If you are going to talk about "tiers", then you'd better own that

SevenTierCakeOccasionally during the review of careers or grant applications you will see dismissive comments on the journals in which someone has published their work. This is not news to you. Terms like "low-impact journals" are wonderfully imprecise and yet deliciously mean. Yes, it reflects the fact that the reviewer himself couldn't be bothered to actually review the science IN those paper, nor to acquaint himself with the notorious skew of real world impact that exists within and across journals.

More hilarious to me is the use of the word "tier". As in "The work from the prior interval of support was mostly published in second tier journals...".

It is almost always second tier that is used.

But this is never correct in my experience.

If we're talking Impact Factor (and these people are, believe it) then there is a "first" tier of journals populated by Cell, Nature and Science.

In the Neurosciences, the next tier is a place (IF in the teens) in which Nature Neuroscience and Neuron dominate. No question. THIS is the "second tier".

A jump down to the IF 12 or so of PNAS most definitely represents a different "tier" if you are going to talk about meaningful differences/similarities in IF.

Then we step down to the circa IF 7-8 range populated by J Neuroscience, Neuropsychopharmacology and Biological Psychiatry. Demonstrably fourth tier.

So for the most part when people are talking about "second tier journals" they are probably down at the FIFTH tier- 4-6 IF in my estimation.

I also argue that the run of the mill society level journals extend below this fifth tier to a "the rest of the pack" zone in which there is a meaningful perception difference from the fifth tier. So.... Six tiers.

Then we have the paper-bagger dump journals. Demonstrably a seventh tier. (And seven is such a nice number isn't it?)

So there you have it. If you* are going to use "tier" to sneer at the journals in which someone publishes, for goodness sake do it right, will ya?

___
*Of course it is people** who publish frequently in the third and fourth tier and only rarely in second tier, that use "second tier journal" to refer to what is in the fifth or sixth tier of IFs. Always.

**For those rare few that publish extensively in the first tier, hey, you feel free to describe all the rest as "second tier". Go nuts.

22 responses so far

  • namnezia says:

    "Second tier" refers always to "lower-tier than where I usually publish".

  • Hermitage says:

    Which tier has Satan stuck in a frozen lake, beating his wings so ferociously he freezes all the miscreants surrounding him? I need the info for my...notes.

  • zb says:

    One can use the "tier" terminology while making arbitrary assumptions on the standard deviations of IF (and general wishy washy rankings) that differ from yours, and just divide journals into namnezia's categories (first tier = where one publishes and second tier = where other people not as good as you publish). Under such a scheme Nature/Science are in the first tier for everyone, but for other Nature Neuro, Neuron, J Neurosci, PNAS, and maybe even Cerebral Cortex can pop in depending on field. And, I don't know JBC and Biological Psychiatry, but I'm guessing they pop in, too. There's clearly subfield specific variations in journal rankings that determine "tier" designations.

    And, yes, this is all for people who judge based on journal rather than science. But that's not going to stop until there's basically only one journal that almost everything is published in (i.e. ArchivX, or something like that).

  • physioprof says:

    "Have you ever noticed that anybody driving slower than you is an idiot, and anyone going faster than you is a maniac?"

    George Carlin

  • drugmonkey says:

    30 is one hell of a lot farther away from 7 than 4 is. Just sayin'

  • Dave says:

    After CNS, there are only substandard journals. We all know this and so most of us are outrageous failures and our work cannot be important or interesting. We are not worthy of funding. It is a meritocracy....don't you know?

  • jipkin says:

    in partial defense of those who do this, perhaps overall impact factor doesn't capture within-field (or within-subfield) impact accurately. CNS have high impact because they're broadly read. But a good paper on inhibitory shunting in CNS or neuron or nat. neuro. or jneuro is going to be read by the same crowd no matter what. So perhaps in conversations by those people amongst themselves, they're justified in talking about publishing in a top-tier journal.

  • kevin. says:

    I might consider reading a paper on shunting inhibition in CNS, maybe Neuron or Nat. Neuro, but you'd probably have to put a gun to my head to read it elsewhere. If that isn't the definition of impact factor, I don't know what is.

  • drugmonkey says:

    Yeah, everybody knows J Neuro is a shitteasse dump journal. Geez.

  • physioprof says:

    If J Neurosci fluffed its IF with a shittetonne of review articles like Neuron and Nature Neuroscience do, I bet it would be well above 10.

  • miko says:

    IF is meaningless bullshit of course, and tiers need operational definitions that are useful.

    For me, C/N/S, Neuron, Nat Neuro, Plos Bio, J Neuro, PNAS Current Bio are the "same" in terms of some messy combination of how well the authors have sold the work, it's potential level of interest, etc.

    Below that are journals I've heard of.

    Below that are journals I haven't.

    = 3 tiers.

  • zb says:

    Putting j neuro, bio psychiatry, and neuropsycopharmo in the same tier doesn't make sence based on etymology alone. Tack the word science onto each one, and the have 1, 2, and 3 modifiers.

    Outside of the strict if bean counters that's how it works, right? people used to look in certain journals for certain papers, hence higher prestigiosity, which what "tiers" refer to.

    (don't know exactly how to explain the relative prestigiosity of NEJM, which should loose points for the location reference.

  • Spiny Norman says:

    Sepsis. (See today's NYT; PNAS).

  • Spiny Norman says:

    The original paper on the ubiquitin E1/E2/E3 ligases was in PNAS. And it has only a few hundred cites. And it won its authors a Nobel Prize. Glamour -- and even citations -- are not everything, suckas.

  • David says:

    Lumpers and splitters, dude. To me, "second tier" consists of the various ways to publish that aren't indexed in medline.

  • whimple says:

    Agree with miko on three tiers. In connection with grant application reviews, the tiers are:

    1) top tier: helps your new application get funded
    2) middle tier: expected productivity - neither helps nor harms your new application, but keeps you in the running
    3) bottom tier: ignored.

  • DJMH says:

    I like to read papers about shunting inhibition, because it's the best way to intimidate non-physiologists. That and dynamic clamp.

    But I agree with namnezia's definition, second-tier is defined by one's own publications....one tier down from that.

  • Spiny Norman says:

    "A jump down to the IF 12 or so of PNAS most definitely represents a different "tier" if you are going to talk about meaningful differences/similarities in IF."

    In general I'm not sure one can compare the IF's of general-interest journals with specialist journals. Maybe if ISI were to generate field-specific IF's for subsets of papers in the general-interest journals...

  • AcademicLurker says:

    In general I'm not sure one can compare the IF's of general-interest journals with specialist journals.

    Another issue is that citation metrics are, among other things, an index of field size. The field of neuroscience is so freaking huge* that a middling NS journal will have a higher IF than the top mathematics journal.

    *30,000 people at your annual meeting? WTF?!

  • drugmonkey says:

    Yay Neuroscience!

  • whimple says:

    The field of neuroscience is so freaking huge*... 30,000 people at your annual meeting? WTF?!

    Is that mostly trainees? Wonder where all those neuroscience jobs are going to come from?

  • DrugMonkey says:

    Trainees, yes. Yay Neuroscience!!!!

Leave a Reply


× 6 = twelve