barschall.stanford.edu header gif

Main Index: Trial Testimony June 10, 1997


                                                                160


   1    UNITED STATES DISTRICT COURT
        SOUTHERN DISTRICT OF NEW YORK
   2    ------------------------------x

   3    GORDON & BREACH SCIENCE
        PUBLISHERS S.A., STBS., LTD.
   4    and HARWOOD ACADEMIC
        PUBLISHERS GMBH,
   5
                       Plaintiffs,
   6
                   v.                           93 CV 6656 LBS
   7
        AMERICAN INSTITUTE OF PHYSICS
   8    and THE AMERICAN PHYSICAL
        SOCIETY
   9
                       Defendants.
  10
        ------------------------------x
  11
                                                June 10, 1997
  12                                            10:10 a.m.
        Before:
  13
                            HON. LEONARD B. SAND
  14
                                                District Judge
  15

  16

  17                            APPEARANCES

  18    ORANS, ELSEN & LUPERT, LLP
             Attorneys for Plaintiffs
  19    BY:  LESLIE A. LUPERT
             ROBERT L. PLOTZ
  20         PETER E. SEIDMAN

  21    COVINGTON & BURLING
             Attorneys for Defendants
  22    BY:  RICHARD A. MESERVE
             JEFFREY G. HUVELLE
  23         SUSAN L. BURKE

  24

  25




                                                                161


   1               (Trial resumed)

   2               THE COURT:  The witness may resume the stand.

   3     BRUCE KINGMA, Resumed.

   4               THE COURT:  I remind you, you are still under

   5    oath.

   6    CROSS-EXAMINATION (Continued)

   7    BY MR. HUVELLE:

   8         Q.    Dr. Kingma, I would like to start off by just

   9    talking briefly about the role of scientific journals.

  10               Are you familiar with the history of scientific

  11    journals?

  12         A.    Briefly familiar with the history.

  13         Q.    You are aware that they have been around for

  14    several hundred years?

  15         A.    Yes.

  16         Q.    Is it your understanding that scientific journals

  17    have been the principal means by which advances in science

  18    have been communicated within the scientific community?

  19         A.    Principally, yes.  There's obviously a number of

  20    other means of communication too, I guess.  I would say that

  21    they are an important part.

  22         Q.    But isn't it true that it is widely considered

  23    that, in science, the scientific journal has been the

  24    principal medium for the transmission of scientific

  25    knowledge?




                                                                162


   1         A.    Yes.

   2         Q.    Over the hundreds of years that scientific

   3    journals have been in use, there have developed traditions

   4    regarding the writing of articles for scientific journals;

   5    is that correct?

   6         A.    Yes.

   7         Q.    One of those -- one of the academic traditions

   8    relating to articles for scientific journals is that the

   9    author cite in connection with his article as references the

  10    prior works that have contributed in a significant way to

  11    his analysis in his own article; is that correct?

  12         A.    Correct.

  13         Q.    It's part of the academic tradition to treat such

  14    citations as an important part of a scientific article?

  15         A.    As an important part of it, yes.

  16         Q.    And it is true that, for most scientific

  17    journals, there is a peer review process that looks over the

  18    article and makes sure that it is of sufficient quality; is

  19    that correct?

  20         A.    Correct.

  21         Q.    In that it conforms to the standards of academic

  22    scholarship?

  23         A.    Correct.

  24         Q.    Including the citation of prior works that have

  25    contributed to the author's own analysis?




                                                                163


   1         A.    Correct.

   2         Q.    These academic traditions apply to the field of

   3    economics, too, your own field?

   4         A.    Correct.

   5         Q.    When you write an article, you exercise care and

   6    responsibility in citing those prior articles that have

   7    contributed in a significant way to your own analysis?

   8         A.    Correct.

   9         Q.    And other scientists and academicians with whom

  10    you are familiar have likewise exercised the same care and

  11    responsibility in citing the prior works, prior articles,

  12    that have contributed to their own analysis, correct?

  13         A.    Correct.

  14         Q.    Is it true that numerous citations to a

  15    scientific journal tell us that the scientists most engaged

  16    in advancing scientific knowledge find that articles in that

  17    journal contribute in a significant way to their own work?

