Tuesday, April 30, 2013

Results from new Survey of Climate Scientists coming soon

If only you knew how shocked I was! Scouting through the internet I recently came across a blog posting titled Lewandowsky: study “Useless” unless authors demonstrate “data integrity”, at http://climateaudit.org/2012/09/12/lewandowsky-study-useless-unless-authors-demonstrate-data-integrity/.  The blog posting is written by Steve McIntyre.  Apparently Lewandowsky, in reference to the 2003 Survey of Climate Scientists, accuses me of violating “all internet survey methodological standards by not recording dates, times, and IP numbers of respondents”.
 Just for the record, I would like to point out that many people working in institutes have no choice but to use the static IP of the institute’s firewall.  I doubt they are aware or even care.  But this would mean that if I sampled 100 scientists at 10 institutes, likely in the recorded responses there would be only 10 IP addresses.  Also, it is not difficult to manipulate time and date on a PC. So, please, for all of the conspiracy theory (of any persuasion) critics of the surveys of climate scientists, try to know a little of what you speak before you speak.  Apparently Lewandowsky complained that there was “no way to check or verify the integrity of the data” and therefore the data was probably “useless”. Lewandowsky said that the study should not have been published “without the authors demonstrating the integrity of their data”.  Steve McIntrye goes on to point out that the work of Lewandowsky is not without blemish.

So, a brief note to Mr. Lewandowsky and kind,

Well, Mr. L. the results of the survey conducted in late 2012 will be released soon, so you and the crew at Deltoid  (http://scienceblogs.com/deltoid/) can again amuse yourselves conjuring up this or that conspiracy story.  Too late now but maybe I should have asked
Question 1.   Are you   1 a skeptic     2. an alarmist

Question 2.  How many copies of this survey do you intend to submit?
Next time.

I interact with climate scientists most days in one way or another.   Most are conscientious people who have personal opinions about what they are working on (they are human) but do not conscientiously go out of their way to cheat to try and get a point across.  No doubt in the grand scheme of climate science there are few in the upper echelons who might stretch a point or two when the opportunity arises for some publicity, but they likely constitute a very small minority (and more and more are being taken less and less seriously in the grand scheme of things external to the inner world of climate change).  But I often have to wonder about the minds and integrity of all of the non-contributory constant critics.  So, while you await the new results, perhaps you can dream up some more devious plots to test against the data, in the case that the results do not match your personal opinion.  In the meantime, I will prepare a report of what was actually reported.  Don't worry about the integrity of the data.  This time the survey was embedded with a Send-Tracking-Agent-Signature-Integration protocol, a tried and true means of obtaining the 'truth'.

4 comments:

None said...

"Just for the record, I would like to point out that many people working in institutes have no choice but to use the static IP of the institute’s firewall. I doubt they are aware or even care. But this would mean that if I sampled 100 scientists at 10 institutes, likely in the recorded responses there would be only 10 IP addresses. Also, it is not difficult to manipulate time and date on a PC. So, please, for all of the conspiracy theory (of any persuasion) critics of the surveys of climate scientists, try to know a little of what you speak before you speak."

Regarding the IP address point, it should be recorded anyway. Then you've got 10 responses, at different times, with the same IP address coming through a proxy from a university and you can say "these could conceivably be different people as it's coming through a uni proxy. You could also say, if 10 results were coming through a private address IP, all submitted within 5 minutes, you can say "this data is suspicious"; particularly if the data represents outliers.

Regarding the timestamping issue, the timestamps are taken from the server when the data is received, so time on the machine from which they were submitted is irrelevant.

FWIW Lewandowsky's pieces are the worst "peer reviewed literature" ever brought to my attention. It's a joke they got published (if indeed his second piece gets published at all).

Anonymous said...

Lewandowski an die Latte (51. min)! Aargh. Oh, ich verwechsle gerade etwas, Lewandowsky ist heute nachrangig. ;-)

Andreas

Anonymous said...

Prof. Lewandowsky?

He just got kicked upstairs ... the Royal Society has decided to award him the Wolfson Research Merit award for his "outstanding achievement and potential" ...

http://bishophill.squarespace.com/blog/2013/4/28/paul-and-the-pug-dog.html

http://joannenova.com.au/2013/04/royal-society-calls-lewandowsky-outstanding-gives-him-money-loses-more-scientific-credibility

Learning from Lewandowski means learning to win ... Lysenkoism in full blossom.

V. Lenzer

ghost said...

ah, Vergleich mit Massenmördern.... Man, man, man... manchmal frage ich mich wirklich.

@Dennis

das war doch eine lustige Studie, aber ich finde, in den Kommentaren der Wissenschaftler wird doch oft deutlich, dass sie viele Schwächen hatte. "useless" ist dennoch unfair. (Ihr letzter Satz im Post aber genauso...)

Mein Problem ist, dass Klimaforschung doch ein sehr großes Gebiet ist, und Meinungen sehr unterschiedlich gewichtet werden müssten bzw. gleiche "Punktzahl" völlig unterschiedliche Aussagen sein können. Ich glaube, die Umfrage war zu generell. Ich fand auch die Fragen etwas verquer.

Gut fand ich, dass ich doch ein bisschen sehen konnte, wer gefragt wurde, und warum Antworten zustande kamen. Wenn ich als Laie wissenschaftliche Aussagen "evaluieren" will, soweit ich es kann, sehe ich immer zwei Hauptmöglichkeiten:

1. Argumente für eine Meinung angucken (soweit ich kann)
2. die Vertrauenswürdigkeit und "gefühlte" Kompetenz der Person, Organisation, Quelle usw. ansehen (geht einfacher)

Beides geht bei einer solchen Studien nicht so richtig, finde ich. Deswegen habe ich so meine Schwierigkeiten damit.

Naja, darum geht es hier ja nicht... oder?