cmoore@brl.mil (10/17/90)
Some telecom items, in this newspaper excerpt referring to an opinion poll. New York Times, Oct. 14, 1990, page 25 column 1, first 3 paragraphs only. How the Poll Was Conducted "The latest New York Times/CBS News Poll is based on telephone interviews conducted Oct. 8-10 with 960 adults around the nation, excluding Alaska and Hawaii. "The sample of telephone exchanges called was selected by a computer from a complete list of exchanges in the country. The exchanges were chosen so as to insure that each region of the country was represented in proportion to its population. For each exchange, the telephone numbers were formed by random digits, thus permitting access to both listed and unlisted numbers. The numbers were then screened so that only residences would be called. "The results have been weighted to take account of household size and number of residential telephone lines and to adjust for variations in the sample relating to region, race, sex, age and education."
oplinger@sol.crd.ge.com (B. S. Oplinger) (10/19/90)
cmoore@brl.mil describes the process for a NY Times/CBS News poll: How pray tell can they have generated 'telephone numbers ... formed by random digits, thus permitting access to both listed and unlisted numbers' and then caused them to be 'screened so that only residences would be called?' Is there some magic way to tell if a number is residential or commercial, especially the unlisted ones. Or is this simply a case of a newspaper article mixing facts and fiction? brian oplinger@crd.ge.com <#include standard.disclaimer> [Moderator's Note: I think they made the assumption (mostly correct) that business phones would probably not be non-pub; thus in the process of sorting out who to call and who not to call, all non-pubs were assumed to be residential for the purpose of filing the number in one compartment or another. PAT]
friedl@mtndew.tustin.ca.us (Stephen Friedl) (10/21/90)
> [Moderator's Note: I think they [NYT] made the assumption (mostly correct) > that business phones would probably not be non-pub; thus in the > process of sorting out who to call and who not to call, all non-pubs > were assumed to be residential for the purpose of filing the number in > one compartment or another. PAT] I'm not so sure about this. Virtually all business have many more than one telephone line, and only the main "entry point" numbers will be published -- it's the default for additional lines. Four out of my five telephone lines at home are non-pub, but I certainly don't have an "unlisted" number in the traditional sense. Stephen J. Friedl, KA8CMY / I speak for me only / Tustin, CA / 3B2-kind-of-guy +1 714 544 6561 / friedl@mtndew.Tustin.CA.US / {uunet,attmail}!mtndew!friedl [Moderartor's Note: Please see my message in the last issue. Non-pub is not the same as not-listed. If I cross check all the supplementary lines into your business, they will still show up with your business name even though they won't all appear in the directory. PAT]
jeh@dcs.simpact.com (Jamie Hanrahan) (10/21/90)
In article <13818@accuvax.nwu.edu>, oplinger@sol.crd.ge.com (B. S. Oplinger) writes: > cmoore@brl.mil describes the process for a NY Times/CBS News poll: > Is there some magic way to tell if a number is > residential or commercial, especially the unlisted ones. Or is this > simply a case of a newspaper article mixing facts and fiction? I think the latter. I was at a friend's house when they received a survey call. There was some confusion because this house has two lines, one private and one business. When the survey folks learned that they had called the business line, they didn't want to talk further. In this case, they were just calling every randomly-generated number and asking. Jamie Hanrahan, Simpact Associates, San Diego CA Internet: jeh@dcs.simpact.com, or if that fails, jeh@crash.cts.com Uucp: ...{crash,scubed,decwrl}!simpact!jeh
brian@ucsd.edu (Brian Kantor) (10/22/90)
>[Moderator's Note: I think they made the assumption (mostly correct) >that business phones would probably not be non-pub; thus in the >process of sorting out who to call and who not to call, all non-pubs >were assumed to be residential for the purpose of filing the number in >one compartment or another. PAT] Hearty Guffaw! Here at UCSD we have about 200 dial-in modem lines, of which some 75 or so are in the same prefix (558) as those assigned to the student housing, and another 20+ in the 452 prefix, which is split between homes in the surrounding community, and various small centrexen serving nearby firms. Our modem numbers are NOT en-bloc, since we add lines typically 8 or 16 at a time as demand grows, and it often seems that we often get all the numbers ending in odd digits. We only publish the pilot numbers for the six hunt groups these lines all belong to (each is a different grade of modem service). All the rest are non-pub, as are most business centrex numbers, so such a survey (or solicitation attack) would wind up targeting a whole lot of business and modem lines among the residences. I don't think they would be able to tell which is which ahead of time; it's a real mixed bag. This is beside the 534 prefix, which we completely own, and 543 and 294 each of which is about half ours. In a year or so, if campus telecommunications removes all PacBell service from the dorms and replaces student phones with campus lines, as they've been discussing, we'll probably own half of another prefix as well - about half our students live in campus housing. I suspect the newspaper's screening process involves a lot of apologizing to inappropriate dialees. That and hanging up on machines. (Actually, it's pretty easy to tell which are our dial-in modem lines. They're the ones that are busy all evening long!) Brian [Moderator's Note: You refer to those numbers as non-published, but I think you mean 'non-listed' or 'not listed in the directory'. The difference between these two conditions is the one is unavailable, period, and the other, while not listed in the directory will still usually show up in a cross-reference directory, typically with just the company name and no address given, or a reference to the lead number in the group. I think people preparing the list of 'random' numbers for these polls do first select randomly, then use criss-cross directories to backtrack them into residence, business or payphone categories. So if they pick (for example) random number 708-491-1234 in Evanston, IL, looking at the criss-cross will show only a relative handful of entries under 708-491. No where near the 9000+ numbers the exchange might have are listed. But the several dozen that are listed all say 'Northwestern University such and such' ... so the random number compiler assumes the whole exchange is probably Northwestern. Likewise if the criss-cross directory lists a number with a business name and then skips the next two dozen numbers before it starts listing again, its reasonable to say the numbers following are linked to the first one. For exchanges that go on page after page in the criss-cross listing one residence after another, a skipped entry is probably a residence with a non-pub number. But it is not that often that a business will have its main incoming number non-pub. PAT]
gwangung@milton.u.washington.edu (Roger Tang) (10/22/90)
In article <13818@accuvax.nwu.edu> oplinger@sol.crd.ge.com (B. S. Oplinger) writes: !cmoore@brl.mil describes the process for a NY Times/CBS News poll: !How pray tell can they have generated 'telephone numbers ... formed by !random digits, thus permitting access to both listed and unlisted !numbers' and then caused them to be 'screened so that only residences !would be called?' Is there some magic way to tell if a number is !residential or commercial, especially the unlisted ones. Or is this !simply a case of a newspaper article mixing facts and fiction? ![Moderator's Note: I think they made the assumption (mostly correct) !that business phones would probably not be non-pub; No, they don't do this. Based on my days as a telephone survey taker (NOT a solicitor!), we ASKED (or added two plus two when somebody answered, 'Joe's Pizza.'). There's all sorts of methodological screening techniques; some of theme are quite sophisticated. This one, however, really just requires the brains of an avocado.