  18         A.    Most citations to a journal show that that

  19    particular journal -- is that what you said -- makes a

  20    substantial contribution, that would be correct.

  21         Q.    Do you agree that the total number of citations

  22    to a journal is useful information for a librarian to have

  23    in connection with the assessment of the effectiveness of

  24    journals; is that correct?

  25         A.    I would say there is better information than




                                                                164


   1    citations, because you're now taking it from the world of

   2    journals down to the world of my library, my patrons, and

   3    that's where the analysis starts to crumble in that a

   4    particular journal might have a lot of citations but not be

   5    particularly useful to the patrons at my particular library.

   6         Q.    You said there are other better pieces of

   7    information, but would you agree that the total number of

   8    citations to a journal is useful information for a librarian

   9    to have in connection with the assessment of the

  10    effectiveness of that journal?

  11         A.    I would say that there is better information in

  12    terms of making collection management decisions.  Citations

  13    might provide you with -- the total citations might provide

  14    you with a piece of information about a set of journals, but

  15    those decisions still have to be made in concert with what

  16    the faculty at your library want.

  17         Q.    But do you agree that the piece of information

  18    provided by the total number of citations to a journal is

  19    useful to a librarian?

  20         A.    It may be useful to a librarian.

  21         Q.    You have written a textbook on information

  22    science?

  23         A.    I would call it economics information.  I

  24    wouldn't call it a textbook on information science.

  25         Q.    OK.  I will accept your characterization of your




                                                                165


   1    book readily.

   2                And this is a book for students?

   3         A.    Correct.

   4         Q.    And it goes over the basics of economics as

   5    applied to information sciences?

   6         A.    Correct.

   7         Q.    And one of the issues it addresses is the subject

   8    we mentioned yesterday, market failure?

   9         A.    Correct.

  10         Q.    The failure of the consumer to have appropriate

  11    information about the price or quality of a product?

  12         A.    Correct.

  13         Q.    In your textbook, you identified a number of

  14    examples of markets where there might be a problem with the

  15    available information, or in terms of the lack of relevant

  16    information?

  17         A.    Correct.

  18         Q.    One of the markets that you identified was the

  19    market for journal subscriptions?

  20         A.    Correct.

  21         Q.    The problem you identified was that the

  22    individual subscriber or library cannot determine the

  23    quality of a journal until after paying for it and receiving

  24    the subscription?

  25         A.    Correct.




                                                                166


   1         Q.    Do you recall that?

   2         A.    Correct.

   3         Q.    You suggested a market solution for that absence

   4    of information?

   5         A.    I believe it is a table that you are reading

   6    from --

   7         Q.    Right.

   8         A.    -- in which a number of market solutions are

   9    suggested.

  10         Q.    Right.  Four solutions?

  11         A.    I believe it's four.  I would like to refresh my

  12    memory.

  13               MR. HUVELLE:  I would like to show the witness

  14    Defendant's Exhibit DDD.

  15         Q.    At the bottom of the page, you identify the

  16    market for journal subscriptions?

  17         A.    Correct.

  18         Q.    One of the solutions that you propose for the

  19    problem of lack of information regarding quality of journals

  20    is the number of citations to published articles?

  21         A.    The third solution there, correct.

  22         Q.    Is it also true that, in your writings, you have

  23    referred to the problems that have arisen because of the

  24    increase in the number of scholarly journals that publish

  25    many articles never cited in later research?  Do you recall




                                                                167


   1    that?

   2         A.    Yes, I -- it's from one of the books that I've

   3    written, yes.

   4         Q.    You said that these articles seem to some people

   5    to exist solely for the purpose of padding the author's

   6    bibliography; do you recall that?

   7         A.    Correct, yes.

   8         Q.    Isn't it true that in writing this article and

   9    making that observation, you used the fact that some

  10    articles are never cited in later research as an indication

  11    of the low quality of the articles; is that true?

  12         A.    As a signal of this, correct.

  13         Q.    When did you receive tenure?

  14         A.    September of 1996.

  15         Q.    And you had to apply and be reviewed for that?

  16         A.    Yes.

  17         Q.    In connection with that, you had the opportunity

  18    to submit a short statement of your accomplishments; is that

  19    correct?

