Hi,
I am hoping you can clarify the VOSUPPWT. I am using the dataset linked to the Fall Data Challenge to teach my high school students. We are using Excel for our data analysis.
How are these values generated? And why are they so high?
How do I use the VOSUPPWT to analyze the data? Do I just multiply the weight with the variable of interest, e.g. how they voted? What if we want to do a two-way analysis?
Thank you so much for your help. I am looking forward to hearing from you.
Varada
Data used in the American Statistical Analysis Fall Data Challenge are from IPUMS CPS; the CPS is a monthly survey of the U.S. population that is designed to be nationally representative of the entire population without actually surveying each individual. Therefore, a sample weight is assigned to each individual in the data to estimate how many people they ‘represent’ in the U.S. population. These are offered as sample weight variables (such as VOSUPPWT) that allow data users to produce representative statistics. VOSUPPWT is the weight specific to the voter supplement; which was calculated by the U.S. Census (detailed information about how weights are calculated can be found on this weighting web page and Chapter 10 of technical paper 66). You can read more about sample weights in general on this blogpost and more about CPS sample weights on this FAQ page, this webpage on CPS weights, and this webpage on CPS sample design. In Excel, you can incorporate weights into an analysis in a few ways; by calculating the percent of a certain value by summing the weights for the value of interest divided by the sum of the number of responses (described here), through the use of SUMPRODUCT (described here and here), or by summing weights in the value field of a pivot table (described here).
Thank you very much, Grace.