Sign in or Join FriendFeed
FriendFeed is the easiest way to share online. Learn more »
Walt Crawford

Walt Crawford

Mostly retired library person/researcher/writer/speaker. All original FF contributions CC0 (public domain).
Classic example of Business Insider at its best: A headline implying that Gartner says PCs are dying (which Stephen A. gladly repeats)...but, of course, if you click through to Gartner it actually says PC growth is slowing. Not *precisely* the same as dying....
Unsurprising side-effect of checking Gold OA journals published by (some) colleges & universities: sidebar ads on various other sites extolling the virtues of attending those colleges & universities, especially the private ones.
Serious question and quandary: Someone hearing about the new OA journal analysis I'm doing says, in essence, "I assume you'll make the spreadsheet available for crowdsourced verification and further analysis." I do not see this as being a good idea, and if pushed on it would be inclined to drop the project ...
I'd hate to do that at this point, since I've already looked at (individually) 7,818 out of around 11,500 journals and "journals" in the project. But I'm using a variety of approximation techniques to make the project less overwhelming, and I can just see, e.g., Beall downloading the spreadsheet, finding that my numbers (which will NEVER be used citing individual journals) are off for two or three journals, and... - Walt Crawford
... proclaiming "See? I said Crawford's stuff was bilge." (He's already used that word.) Whereas, of course, JB simply avoids facts, sticking with "I suspect" and "It seems unlikely" and the like. - Walt Crawford
Anyway, if it's really Improper for me to do this work--which will lead, as in the July '14 C&I, to discussions of patterns, NOT of individual journals--and NOT make the spreadsheets available, then I may very well give it up, blowing seven weeks of effort but saving 4-6 more weeks. - Walt Crawford
So: the question: Am I being unreasonable? If I'm unwilling to make the spreadsheets available, should I just drop the project? Advice very much wanted. (As I go off and do the next 10 journals...) - Walt Crawford
It's a good open data question. I can see a lot of value in the release of the data--if you're clear both a readme file and the final document about how/what you've summarized and approximated. JB's going to be skeptical anyway;he's already decided against you, I would try not to make this decision based on him (which is hard!). In my mind sharing the data gives your work/words more... more... - Hedgehog
You are in no way obliged to. I think it would be interesting and useful to others though. Whether you do or don't, you should be transparent about your approximation techniques. Then if you do make the dataset open and people complain a number is off you can just say "Well duh, see how I talked about approximation techniques? Feel free to make a better dataset if you've got the months/years of time spare." But if you don't want to make the dataset open, that doesn't mean the rest of the project is useless. - Deborah Fitchett
I think my reaction might very well be the same. The spreadsheets weren't "designed" for crowdsourcing, and the whole thing makes me nervous as hell. Of course, I could just blank out all the journal names, but that would make the spreadsheets close to worthless... - Walt Crawford
Deborah: My response was to Hedgehog. Yes, I do plan to discuss the shortcuts/approximations used, as I normally would. Thanks for the additional thoughts. - Walt Crawford
Is there a small representative sample that you could show to explain how you evaluated the data, and to explain how some people might interpret some things differently, and then say that the big spreadsheets are available on request? But then I wonder if JB would bother to ask for the data. Would he do that so he could nitpick? - Joe
Joe: Yes, if I go to the extra work on a sample--although there are two issues, evaluation (is this an A, B, or C?) and actual counting (where I've used shortcuts at times). In either case, then I'd need to (a) prepare a readme in some detail and (b) respond to requests. Grant funding for this project has been notoriously lacking... - Walt Crawford
I guess what I'm saying is "Is it irresponsible to not make the data available?" in which case maybe the best bet is to shut down the project. Deborah's saying it's *not* irresponsible; I appreciate that. - Walt Crawford
bump: I could really use a little more feedback on this, if anybody else is interested. - Walt Crawford
It's much easier saying what you should do with your data than doing it with my data. I totally get what you're saying. My thoughts were: I didn't document enough, my code (or calculations) done in Excel were done the hard way, the data are dirty, maybe someone else will see something in there I missed.... But as not you, I really think you should try to release under a cc-0 license or... more... - Christina Pikas
