Announcing improved table1_mc: Stata module to create "Table 1.docx" of baseline characteristics for a manuscript

[Mark Chatfield]

Dear Statalisters,

I have just posted an updated version of my program table1_mc on SSC. To make use of the ability to put the output into a docx file, version 15.1 is required, otherwise version 14.2.

. sysuse auto, clear
. generate much_headroom = (headroom>3)
. table1_mc, by(foreign) vars(price conts \ weight contn %5.0f \ rep78 cate \ much_headroom bin) nospace clear
. table1_mc_dta2docx using "C:\example Table 1.docx", replace tablenumber("Table 1.") tabletitle("Car characteristics by car type.") footnote("Foreign cars included Audi, Datsun, Toyota and VW cars.")

PS. First announcement was in May 2017 https://www.statalist.org/forums/for...r-a-manuscript

[Marcos Almeida]

Congratulations, really awesome!

I have just two questions:

First, I wonder whether it is possible to add a * sign to p-values, crossed with, say, the information about the respective test on the footnote, such as p = 0.07* and in the foootnote: * = Pearson chi-square; ** Mann-Whitney, etc.

Second, I know we may add the total and get, say, the mean value for the total sample, but I wonder whether we could get a column with the "available" cases according to each variable.

Thank you for this program.

[Mark Chatfield]

Hi Marcos

I'm glad you like the program!

The following option is probably best to use to say what test was used.
test include column describing the significance test used

As for adding available cases, here's the beginnings of how you could do that:
sysuse auto, clear
table1_mc, by(foreign) vars(price conts \ weight contn %5.0f \ rep78 cate) extraspace clear total(before)
browse
rename N_T AvailableCases
drop *T
table1_mc_dta2docx using "C:\example Table 1.docx", replace

Cheers,
Mark

[Marcos Almeida]

Thank you for the tips, Mark. I’m on a trip and, as soon as I come back to ‘my Stata’, I’ll add the options you suggested.

Again, congratulations for such a helpful program.

[Marcos Almeida]

Mark Chatfield : I wonder whether it is possible to abridge a specific command.

For example: instead of typing each variable name, then selecting for each one whether it is, say, "countn", than formating for each one the decimals, is there some way we could select a given specification for a row of variables, as we do when using, say, var1-var50 in "regular" Stata.

I tried with - vars(var1-var100 countn %5.1f), to no avail.

Considering I'd have to type 50 variables plus the selected test plust the format for decimals, it would give quite a big line for the command.

Thanks in advance.

[Mark Chatfield]

Hi Marcos

Not easy for me to do that, but easy for you to do that:

. foreach v of var var1-var100 {
. local vlist "`vlist' `v' contn %5.1f "
. }
. table1_mc, vars(`vlist')

ps There is the format option to set the default for formatting continuous variables which you might like to use.

Cheers,
Mark

[Marcos Almeida]

Hi Mark Chatfield , thank you for the reply and the command. Perhaps there is something missing. I tried it with the example below, and I still keep getting just the first variable:

. sysuse auto
(1978 Automobile Data)

. foreach v of var price-displacement {
  2. local vlist "`vlist' `v' contn %5.1f "
  3. }

. table1_mc, by(foreign) vars(`vlist')

  +--------------------------------+
  | factor   N_0   N_1   m_0   m_1 |
  |--------------------------------|
  | Price     52    22     0     0 |
  +--------------------------------+
   N_ ... #records used below,   m_ ... #records not used
 
  +------------------------------------------+
  |          Domestic    Foreign     p-value |
  |------------------------------------------|
  |          N=52        N=22                |
  |------------------------------------------|
  | Price    6072.4 ()   6384.7 ()    0.68   |
  +------------------------------------------+
Data are presented as mean (SD).

I wonder whether some extra fiddling with the code could solve the problem.

Thanks in advance.

[Mark Chatfield]

Hi Marcos

It just needs the backslash (\) added

. local vlist "`vlist' `v' contn %5.1f "

. display "`vlist'"
would have revealed this.

Cheers,
Mark

[Mark Chatfield]

Sorry, the backslash after %5.1f and before " gets deleted whenever I save

[Marcos Almeida]

Mark Chatfield: Thank you for the helpful command. The added backlash provided the desired output:

. sysuse auto
. foreach v of var price-displacement {
  2. local vlist "`vlist' `v' contn %5.1f \"
  3. }

. table1_mc, by(foreign) vars(`vlist')

  +------------------------------------------------+
  | factor                   N_0   N_1   m_0   m_1 |
  |------------------------------------------------|
  | Price                     52    22     0     0 |
  |------------------------------------------------|
  | Mileage (mpg)             52    22     0     0 |
  |------------------------------------------------|
  | Repair Record 1978        48    21     4     1 |
  |------------------------------------------------|
  | Headroom (in.)            52    22     0     0 |
  |------------------------------------------------|
  | Trunk space (cu. ft.)     52    22     0     0 |
  |------------------------------------------------|
  | Weight (lbs.)             52    22     0     0 |
  |------------------------------------------------|
  | Length (in.)              52    22     0     0 |
  |------------------------------------------------|
  | Turn Circle (ft.)         52    22     0     0 |
  |------------------------------------------------|
  | Displacement (cu. in.)    52    22     0     0 |
  +------------------------------------------------+
   N_ ... #records used below,   m_ ... #records not used
 
