Tuesday, April 28, 2026

Tips on how to efficiently ask a query on Statalist


As everybody is aware of, I’m an enormous proponent of Statalist, and never only for egocentric causes, though these causes play a task. Almost each member of the technical employees at StataCorp — me included — are members of Statalist. Even once we don’t take part in a selected thread, we do listen. The discussions on Statalist play an necessary position regarding Stata’s growth.

Statalist is a dialogue group, not only a question-and-answer discussion board. Nonetheless, new members typically use it to acquire solutions to questions and that works as a result of these questions generally change into gist for subsequent discussions. In these instances, the questioners not solely get solutions, they get way more.

Among the finest options of Statalist is that, irrespective of how poorly you ask a query, you’re unlikely to be flamed. Not solely are the members of Statalist good — simply as are the members of most lists — they act simply as good on the record as they are surely. You might be unlikely to be flamed in the event you ask a query poorly, however you’re additionally unlikely to get a solution.

Right here is my recipe to extend the possibilities of you getting a useful response. You must also learn the Statalist FAQ earlier than writing your query.

Topic line

Make the topic line of your electronic mail significant. Some good topic strains are:

Survival evaluation

Confusion about -stcox-

Sudden error from -stcox-

-stcox- output

The primary two sentences

The primary two sentences are an important, and they’re the simplest to write down.

Within the first sentence, state your downside in Stata phrases, however don’t go into particulars. Listed below are some good first sentences:

I’m having an issue with -stcox-.

I’m getting an surprising error message from -stcox-.

I’m utilizing -stcox- and don’t know interpret the outcome.

I’m utilizing -stcox- and getting a outcome I do know is unsuitable, so I do know I’m misunderstanding one thing.

I need to use -stcox- however don’t know begin.

I believe I need to use -stcox-, however I’m uncertain.

I need to use -stcox- however my information is sophisticated and I’m uncertain proceed.

I’ve an advanced dataset that I one way or the other want to remodel right into a kind appropriate to be used with -stcox-.

Stata crashed!

I’m having an issue which may be extra of a statistics subject than a Stata subject.

The aim of the primary sentence is to be a focus for members who’ve an curiosity in your subject and let the others, who had been by no means going to reply you anyway, transfer on.

The second sentence is even simpler to write down:

I’m utilizing Stata 11.1 for Home windows.

I’m utilizing Stata 10 for Mac.

Even in case you are sure that it’s unimportant which model of Stata you’re utilizing, state it anyway.

Write two sentences and you’re accomplished with the primary paragraph.

The second paragraph

Now write extra about your downside. Attempt to not be overly wordy, but it surely’s higher to be wordy than curt to the purpose of unclearness. Nevertheless you write this paragraph, be express. For those who’re having an issue making Stata work, inform your readers precisely what you typed and precisely how Stata responded. For instance,

I typed -stcox weight- and Stata responded “information not st”, r(119).

I typed -stcox weight sex- and Stata confirmed the standard output, besides the usual error on weight was reported as dot.

The type of the second paragraph — which can prolong into the third, fourth, … — is determined by what you’re asking. Describe the issue concisely however fully. Sacrifice conciseness for completeness in the event you should otherwise you assume it would assist. To the extent attainable, simplify your downside by eliminating extraneous particulars. As an illustration,


I've 100,000 observations and 1,000 variables on companies, however 4
observations and three variables might be sufficient to point out the issue.
My information appears to be like like this

        firm_id     date      x
        -----------------------
          10043       17     12
          10043       18      5
          13944       17     10
          27394       16      1
        -----------------------

I would like information that appears like this:

        date    no_of_firms   avg_x
        ---------------------------
          16              1       1
          17              2      11
          18              1      12

That's, for every date, I would like the variety of companies and the
common worth of x.

Right here’s one other instance for the second and subsequent paragraphs:


The substantive downside is that this:  Sufferers enter and go away the 
hospital, generally greater than as soon as over the interval.  I believe  
on this case it will be acceptable to mix the 
separate stays so {that a} affected person who was in for two days and 
later for 4 days may very well be handled as being merely in for six days,  
besides I additionally file what number of separate stays there have been, too.