  20         A.    Yes.

  21         Q.    You had to specifically outline your achievements

  22    in the area of research, teaching, and university services?

  23         A.    Yes.

  24         Q.    Isn't it true as an indication of the quality of

  25    your work you included in that submission the number of




                                                                168


   1    times your journal articles had been cited in subsequent

   2    articles?

   3         A.    Yes.

   4         Q.    You referred yesterday to the Bensman article,

   5    Bensman being a professor at Louisiana State University, I

   6    believe?

   7         A.    He's a librarian at Louisiana State, actually.

   8         Q.    Librarian.  And you cited that article to support

   9    the proposition that faculty perceptions at Louisiana State

  10    University in the Chemistry Department do not correlate with

  11    rankings based on the impact factor?

  12         A.    Correct.

  13         Q.    Isn't it true that one of the central conclusions

  14    of Mr. Bensman was that total citations to a journal highly

  15    correlated with those faculty rankings?

  16         A.    Yes.

  17         Q.    An impact factor normalizes the size of journals;

  18    is that correct?

  19         A.    That's one of the things it does.  Obviously it

  20    differs from total citations in another way, too.

  21         Q.    If you had total citations, you might have a huge

  22    journal with many articles and thousands of citations?

  23         A.    Right.

  24         Q.    Impact factor divides the total citations by the

  25    number of articles?




                                                                169


   1         A.    It's -- it divides --

   2         Q.    In a two-year period.

   3         A.    In a two-year period, right.  So it's a snapshot

   4    and then divides by the number of articles, correct.

   5         Q.    Divides the total citations in a one-year period

   6    by the number of articles in the prior two-year period?

   7         A.    Correct.

   8         Q.    And so you get an average number of citations per

   9    article?

  10         A.    Correct.

  11         Q.    And so that's helpful, or it's fair to a smaller

  12    journal because you're dealing with averages on a per-

  13    article basis; is that correct?

  14         A.    I wouldn't call it fair or helpful.  I mean, what

  15    you're doing is normalizing in a way that I disagree with.

  16    But, nonetheless, you can take a smaller -- your statement,

  17    part of your statement is correct:  If you take a smaller

  18    journal and divide it by the number of articles, that might

  19    improve the impact factor for that smaller journal.

  20         Q.    You could have two journals, in each one, each

  21    and every article cited five times during a particular year?

  22         A.    Right.

  23         Q.    Every prior article is cited five times.  So the

  24    impact factor is going to be five for both journals?

  25         A.    Correct.




                                                                170


   1         Q.    But if one journal has a thousand articles,

   2    they're going to have 5,000 citations, and if the other

   3    journal has 20 articles, they'll have 100 citations?

   4         A.    Correct.

   5         Q.    So it's a different analysis?

   6         A.    It's a different number of citations.

   7         Q.    It's a different number?

   8         A.    Yes.

   9         Q.    It's a different number?

  10         A.    Right.

  11         Q.    Both are true numbers?

  12         A.    Right.

  13         Q.    They're both based on objective facts.  It's just

  14    different ways of looking at the data.

  15         A.    I don't know if I would agree that it's objective

  16    facts.  I mean, there are certain other issues going on with

  17    citations.  But it's -- you're right.  It's two -- the

  18    mathematics of that are that there are two different bases,

  19    two different numbers you are producing out of the impact

  20    factor.

  21         Q.    Well, the objective fact is the number of

  22    articles, the number of citations.

  23         A.    There is a number of citations, number of

  24    articles, correct.

  25         Q.    Isn't it true in Bensman's analysis, that he




                                                                171


   1    suggested that the reason impact factors did not correlate

   2    with faculty rankings is that the faculty gave -- ranked

   3    higher the large journals with lots of citations?

   4         A.    Correct.

   5         Q.    And so the faculty rankings tended to disfavor

   6    the small niche journals at the expense of the large society

   7    journals?

   8         A.    Correct.

   9         Q.    Your criticism of the impact factor is that it is

  10    too small of a window?

  11         A.    That's one of my criticisms of it, correct.

  12         Q.    Impact factor looks at citations in a particular

  13    year to articles in the two prior years?