  +----------------------------------------------------------------------+
  |                          Domestic          Foreign           p-value |
  |----------------------------------------------------------------------|
  |                          N=52              N=22                      |
  |----------------------------------------------------------------------|
  | Price                    6072.4 (3097.1)   6384.7 (2621.9)    0.68   |
  |----------------------------------------------------------------------|
  | Mileage (mpg)            19.8 (4.7)        24.8 (6.6)        <0.001  |
  |----------------------------------------------------------------------|
  | Repair Record 1978       3.0 (0.8)         4.3 (0.7)         <0.001  |
  |----------------------------------------------------------------------|
  | Headroom (in.)           3.2 (0.9)         2.6 (0.5)          0.011  |
  |----------------------------------------------------------------------|
  | Trunk space (cu. ft.)    14.8 (4.3)        11.4 (3.2)         0.002  |
  |----------------------------------------------------------------------|
  | Weight (lbs.)            3317.1 (695.4)    2315.9 (433.0)    <0.001  |
  |----------------------------------------------------------------------|
  | Length (in.)             196.1 (20.0)      168.5 (13.7)      <0.001  |
  |----------------------------------------------------------------------|
  | Turn Circle (ft.)        41.4 (4.0)        35.4 (1.5)        <0.001  |
  |----------------------------------------------------------------------|
  | Displacement (cu. in.)   233.7 (85.3)      111.2 (24.9)      <0.001  |
  +----------------------------------------------------------------------+
Data are presented as mean (SD).

However, for a reason I cannot envisage, the .docx file gets an error.

We get the results of "make" variable" (and I didn't ask for that in the command), the firts two lines are in bold, and we don't get the results by "foreign", according to the command I typed.

Please see below.

The output in Word is like this:

Table 6. Car characteristics by car type.

AMC Concord	4,099	22	3	2.5	11	2,930	186	40	121	3.58	Domestic
AMC Pacer	4,749	17	3	3.0	11	3,350	173	40	258	2.53	Domestic
AMC Spirit	3,799	22	.	3.0	12	2,640	168	35	121	3.08	Domestic
Buick Century	4,816	20	3	4.5	16	3,250	196	40	196	2.93	Domestic
Buick Electra	7,827	15	4	4.0	20	4,080	222	43	350	2.41	Domestic
Buick LeSabre	5,788	18	3	4.0	21	3,670	218	43	231	2.73	Domestic
Buick Opel	4,453	26	.	3.0	10	2,230	170	34	304	2.87	Domestic
Buick Regal	5,189	20	3	2.0	16	3,280	200	42	196	2.93	Domestic
Buick Riviera	10,372	16	3	3.5	17	3,880	207	43	231	2.93	Domestic
Buick Skylark	4,082	19	3	3.5	13	3,400	200	42	231	3.08	Domestic
Cad. Deville	11,385	14	3	4.0	20	4,330	221	44	425	2.28	Domestic
Cad. Eldorado	14,500	14	2	3.5	16	3,900	204	43	350	2.19	Domestic
Cad. Seville	15,906	21	3	3.0	13	4,290	204	45	350	2.24	Domestic
Chev. Chevette	3,299	29	3	2.5	9	2,110	163	34	231	2.93	Domestic
Chev. Impala	5,705	16	4	4.0	20	3,690	212	43	250	2.56	Domestic
Chev. Malibu	4,504	22	3	3.5	17	3,180	193	31	200	2.73	Domestic

I wonder whether you know what is happening and would kindly provide a solution for this.

Thank you in advance.

[Marcos Almeida]

Mark Chatfield : I tried to use table1_mc_dta2docx in another computer (also Windows 10), and I got the same odd display. I wonder whether this is a bug or I’m missing something.

Thanks in advance.

[Mark Chatfield]

Hi Marcos, if you would like to email me directly (my email address can be found on http://researchers.uq.edu.au/researcher/20374), I can look into this if you haven't solved it yet. Please provide me with all the Stata syntax you used.

[Mark Chatfield]

Hi all
I've uploaded a newer version of table1_mc. Along with one bug fix (N and m now reported correctly when row -rather than the default column- percentages), I've expanded the Remarks section to mention Susan Donath's new -baselinetable- command (Stata Journal 18-2), and a command -tabxml- for use in surveys:

Remarks

table1_mc is an extension and modification of Phil Clayton's table1 command.

Other user written commands that do similar things include:
baselinetable ... very nice, but no p-values. Can use any statistics from summarise, can't report geometric
mean and geometric SD.
tabout ... http://tabout.net.au/docs/home.php
sumtable ... two stats in two (not one) columns, generally not as flexible, no p-values
partchart ... quite similar but strangely doesn't seem to express IQR as (Q1-Q3), reporting instead the
difference between the two quartiles
tabxml ... survey data frequencies and/or % for categorical variables, means and SE / CI for continuous
variables. To avoid matsum error message: . net install sg100, from(http://www.stata.com/stb/stb47)

See also
http://blog.stata.com/2017/04/06/cre...ts-for-arbitrary-variables/

While table1_mc does not report an effect size (e.g. differences in means, medians or proportions) &
associated 95% CI for a variable when there are 2 or more groups, readers might consider calculating and
reporting these also.

[Saranya Seetharaman]

Dear Mark,

Could you please illustrate how to get the row percentages (rather than the default column percentage) for categorical variables? From the example above :
table1_mc, by(foreign) vars(price conts \ weight contn %5.0f \ rep78 cate \ much_headroom bin) ?

Thank you!
Saranya

[Mark Chatfield]

Hi Saranya
The answer is clearly documented in the help file.
. table1_mc, by(foreign) vars(price conts \ weight contn %5.0f \ rep78 cate \ much_headroom bin) catrowperc
BW,
Mark