Thus spake Nicholas Thompson circa 29/01/09 12:37 PM: > [...] it all > boils down to a one-bit decision: either you are going to read the sucker > or you arent.
Not that I'm argumentative or anything; but it's not just binary. I have at least 3 modes of reading: 1) read and integrate, 2) sloppy reading, and 3) skim. I do (1) when I want/expect to use the content for some task. I do (2) when I merely need to carry some context for understanding or communicating with others. And I do (3) when I want to determine whether I need to do (1) or (2), or when I just want an entry into some topic. So, the decision is, at least, quaternary. And much of which type of reading I do depends on the character of the publication pathway. And this is one of the reasons I hate the way /. and digg work. For whatever reason, I tend to get the most benefit out of obscure articles ... perhaps similarly, I seem to get the most enjoyment out of obscure music. Homogeny seems to be the enemy. > If one knows who the reviewers are ... knows their tastes, ete, perhaps > each consumer could rate reviewers and the program could give a stars by > reviewer-weighting customized for each consumer. Software could be > provided to do this. Very close to what Amazon provides right now, except > that each reader could accumulate his own personal judgments of reviewers, > rather than relying on swarm evaluation. This is a close approximation to what I'd like, except why approximate if you can shoot for the ultimate? If we were to develop a complicated projection from many to one dimensions, we may find that as difficult as implementing the multi-dimensional measure right off the bat. I suppose there would be marketing reasons... a competent funder might demand we start accumulating users immediately via a reducing projection and build out the multi-dimensional rating interface over time. > I forgot to say: Let's say the journal has an editorial board that rates > and comments on articles as they are submitted. Let;s say we start a > journal called THE FRIAM JOURNAL OF APPLIED COMPLEXITY. Every article is > sent out to five reviewers. So now we have a possiblity of 25 stars, say. > Now, the editor passes along the ratings and suggestions of the readers > and the author now can make choice. He can carry on publication with a low > rating, or he can revise and resubmit to get a better rating. This would be a nice evolution of the system we currently have. If I were the editor of an extant journal, I might find it attractive. But if I were to start an entirely new publication intent on revolutionizing peer-review, I would be more inclined to adopt a multi-dimensional rating system not based solely on number of stars. Of course, there are all sorts of compromises. Perhaps the stars are colored according to the domain expertise of the reviewer. Or perhaps we have multiple symbols for types of rating (innovation vs. clear communication vs. scientific impact etc.). > this led to another thought. A group such as this one wouldnt need even to > start its own journal. It could just start a rating service of some other > publication. We could, for instance, start by rating JASSS and putting the > ratings up on the web. The trouble is we wouldnt be rating or seeing the > articles that JASSS had rejected. I suppose we could ask JASSS to send us > all their rejected articles! Actually, I'd like to see something like this for hubs like pubmed or repositories like citeseer or the acm's digital library. I wouldn't want it to be publication-specific, though I might want it to be domain-specific. > I am probably too lazy to do anything like this, but I really like thinking > about it. [grin] Oh ... uh ... what? ... you were talking about actually _doing_ something?!? Umm ... ok ... perhaps I'm in the wrong place... [patting pockets, grabbing jacket, retreating from the room] ;-) -- glen e. p. ropella, 971-222-9095, http://agent-based-modeling.com ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