also, you are looking for sponsorship or more publishing opportunities, I think, and this could bring more attention and respect for your work. More trust from the people you want to attract. - Christina Pikas
oh and it's not irresponsible to not share. but it is generous to share. i doubt beall would take the time and effort to re-do your analysis. I really do. And if he does, so what! - Christina Pikas
Christina: Thanks for that. The odds of sponsorship are so low that I've mostly given up, and I've cut back on publishing opportunities for the moment. I'll have to think about this. - Walt Crawford
So its a bit late now but I always now try to figure out at the beginning of an analysis what can and can't be (easily) shared and bear that in mind as I work on something. I would always err on the side of release, even where its messy because the upsides generally are larger than the downsides. Problem is the upsides are all less noisy than certain downsides. If it were me, under... more... - Cameron Neylon
Turns out the person suggesting release is actually suggesting release of the spreadsheet(s) used in the July article--and alternatively suggests that I recruit a small team of experts who could reanalyze/deepen the analysis. Not likely to happen, honestly. I'm now satisfied that it's not reprehensible/irresponsible to *not* release the spreadsheets. Beyond that...well, I'll keep thinking about it. Thanks all for your help. - Walt Crawford
[Curiously, just reading a long discussion involving "reanalysis" of data, in an entirely different field, and it doesn't encourage me: basically says that the only reason people reanalyze is to attempt to refute] - Walt Crawford
… in related news, this conversation is useful for us as we think about trying to persuade our colleagues to release their data. it nicely illustrates the pros & cons of data sharing in our discipline. And I agree with all the points listed above. :-/ - Stephan!e•CogSc!L!brar!an
Bumping only because I just did a blog post about this... - Walt Crawford
Hmmm...interesting. The clinical trial analogy isn't a bad one. Also makes you realise that if you 'anonymised' the dataset then re-identification wouldn't be hard. I think there is some value in such a dataset (looking at correlations, subsets etc) but given the re-identification risk the work involved in effective anonymisation might be more than its worth. - Cameron Neylon
Thanks, Cameron. Others? (I should probably publicize the blog post on Google+, since OAfolk seem to hang out there. Will do so.) (Have now done so.) - Walt Crawford
I could see value in releasing *just* the names (and maybe URLs if you're feeling generous) of the journals/publishers you looked at, with NO accompanying information -- well, maybe column headers, but no more than that. That way, if we (writ large; it's a tempting crowdsourcing project) want to do the reanalysis, we can gather our own damn data and it's not on you if we do. No obligation, just a thought. - RepoRat
As I noted elsewhere, I'm perfectly willing to do that. (For the Beall and OASPA lists, the URLs "come with"--they're hyperlinked journal and/or publisher names. For the DOAJ list, which is of course just a subset of the downloadable DOAJ dataset as of May 2014, they're a separate column, but leaving 'em in is no more work than taking 'em out. For Beall, some fraction of journal cells would *not* have hyperlinks because of problems in downloading title lists from the publishers.) - Walt Crawford
Open Data, Crowdsourcing, Independent Research and Misgivings -
Perfect sentence in introductory essay for a journal from a semi-accredited (U.S.) university: "Special thanks go to Dr. [redacted] for his tiresome support of the journal." That support certainly can get tiresome...
gotta love this: a new Denny's in NYC with a "Grand Cru Slam"--two Grand Slam breakfasts and a bottle of Dom Perignon--for a mere $300. And I'll bet they sell a few of them.
Just got an email (from a vendor I explicitly allow them from) re a "Huge Distressed Cruise Inventory Sale." I know what they mean, but somehow booking a distressed cruise isn't high on my list...
Apropos of nothing much, it appears that Romania has a very well-developed English-language gold OA journal collection (mostly no-fee), at least in humanities & social sciences.
For anyone who cares (and sees the little pictures next to people): the new one was taken August 20, 2014, in Morgan Territory Regional Preserve--one of the many hiking spots near here. (I wear a floppy gardening hat because I have a fat head, making more elegant sun-protecting hats difficult/expensive.)
