The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

================

Introduction

xtsum is an R wrapper based on STATA xtsum command, it used to provide summary statistics for a panel data set. It decomposes the variable \(x_{it}\) into a between \((\bar{x_i})\) and within \((x_{it} − \bar{x_i} + \bar{\bar{x}}\), the global mean x being added back in make results comparable) (StataCorp 2023).

Installation

install.packages("xtsum")

# For dev version
# install.packages("devtools")
devtools::install_github("macosso/xtsum")

Getting Started

# Load the librarry
library(xtsum)

xtsum

This function computes summary statistics for panel data, including overall statistics, between-group statistics, and within-group statistics.

Usage

xtsum(
  data,
  variables = NULL,
  id = NULL,
  t = NULL,
  na.rm = FALSE,
  return.data.frame = TRUE,
  dec = 3
)

Arguments

• data A data.frame or pdata.frame object representing panel data.

• variables (Optional) Vector of variable names for which to calculate statistics. If not provided, all numeric variables in the data will be used.

• id (Optional) Name of the individual identifier variable.

• t (Optional) Name of the time identifier variable.

• na.rm Logical indicating whether to remove NAs when calculating statistics.

• return.data.frame If the return object should be a dataframe

• dec Number of significant digits to report

Example

Genral example

Based on National Longitudinal Survey of Young Women, 14-24 years old in 1968

data("nlswork", package = "sampleSelection")
xtsum(nlswork, "hours", id = "idcode", t = "year", na.rm = T, dec = 6)

The table above can be interpreted as below paraphrased from (StataCorp 2023).

The overall and within are calculated over N = 28,467 person-years of data. The between is calculated over n = 4,710 persons, and the average number of years a person was observed in the hours data isT = 6.

xtsum also reports standard deviation(SD), minimums(Min), and maximums(Max).

Hours worked varied between Overal Min = 1 and Overall Max = 168. Average hours worked for each woman varied between between Min = 1 and between Max = 83.5. “Hours worked within” varied between within Min = −2.15 and within Max = 130.1, which is not to say that any woman actually worked negative hours. The within number refers to the deviation from each individual’s average, and naturally, some of those deviations must be negative. Then the negative value is not disturbing but the positive value is. Did some woman really deviate from her average by +130.1 hours? No. In our definition of within, we add back in the global average of 36.6 hours. Some woman did deviate from her average by 130.1 − 36.6 = 93.5 hours, which is still large.

The reported standard deviations tell us that the variation in hours worked last week across women is nearly equal to that observed within a woman over time. That is, if you were to draw two women randomly from our data, the difference in hours worked is expected to be nearly equal to the difference for the same woman in two randomly selected years.

More detailed interpretation can be found in handout(Porter n.d.)

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.