Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Predicting the webs future
- Umar A Ahmad (26 December 2007)
-
Reviewer Bias
- Bruce W. Thomas, Wichita, Kansas 67208 USA (28 December 2007)
-
Wiki for Executable English / Drug Database
- Adrian Walker (7 January 2008)
|
|
|||
|
Umar A Ahmad, 4th year medical student BS8 1TH
Send response to journal:
|
If the history of the internet is to be our guide, we can conclude only one undeniable truth! This is that predicting its future is nothing but foolhardy. This history of the web has shown us that the many predictions of its future have been wrong, still not a day goes by when we hear the next vision of its future. This reminds me of a quote by the American writer Peter Drucker; “Trying to predict the future is like trying to drive down a country road at night with no lights looking out of the back window”. (ua4309@bris.ac.uk) Competing interests: None declared |
|||
|
|
|||
|
Bruce W. Thomas, Concerned reader 806 N. Glendale St., Wichita, Kansas 67208 USA
Send response to journal:
|
Recent discussion among collegues regarding "publication bias" stirred up the notion that when manuscripts are submitted to reviewers that author and institution names are NOT redacted in order to eliminate the clear bias that such identifiers would introduce. I was astonished to learn that others not only assumed that these identifiers were included for the reviewers but at the casual flippancy toward the practice, e.g., "wink-wink, nod-nod...of course that's the way it works!" Wishing to prove that such folly could not possibly survive in the era of Evidence Based Medicine, I called the editorial line at a major medical journal to ask them of their policy. I was told that, indeed, the author and institution names are NOT redacted but that, if I asked, they COULD be redacted from any manuscript I wished to submit (which, of course, would introduce further bias and inconsistency into the policy!). I asked why such a policy, which adds such a risk of publication bias, would be allowed. I was told that a reviewer must have that information so that she/he could recuse if there should be a "conflict of interest" (COI). Of course that is non-sensical since only the redaction of identifiers could ensure this. So, it seems, the COI decision is left to the reviewer to determine, subjectively?! But wouldn't elimination of identifiers remove that subjectivity altogether? Am I left to believe that after a trial is designed with strict methodology to eliminate bias, including that which is subliminal, e.g., prospectivity, double-blinding, strict randomization prcesses, etc., that the dissemination of the results may be hindered or even obviated by such a bias? Even if reviewers embrace the "honor system" and refuse to review manuscipts authored by friends, collegues or enemies, the fact that well- known authors/institutions or authors that have recently been read/noticed by the reviewers or simply have NAMES that are more or less likely to be of similar ethnic or national or cultural origins as the said reviewer, introduces a degree of bias that can never be ascertained, let alone rejected. It is astounding to me that if this glaring flaw in the process of the publication of scientific information exists that it has not been thoroughly discredited and abandoned. Please say it ain't so. Please tell me, at least, this practice is not widespread or that, if it is, that there is an overarching reason for it. Name Redacted Competing interests: None declared |
|||
|
|
|||
|
Adrian Walker, CTO Bristol CT 06010
Send response to journal:
|
There's some emerging technology that may be of interest in the context of Web 3.0 and medical data. It takes the form of a Wiki for knowledge -- in the form of rules in open vocabulary, executable English. It's online at [1], and shared use is free. There's a short video [2], describing how the technology is used with at database of drugs and their interactions. From knowledge in the form of English rules, the technology automatically generates and runs distributed SQL that would be too complex to write reliably by hand. The results are explained, in English, at the scientific or business level. The technology can also be used as a Service Oriented Architecture endpoint on the web [3]. [1] www.reengineeringllc.com [2] www.reengineeringllc.com/ibldrugdbdemo1.htm [3] www.reengineeringllc.com/iblClient1.java Competing interests: None declared |
|||