Table-related commands in STATA

Abbreviations of commands

table	(no abbv.)
tabulate	ta tab
tabstat	(no abbv.)
summarize	su

Differences between table, tabulate, tabstat, summarize

	one-way	two-way	options
table	table v1 create a one-way table of v1 with simple statistics	table v1 v2 create a two-way table of v1 in row† and v2 in column†	,statistics( )
tabulate	tabulate v1 create a one-way table of v1 with detailed statistics	tabulate v1 v2 create a two-way table with v1 in row† and v2 in column†	,chi2 Pearson's chi-squared test; *only for two-way ,summarize(v3) detailed statistics for v3
tabstat	tabstat v1 create a one-way table of v1 with detailed statistics	*no two- or multiple-way table	,statistics( ) ,by(v3) detailed statistics for each of v3
summarize	summarize v1 detailed statistics of v1	*no two- or multiple-way summary	,detail

† row = transverse direction, column = longitudinal direction

Sample data

Suppose we have such a dataset in STATA.

Where,

id	discrete	:Identification number
sex	binary	:Male=0, Female=1
data1	continuous	:Results of a certain test
factorA, B, C	binary	:Negative=0, Positive=1
SES	categorical	:Categories of Socio-Economic Status, divided into four
disease	binary	:Free from a certain disease=0, Having the disease=1

One-way

Summary of sex, a binary variable

table		Both reports frequency but tabulate is more detailed
tabulate
tabstat		Both reports mean but summarize is more detailed
summarize

Summary of data1, a continuous variable

table		Both reports frequency of each value, which does not make sense
tabulate
tabstat		Both reports mean but summarize is more detailed
summarize

Summary of SES, a categorical variable

table		Both reports frequency but tabulate is more detailed
tabulate
tabstat		Both reports mean but summarize is more detailed
summarize

One-way, multiple

table	*Both do not create one-way multiple table
tabulate
tabstat		Reports mean in row (transverse) direction
summarize		Reports more details in column (longitudinal) direction

Two-way

Summary of factorA based on sex

table		Both creates the same table but tabulate is better visualized
tabulate
tabstat		Both reports mean but summarize is more detailed; needs bysort option before the command
summarize

Summary of sex based on factorA

table		Both creates the same table but tabulate is better visualized
tabulate
tabstat		Both reports mean but summarize is more detailed; needs bysort option before the command
summarize

Summary of data1 based on disease

table	*Both do not create a meaningful table for continuous variable
tabulate
tabstat		Both reports mean but summarize is more detailed; needs bysort option before the command
summarize

Two-way with proportions

Summary of factorA based on sex with proportions

table		This calculates proportions of cells compared to the whole without showing raw values
		This calculates proportions in column (longitudinal) directions without showing raw values
		This calculates proportions in row (transverse) directions without showing raw values
tabulate		This calculates proportions in column (longitudinal) directions
		This calculates proportions in row (transverse) directions

Two-way, multiple

Summary of factorA, factorB, factorC based on disease

tabstat
		factorA,B,C are binary variables so summations of values provide the positivities of factorA,B,C
		statistic(n) (statistic(count) is the same) only counts observations with real values, which only tell non-missing observations

Summary of factorA, factorB, factorC based on SES

tabstat
		factorA,B,C are binary variables so summations of values provide the positivities of factorA,B,C
		statistic(n) (statistic(count) is the same) only counts observations with real values, which only tell non-missing observations