#Psid codebook code#

datafile$GeogMobilityHd <- NAĭatafile$GeogMobilityHd <- NA First, I convert NAs as marked in the survey to NAs as read by the computer. The survey includes a measure for geographic mobility: whether the head of household lives in same state, the same region but a different state, or a different region from where they grew up. #This variable equals 1 for all those (where it being 1 is TRUE) and 0 for the restĭatafile$MetroAdjacent <- 5ĭatafile$MetroAdjacent <- as.numeric(datafile$MetroAdjacent = 5)ĭatafile$SmallUrban <- 6ĭatafile$SmallUrban <- as.numeric(datafile$SmallUrban = 6)ĭatafile$Rural <- as.numeric(datafile$BealRuralUrbanCode = 9) datafile$Metropolitan <- datafile$BealRuralUrbanCodeĭatafile$MetroAdjacent <- datafile$BealRuralUrbanCodeĭatafile$SmallUrban <- datafile$BealRuralUrbanCodeĭatafile$Rural <- datafile$BealRuralUrbanCodeĭatafile$Metropolitan <- 1ĭatafile$Metropolitan <- as.numeric(datafile$Metropolitan = 1) #Example19xxRelated 250k (1,2,3,4), small urban metro-adjacent (5,7), small urban area not adjacent to a metropolitan area (6,8), and completely rural (9). #In 19xx data, a data entry issue led all 2s to accidentally be marked 3s

#Psid codebook code#

While the rest of this guide goes step by step through the data cleaning process for my PSID charitable giving research, the following block is just an example to illustrate this: #This next line of code does something that you dont need to run except for 19xx data I also put a # before “optional” lines of code which I would not regularly use, but might in certain circumstances. When you run your code, those lines will be skipped. You can add comments by using the # key at the start of any line. It also makes spotting errors easier, especially when youre tired and everything begins to look like gibberish. When writing code, documenting what you’re doing as clearly as possible makes life much easier for myself as well as anyone who might be using my code. I also do this with the R script files for the code I use. The same goes if you notice a mistake in your code much later– you can go right back to the most recent version before that change. If anything goes wrong, you can go right back to the version of the data file you saved right before, instead of having to start over from scratch. I create a copy of the data file before doing anything with it, and save a version to a separate file after each change or edit I make.

The goal is for any researcher with the data, even with next to no knowledge of R, to be able to easily edit any part they need (variable names, assignments, etc) without having to change any of the code and without it affecting any of the other variables. Programming elegance by something like using more if loops is ignored, in favor of the code being as easy to understand and modify as possible. This code cleans a downloaded selection of yearly PSID data to output a usable, clean dataset with the list of variables named for easy analysis.