How to Create Publication-Ready Tables from R Regression Output with Stargazer

You can utilize the powerful stargazer package within R to produce exceptional quality tables suitable for immediate use in academic publications and reports. The following detailed example walks through the necessary steps, starting with installation and moving toward generating both descriptive and inferential statistical output using a standard, built-in R dataset.

Prerequisites and Installation

Before utilizing any package in R, it must be installed on your system and subsequently loaded into the active environment. The stargazer package is available on CRAN (Comprehensive R Archive Network), making the installation process straightforward using the install.packages() command. Following installation, the library() function ensures that the package’s functions are available for immediate use in your script or console session.

We begin our practical example by executing the required code block to set up the environment:

#install stargazer package
install.packages('stargazer')

#load stargazer package
library(stargazer)

Core Syntax of the stargazer Function

Once the package is successfully loaded, the primary function, stargazer(), is ready for use to produce high quality statistical tables. This function is highly customizable, allowing users to control the input data, the output format, the title of the table, and the file destination for exporting the results. Understanding the basic syntax and its key arguments is crucial for generating tailored output.

The generalized structure of the function call is as follows:

stargazer(df, type=’text’, title=’my_title’, out=’my_data.txt’, …)

The key arguments govern how the function processes the data and formats the resulting table:

• df: This argument specifies the name of the statistical object being summarized. This could be a data frame containing raw data for descriptive summary statistics, or a fitted statistical object, such as a linear model (lm) or generalized linear model (glm), for regression model summaries.
• type: This parameter determines the output format. Common options include 'text' (for console display or plain text files), 'html' (for web display or easy integration into word processors), and 'latex' (standard for academic journals).
• title: A string defining the descriptive title that will appear at the top of the generated table.
• out: Specifies the file path and name to which the generated table will be saved.

It is important to note the flexibility of the out parameter regarding file type. For the examples presented here, we will utilize the .txt extension to export simple text files suitable for console viewing. However, researchers frequently use the .html extension to export tables as high-fidelity HTML files, or the .tex extension for LaTeX integration, depending on the publication requirements.

Two Primary Uses of stargazer

The stargazer() function is predominantly utilized to generate two distinct categories of statistical tables, each serving a critical purpose in data analysis and reporting:

• Generation of comprehensive descriptive summary statistics for all relevant variables contained within a data frame.
• Production of highly standardized and easily interpretable summary tables for the results of a fitted regression model, detailing coefficients, statistical significance, and overall model fit statistics.

Generating Descriptive Summary Statistics

The first common application involves producing a table of descriptive statistics. We will demonstrate this using the well-known mtcars dataset, which is built into the base R programming language environment. This dataset provides information on 32 automobiles regarding 11 variables, including fuel consumption (mpg), cylinder count (cyl), and horsepower (hp). By passing the mtcars object directly to stargazer(), we instruct the package to calculate and format key descriptive metrics for every column.

The following code snippet creates a descriptive table, outputs the result to the console (by setting type='text'), and simultaneously saves the output to a text file named mtcars_data.txt.

#create table that provide summary statistics of each variable in dataset
stargazer(mtcars, type='text', title='Summary Statistics', out='mtcars_data.txt')

Summary Statistics
=============================================================
Statistic N   Mean   St. Dev.  Min   Pctl(25) Pctl(75)  Max  
-------------------------------------------------------------
mpg       32 20.091   6.027     10     15.4     22.8     34  
cyl       32  6.188   1.786     4       4        8       8   
disp      32 230.722 123.939    71    120.8     326     472  
hp        32 146.688  68.563    52     96.5     180     335  
drat      32  3.597   0.535   2.760   3.080    3.920   4.930 
wt        32  3.217   0.978   1.513   2.581    3.610   5.424 
qsec      32 17.849   1.787   14.500  16.892   18.900  22.900
vs        32  0.438   0.504     0       0        1       1   
am        32  0.406   0.499     0       0        1       1   
gear      32  3.688   0.738     3       3        4       5   
carb      32  2.812   1.615     1       2        4       8   
-------------------------------------------------------------

Interpreting Descriptive Output

The resulting output is a meticulously structured table that provides essential summary statistics for every quantitative variable within the mtcars data frame. Each row corresponds to a variable (e.g., mpg, hp), and columns provide statistical measures such as the sample size (N), the Arithmetic Mean (Mean), the Standard Deviation (St. Dev.), the minimum and maximum values (Min and Max), and critical percentiles (Pctl(25) and Pctl(75)). This display is significantly more organized and professional than the default output provided by R’s base functions.

