data file contains about 26,451 Individual Level observations and 5,312 Household Level Observations.
SALDRU Data for 64mb or greater of total memory for
This data file contains about 43,984 Individual Level observations and 8,854 Household Level Observations.