Are all these damn phone calls where there's never anybody at the other end of the line a conspiracy by cell-phone companies to make us give up landlines? Or something more nefarious?
If you've already downloaded Cites & Insights 14:9, you may want to do so again--I've changed one paragraph (where I managed to divide British pounds by 1.7 to get U.S. dollars rather than multiplying them). Here's the actual change:
Correction in Cites & Insights 14:9 -
Flashback thanks to PUBLIB (a discussion of receipt printers): I never realized that Star Micronics--makers of the first (dot matrix) printer I had at home--is still in business (albeit not making printers for PCs). They made solid gear, and based on the PUBLIB comment I guess they still do. Fifty years: not bad in tech.
Cites & Insights 14:9 (September 2014) available -
If you haven't already, go read Wayne Bivens-Tatum's latest "Peer to Peer" at Library Journal. Highly recommended!
I had missed the "pure bilge" comment on WBT's column on the first "The Big Deal..." and am now thinking I should use it as a new subtitle for Cites & Insights. (Or maybe not.) It's good to be pure. - Walt Crawford
Old-guy technology grump 1: Forced to use Yahoo!mail (for freecycle), encountering old-style idiot reply defaults (tendency to wind up replying to yourself). Gmail handles this logically (you probably *don't* want to reply to yourself) and I'd forgotten that idiot email systems worked this way.
Old-guy technology grump 2: Yesterday morning, looking at hike pictures from Gmail, Firefox properly opened new tab for the pictures...but wouldn't close the tab. Wound up with six tabs, none of which would close (using any of several techniques). "Solution": Closed Firefox, reopened it, same six tabs but now they'd close. Still wonder what happened. - Walt Crawford
Not surprisingly, the San Francisco Chronicle is running more pieces about Robin Williams, some pointing out that he was an actor and comedian, not a comedian-actor (he trained at Juilliard), and the local movie critic's list of ten best films (headed by Moscow on the Hudson). Too many to link to; just go to
Warning: At least on my aging computer running Firefox, seems to use a boatload of system resources and slows things down something awful--including getting away from it. - Walt Crawford
An hour ago, on the way to lunch, hearing NPR's "Here and Now" interviewing a hero, a doctor (Harvard grad, professor, high-profile) who travels to Mississippi to do abortions. He's deeply religious and grew up anti-abortion, but after a lot of thought concluded that the deeper calling required him to do them.
Here's the thing: The host almost seemed to be grilling him, with an attitude that seemed to imply "How can a Christian POSSIBLY be doing abortions?" Maybe I was overreacting, but it struck me as, well, very non-NPRish. (I do miss Talk of the Nation; Here and Now is a sad alternative.) - Walt Crawford
Esquire did an incredibly good piece on him: - laura x
I think the reporter who wrote that profile was also on, but by the time I reached Canton Village, the NPR person was still doing her seemingly-hostile questioning. - Walt Crawford
I think NPR gets so much flak for being "liberal" that they periodically get into a thing where they think, "Oh, here, we can show what hard-asses we are!" Bleah. I still remember how appalling I found Neal Conan on 9/11. - laura x
The rarity with this doctor is he is a board certified ob/gyn. Most abortionists are not nor have had any true operating room experience other than their few months of rotation. I have my own personal beliefs but applaud this man. - Janet
Laura x: That sounds about right. Janet: It would be interesting to have proof of that "most abortionists..." but that's another issue. - Walt Crawford
Walt, I have first hand knowledge of the abortion trade, I used to do ultrasound demonstrations. The Planned Parenthood doctors are mostly farmed right out of medical school. There are a private physicians of the ob/gyn board certified type. Sadly being legal does not always mean as safe as the public would assume. Kermit Gosnell's clinic is not alone. Steve Brigham is a bad boy Planned Parenthood finally fired. - Janet
I'll assume you're in favor of improved/expanded training for abortion providers (and those interested in providing abortions), so that safety is ensured :) - Meg VMeg
Meg: I certainly am (can't speak for Janet), and would note that PP would probably be a safer place if clinics weren't under attack all the time. (Of course, PP does a lot more than provide abortions, but that's another issue, and yes, we're members.) - Walt Crawford
While undoubtedly there are uncertified and unlicensed health care providers and outright illegal clinics, if you're a legitimate health care provider, doing surgical abortions without training and/or certification and/or proper back-up seem like a great way to get nuked by a malpractice lawsuit and having your carrier drop you hard and getting banned from ever receiving Medicare or... more... - Victor Ganata
Absolutely Meg, there are more regulations on nail salons and restaurants than in the abortion industry. The likes of Gosnell and Brigham would have been shut down before so many women were permanently injured. - Janet
Another meaningless musical post -
Anyone know of a way to get back a specific Special Offer on a Kindle? (The ones that come up when you first turn it on, if you've saved $20 by buying the special offer models.) Twice now my wife's seen possibly-interesting Kindle book offers, but gone past them and been unable to get back.