I am evaluating value, so for my functions, I believe treating 
value as proportional to days in hospital, no matter their
distribution, might be sufficient.  I am  whole days as a
perform of variety of stays.  The concept is that letting sufferers out
too early ends in a rise in whole days, and I need to
measure this.

I understand that extra stays and days may also come up just because
the affected person was sicker.  Some sufferers die, and that clearly 
truncates keep, so I've omitted them from information.  I've illness
codes, however nothing about well being standing inside code.  

Is there a strategy to incorporate this added info to enhance  
the estimates?  I've acquired a number of information, so I used to be pondering of 
utilizing loss of life charge inside illness code to one way or the other rank the codes 
as to seriousness of sickness, after which utilizing "seriousness" 
as an explanatory variable.  I assume my query is whether or not  
anybody is aware of a manner I would do that. 

Or is there someway I may estimate the mannequin seperately inside
illness code, one way or the other constraining the coefficient on variety of 
stays to be the identical?  I noticed one thing within the handbook about 
stratified estimates, however I am uncertain if this is similar factor.

You’re asking somebody to speculate their time, so make investments yours

Earlier than you hit the ship key, learn what you could have written, and enhance it. You might be asking for somebody to speculate their time serving to you. Assist them by making your downside simple to know.

The simpler your downside is to know, the extra probably you’re to get a response. Mentioned in a different way, in the event you write in a disorganized manner in order that potential responders should work simply to know you, a lot much less offer you a solution, you’re unlikely to get an response.

Glowing prose isn’t required. Correct grammar isn’t even required, so nonnative English audio system can chill out. My recommendation is that, until you’re typically praised for the way clearly and entertainingly you write, write brief sentences. Group is extra necessary than the model of the person setences.

Keep away from or clarify jargon. Don’t assume that the one that responds to your query might be in the identical area as you. When coping with a substantive downside, keep away from jargon apart from statistical jargon that’s widespread throughout fields, or clarify it. Potential responders prefer it while you educate them one thing new, and that makes them extra more likely to reply.

Tone

Write as in case you are writing to a colleague whom you understand effectively. Assume curiosity in your downside. The identical factor mentioned negatively: Don’t write to record members as you may write to your analysis assistant, worker, servant, slave, or member of the family. Nothing is extra more likely to to get you ignored than to write down, “I’m busy and actually I don’t have time filter by all of the Statalist postings, so reply to me straight, and shortly. I would like a solution by tomorrow.”

The constructive strategy, nonetheless, works. Simply as when writing to a colleague, typically you don’t want to apologize, beg, or play on sympathies. Typically after I write to colleagues, I do really feel the necessity to clarify that I do know what I’m asking is foolish. “I ought to know this,” I’ll write, or, “I can’t bear in mind, however …”, or, “I do know I ought to perceive, however I don’t”. You are able to do that on Statalist, but it surely’s not required. Often after I write to colleagues I do know effectively, I simply bounce proper in. The identical rule works with Statalist.

What’s acceptable

Questions acceptable for Stata’s Technical Providers aren’t acceptable for Statalist, and vice versa. Some questions aren’t acceptable for both one, however these are uncommon. For those who ask an inappropriate query, and ask it effectively, somebody will often direct you to a greater supply.

Who can ask, and the way

You could be part of Statalist to ship questions. Sure, you possibly can be part of, ask a query, get your reply, and stop, however in the event you do, don’t point out this on the outset. Checklist members know this occurs, however in the event you point out it while you ask the query, you’ll sound superior and condescending. Additionally, stick round for a couple of days after you get your response, as a result of generally your query will generate dialogue. If it does, you must take part. You need to need to stick round and take part as a result of if there’s subsequent dialogue, the ultimate result’s often higher than the preliminary reply.

I’ve beforehand written on be part of (and stop) Statalist. See http://weblog.stata.com/2010/11/08/statalist/.



Related Articles

Latest Articles