For instance, examining the row for mpg (miles per gallon), we immediately see that the average MPG is 20.091 with a standard deviation of 6.027, ranging from a minimum of 10 to a maximum of 34. This succinct presentation allows readers to quickly grasp the distribution and central tendency of the variables being analyzed. Furthermore, because we used the out parameter, the complete summary statistics table has been exported to the specified text file (mtcars_data.txt) in the working directory, facilitating easy integration into external documents.

Generating Regression Model Summaries

The most frequent and powerful application of the stargazer package is the summarization of fitted statistical objects. Unlike descriptive statistics, which summarize the raw data, regression summaries report the results of inference—that is, the relationship between dependent and independent variables. To demonstrate this, we will first fit a simple linear model (lm) predicting miles per gallon (mpg) based on displacement (disp) and horsepower (hp) using the mtcars data.

Once the model (fit) is created, we pass this fitted object directly to the stargazer() function. The package automatically recognizes that the input is a statistical model and structures the output accordingly, focusing on coefficients, statistical significance, and overall model diagnostics.

#fit regression model
fit <- lm(mpg ~ disp + hp, data=mtcars)

#create table that summarizes regression model
stargazer(fit, type='text', title='Regression Summary', out='mtcars_regression.txt')

Regression Summary
===============================================
                        Dependent variable:    
                    ---------------------------
                                mpg            
-----------------------------------------------
disp                         -0.030***         
                              (0.007)          
                                               
hp                            -0.025*          
                              (0.013)          
                                               
Constant                     30.736***         
                              (1.332)          
                                               
-----------------------------------------------
Observations                    32             
R2                             0.748           
Adjusted R2                    0.731           
Residual Std. Error       3.127 (df = 29)      
F Statistic           43.095*** (df = 2; 29)   
===============================================
Note:               *p<0.1; **p<0.05; ***p<0.01

Interpreting Regression Output

The resulting regression model summary is highly informative. The upper section of the table displays the estimated coefficients for each predictor variable (disp and hp) and the intercept (Constant). Beneath each coefficient estimate, the standard error is provided in parentheses. The asterisks (e.g., ***, *) indicate the level of statistical significance, based on the p-value thresholds defined at the bottom of the table.

The output displays the coefficients for each term in the regression model along with various summary statistics near the bottom of the table. In our example, the coefficient for disp (displacement) is -0.030 and is highly significant (***, p < 0.01). The lower section of the table provides crucial model diagnostics, including the R-squared (0.748), which indicates that approximately 74.8% of the variance in MPG is explained by displacement and horsepower.

Advanced Output Formatting: HTML and LaTeX

While the examples above utilized the type='text' setting for simple console display, the true power of stargazer lies in its ability to generate output in formats optimized for publication quality. By changing the type argument to 'html' or 'latex', the package generates code that can be directly embedded into web pages or academic manuscripts compiled using LaTeX, respectively.

For instance, setting type='html' and specifying an output file name ending in .html will produce a complete HTML table definition, complete with CSS classes and inline styling, ensuring perfect visual fidelity regardless of the destination platform. This capability is vital for researchers aiming for maximum efficiency and professionalism in their statistical reporting.

Note: Comprehensive documentation detailing all customization options, including how to select specific statistics, rename variables, or combine multiple models into a single comparative table, can be found on the official documentation page for the stargazer package.

Further Resources for R Data Manipulation

To enhance your statistical workflow within R, consider exploring related topics on data structure and manipulation:

How to Loop Through Column Names in R
How to Create an Empty Data Frame in R
How to Append Rows to a Data Frame in R