Natureally, I’m delighted -
Bookending August with live musical theater at the Bankhead. Yesterday was Shrek: The Musical, Tri-Valley Rep (all volunteer). Good, except orchestra too loud (so we missed some of the lyrics) and kids miked with too much treble (which young voices already have too much of). End of month: Pirates of Penzance, Lamplighters.
We know from experience that the Lamplighters won't deafen us and that they'll do a first-rate job. Shorter on fart jokes, though. - Walt Crawford
Am I wrong to feel irritable when a free service, where I was encouraged to sign up, says it's going to charge (I've never really used it)--and the site offers no way to say "Remove my account"? (I've sent support email. Clearly, they won't/can't actually charge me, but I don't care for the Second Life "Hotel California" model.)
Not naming the service for now, because they seemed sincere about trying to do the right thing. - Walt Crawford
They took care of this rapidly, so I still won't name the service (which I suspect is an excellent one for some folks, just not for me). - Walt Crawford
I'm not sure they've completely thought through all that and they didn't set it up that way from the beginning so they're trying to retrofit? I'm not sure what i'm going to do because i really can't pay that right now, but I want to and I don't want a waiver. - Christina Pikas
I'm trying to decide because most of the things I have on it aren't 'traditional scholarly' enough for it to really understand so I have conference papers labeled as articles, articles labelled as unknown, all the things labelled as webpages, and a whole failure of deduplication. Which isn't that the system sucks so much as that my stuff's a mess, but I can't manually fix it. --I am more likely to go with the annual subscription model they've just announced. - Deborah Fitchett
Another tech pundit (Dan Tynan, I think) saying in a few years we'll look back at keyboards & mice as quaint, the way we do rotary phones & fax machines now. Right. 'Cuz you're going to write extended text using your fingers in the air, I guess. The death of keyboards has been projected almost as often as the death of books.
Heard a few minutes of Mr. "Six Californias" on NPR today. Maybe he should stay off the radio: It sounded like a pure example of very wealthy privileged Native California nostalgia gone wild: Back in the day when California had great education, was the best place to start a business, had no poverty and wonderful weather...
..before everything went to hell because Gummint. But six Gummints could fix it all. Sure they could. - Walt Crawford
Note that I'm a native Californian although never very wealthy. On the other hand, I grew up in one of those Californias he really wants to get rid of--you know, the one that grows the nation's food and has a high poverty rate, because Food. - Walt Crawford
Beautiful hiking day (a little rain yesterday and LOTS of wind cleared some of the muck out of the air), and we did "Brushy Peak counterclockwise." Which in this case meant that for the Ramblers, the intermediate group, instead of a nice pleasant 4 miles with maybe 500' rise, we had *exactly* the same hike as the Scramblers: 6.35 miles...
...and 1,350' rise. Still, pleasant if tiring. (Brushy Peak is so-called because the top is covered with trees, whereas most local hills are bare most of the time. It also has a long history, indicated in part by the many grinding rocks.) - Walt Crawford
And, TIL: I always figured Brushy Peak was a tall hill, not a short mountain. It's 1,702' (we get within a few feet of the peak, because we start out several hundred feet above sea level). Turns out that's a hill in the UK and Ireland, a mountain by old American usage. I think 2,000' or 600m is a useful lower limit for an actual mountain. - Walt Crawford
Other ways to read this feed:Feed readerFacebook