Thanks, Michael, for your excellent work. I am not sure one can use only
the subject line to make the assessment, but it is a start.
One quick question: Should I take your silence as agreeing that the post
that said essentially 'I know you can do this in stata, but I don't have
time to learn how, so someone on statalist please tell me how to do this
in MSWord' is an off-topic post, a type we should not encourage?
As for your analysis, your conclusion matches exactly my prediction. My
original post said we are early in the diffusion process. It said there
is still time to head off this trend. You seem to find the numbers are
low. Well, just as I said originally, they are low; they are expected to
be low early in the diffusion process.
The one-month check cannot assess whether exponential (or any) growth (or
change) is occuring; a one time-point design cannot assess change.
Many more time points are needed to find out if growth is exponential.
But, we don't need to quibble here. What we are talking about is the
process of socializing people to use resources in particular ways. The
easiest time to do that socialization is when some new thing is low in
incidence. I doubt any serious social scientist could disagree with this
basic conclusion masquerading as an "assertion."
For example, it is easier to tell a few new cellphone users they should
not talk on the phone in the middle of the movie, than it is to wait until
40% of the people in the theatre are on the cell phone to try to get them
to stop. It can be done at that point, but it is harder to do. We
observe the same thing with disease transmission--it is easier to treat a
few carriers early on to prevent an epidemic, than to wait until a large
percentage of the population is infected and then start fighting back.
In this case, it is easier and more productive to have discussion early in
the diffusion process, than to wait until 40% (rather than 1.4%) of the
messages are about off-topic issues to try to reign it back in. That's a
major implication of realizing something is behaving similar to a
diffusion process. And my response, to have the discussion now, is meant
to prevent further diffusion of this behavior. Hence, if it stays at 1.4%
or declines, given that we have now intervened, that won't mean my
prediction was off. Perhaps the discussion had an effect. And, frankly,
that is my hope.
But if it rises, then our inoculation efforts have failed. I submit our
inoculation efforts would have more success if those who used another
package would write to the list, in response to off-topic questions, and
tell the post-er to subscribe to the blah-blah-software list to find an
answer, or in other ways demonstrate the inappropriateness of the posting
to statalist. Not only would this give those users a place to send their
questions, it would serve to show the limits of statalist. In short, 3
and 4-message threads of irrelevance would become fewer 2-message threads.
Maybe asking those who use non-stata products to act to protect statalist
is too much to ask. But, given the importance, it seems worth at least
asking.
Respectfully,
Sam
On Thu, 1 Jul 2004, Michael Blasnik wrote:
> Just a quick follow-up. I took a look at the June 2004 Statalist archive ,
> which was a fairly heavy month (in my opinion) for MS Word related posts. I
> counted one 4 post thread about simply how to copy a graph from Stata to
> Word and an 8 post thread (which I participated in) about copying many
> graphs between Stata and MS Word (including how to structure Stata code to
> produce many graphs). Several of those posts were people asking details and
> thanking posters. I guess it was those two threads that prompted a 3 post
> thread suggesting the creation of a faq about working with MS Office and
> Stata. There was also a 2 post thread on exporting output to Excel (but
> applies to any spreadsheet) and one on copy graphs to officewriter running
> on Linux.
>
> Anyway, excluding the thread about creating a faq, I counted 12 MS Word
> messages in June. There were 878 messages posted, giving us a whopping 1.4%
> of messages in June devoted to MS Word, or 1.6% devoted to MS Office if you
> count the 2 Excel posts. Given that there were only 2 MS Word threads, I'd
> say they received a level of interest from Stata users as least as large as
> many other topics.
>
> If this is exponential growth, we're pretty early on the curve.
>
> Michael Blasnik
> [email protected]
>
>
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
