Codebooks

In short, a codebook is a document which describes the content, structure, and the variables of a dataset. We normally use the codebook to understand a dataset.

A minimum structure of a codebook is:

Introduction: Previous considerations, definitions, etc.
Description of the variables.

And a minimum structure of a dataset is:

Units of analysis and codes.
Relevant variables
Other variables

How to use codebooks (castellano | català)

Democracy and Dictatorship dataset

A simple example of a dataset is the Democracy and Dictatorship (DD) dataset, also known as the Cheibub-Gandhi-Vreeland (CGV) index (Cheibub, Gandhi, and Vreeland 2010). This dataset can be found in the personal webpage of one of the authors, José Antonio Cheibub.

Using Excel

Since the DD dataset is relatively short (9159 observations and 78 variables), a good choice for learning is to open it using Excel or Google Sheets. The video attached to this page shows how to explore the data via Google Sheets dynamic tables. The procedures are the following:

Download the Excel file in the author’s webpage.
Upload the Excel file in Google Sheets.
Select all the rows and colums.
Go to Data -> Create Filter

Using R

R requires a more sophisticated knowledge of data wrangling, but this type of software is essential with large datasets (the DD dataset has more than 9.000 observations). The next lines of code show how to upload the DD dataset in R and to display the first six rows. When we upload a dataset to R we must convert it to an R object, normally a dataframe. In Table 1 we display the result of applying the function head() to the dd dataframe.

library(foreign)
library(dplyr)
dd <- as_tibble(read.dta("https://uofi.box.com/shared/static/bba3968d7c3397c024ec.dta"))
head(dd)

Table 1: Democracy and Dictatorship dataset
order	ctryname	year	aclpcode	cowcode	cowcode2	ccdcodelet	ccdcodenum	aclpyear	cowcode2year	cowcodeyear	flagc_cowcode2	entryy	exity	cid	wdicode	imf_code	politycode	bankscode	dpicode	uncode	un_region	un_region_name	un_continent	un_continent_name	aclp_region	bornyear	endyear	exselec	legselec	closed	nheads	nhead	npost	ndate	ageeh	royal	ehead	epost	edate	tenure08	ecens08	edeath	flageh	regime	flagc	flagdem	flagreg	agedem	agereg
1	Afghanistan	1946	142	700	700	AFG	1	1421946	7001946	7001946	1	1946	2008	700	AFG	512	700	10	AFG	4	34	Southern Asia	142	Asia	9	1919	2008	3	2	2	1	Mohammad Zahir Shah	king	11.08.33	14	1	Mohammad Zahir Shah	king	11.08.33	20	1	1	1	5	1	1	1	18	18
2	Afghanistan	1947	142	700	700	AFG	1	1421947	7001947	7001947	0	1946	2008	700	AFG	512	700	10	AFG	4	34	Southern Asia	142	Asia	9	1919	2008	3	2	2	0	Mohammad Zahir Shah	king		15	1	Mohammad Zahir Shah	king		20	1	1	0	5	0	0	0	19	19
3	Afghanistan	1948	142	700	700	AFG	1	1421948	7001948	7001948	0	1946	2008	700	AFG	512	700	10	AFG	4	34	Southern Asia	142	Asia	9	1919	2008	3	2	2	0	Mohammad Zahir Shah	king		16	1	Mohammad Zahir Shah	king		20	1	1	0	5	0	0	0	20	20
4	Afghanistan	1949	142	700	700	AFG	1	1421949	7001949	7001949	0	1946	2008	700	AFG	512	700	10	AFG	4	34	Southern Asia	142	Asia	9	1919	2008	3	2	2	0	Mohammad Zahir Shah	king		17	1	Mohammad Zahir Shah	king		20	1	1	0	5	0	0	0	21	21
5	Afghanistan	1950	142	700	700	AFG	1	1421950	7001950	7001950	0	1946	2008	700	AFG	512	700	10	AFG	4	34	Southern Asia	142	Asia	9	1919	2008	3	2	2	0	Mohammad Zahir Shah	king		18	1	Mohammad Zahir Shah	king		20	1	1	0	5	0	0	0	22	22
6	Afghanistan	1951	142	700	700	AFG	1	1421951	7001951	7001951	0	1946	2008	700	AFG	512	700	10	AFG	4	34	Southern Asia	142	Asia	9	1919	2008	3	2	2	0	Mohammad Zahir Shah	king		19	1	Mohammad Zahir Shah	king		20	1	1	0	5	0	0	0	23	23

These are most common functions to explore a dataframe, all applied to the dd dataframe:

head(dd): Displays the first 6 rows.
tail(dd): Displays the last 6 rows.
dim(dd): Displays the rows and columns.
glimpse(dd): Displays all the columns and the first observations of the dataframe in a vertical format.
View(dd): Opens a spreadsheet similar to Excel.

Other codebooks

Other codebooks you might want to explore are the following ones:

UCDP Dyadic Dataset Codebook Version 20.1 (Pettersson 2020): The codebooks of the UCDP are usually clean and easy to follow. See more in the UCDP webpage.
COW Trade Data v4.0 (Barbieri and Keshk 2016): It contains sections on the procedures and on missing data. This is one of the bold examples when the creation of a dataset generates a paper with some findings. In this case, the authors realized that missing trade data was not only associated with the level of development of the reporting country, but also with the type of relations within a dyad. Two countries in conflict are less likely to report trade data among them than countries having peaceful relations (Barbieri, Keshk, and Pollins 2009).
Intergovernmental Organizations v3: The paper associated to this data is a good example of the new evidence in form of graphs and charts that provides a new dataset (Pevehouse et al. 2019).

References

Barbieri, Katherine, and Omar M. G. Keshk. 2016. “Correlates of War Project Trade Data Set Codebook, Version 4.0. Online.”

Barbieri, Katherine, Omar M. G. Keshk, and Brian M. Pollins. 2009. “Trading Data: Evaluating our Assumptions and Coding Rules.” Conflict Management and Peace Science 26 (5): 471–91.

Cheibub, José Antonio, Jennifer Gandhi, and James Raymond Vreeland. 2010. “Democracy and Dictatorship Revisited.” Public Choice 143 (2-1): 67–101.

Pettersson, Therese. 2020. “UCDP Dyadic Dataset Codebook v 20.1.”

Pevehouse, Jon C. W., Timothy Nordstron, Roseanne W. McManus, and Anne Spencer Jamison. 2019. “Tracking Organizations in the World: The Correlates of War IGO Version 3.0 datasets.” Journal of Peace Research 57 (3): 492–503.

Last updated on Jul 18, 2021