  14         A.    Correct.

  15         Q.    You think it would be better to look at the total

  16    number of citations, the total number of citations in that

  17    year to all prior editions of the journal?

  18         A.    Citation -- total citations, right, is looking at

  19    the total number of citations, and that's certainly better

  20    than using a two-year/one-year window.

  21         Q.    You would look at all the citations in 1986 to

  22    articles published prior to 1986, back to 1975, 1974,

  23    where --

  24         A.    Correct.

  25         Q.    And of course that would disadvantage a new




                                                                172


   1    journal?

   2         A.    Correct.

   3         Q.    Now, if you look just at the two-year articles

   4    appearing in the two-year, prior year period -- the prior

   5    two-year period, you get information as to what's happening

   6    currently, don't you?

   7         A.    Looking at only a two-year snapshot gives you

   8    exactly that, a two-year snapshot.

   9         Q.    And the reason you're looking at citations or

  10    you're talking to the faculty, you're measuring use, is to

  11    make a judgment about the future; is that correct?

  12         A.    If you were -- if we're talking about librarians

  13    making collection development decisions, yes.  It's a matter

  14    of what's going to happen in the future, what's the expected

  15    value.

  16         Q.    Is the past a reliable indication of the future?

  17         A.    One would hope.  It is the only thing you can

  18    use -- the past by definition is the only thing you can use

  19    to guess at what's going to happen in the future.

  20         Q.    And certainly there are times when if you look at

  21    the 10-year past history or the 20-year past history, that's

  22    less reliable than if you look at the prior two years?

  23         A.    No, you're going to make fewer mistakes looking

  24    at a 20 or 20-year past history than only the prior two

  25    years.




                                                                173


   1         Q.    You don't know whether that's true in the case of

   2    scientific journals, though, do you?

   3         A.    I don't know if that's true in the case of

   4    scientific journals, but the -- when you think about how

   5    you're sampling those journals, it's clear that if you're

   6    looking at a 20-year history you're going to be better off

   7    than only looking at the past two years.  You may be

   8    ignoring pertinent information by discarding 18 years.

   9         Q.    You may be misled, as well, by a journal that had

  10    a glorious past and a troubled present?

  11         A.    Hopefully you would be able to pick out that

  12    troubled present by their -- the editors of that journal,

  13    any kind of knowledge you might have about that journal in

  14    its recent history.

  15         Q.    But isn't it true whether you picked two years or

  16    10 years or 20 years, in some cases that past history may

  17    not be the best slice to look at?

  18         A.    In --

  19         Q.    For some journals.

  20         A.    For some journals, depending on the slice, yes,

  21    correct.

  22         Q.    Have you read the deposition of Ms. Hunter in

  23    this case regarding impact factors?

  24         A.    No, I have not read the deposition of Ms. Hunter.

  25         Q.    Do you know whether other publishers use impact




                                                                174


   1    factors or total citations when they want to know how well

   2    their journals are doing?

   3         A.    Yes.

   4               Let me finish what I was about to say, is, I have

   5    read the report from Ms. Hunter but not the deposition of

   6    Ms. Hunter.  And in the report I believe she says that

   7    Elsevier uses impact factors.

   8               Elsevier is the name of the publisher.

   9         Q.    Let me turn to the price side of the equation.

  10               Do you agree that the librarian must make

  11    decisions regarding competing journals based in part on

  12    price information?

  13         A.    Yes.

  14         Q.    And librarians must make comparisons between

  15    journals in terms of price, among other factors?

  16         A.    Yes.

  17         Q.    Do you agree that the information on price should

  18    be in a format that allows the librarian to make wise

  19    management decisions about journal acquisitions?

  20         A.    Yes.

  21         Q.    You agree that price information can be presented

  22    in a variety of ways?

  23         A.    Yes.

  24         Q.    Do you recall that Professor Barschall, in both

  25    of his articles, displayed the annual subscription price for




                                                                175


   1    each journal as well as the cost per thousand characters?

   2         A.    Yes.

   3         Q.    He displayed that information for two years?

   4         A.    Yes.

   5         Q.    But in addition to the annual price or

   6    subscription price, Barschall normalizes the size of the

   7    journals to provide a common measure of cost per amount of

   8    article information; is that correct?

   9         A.    Yes, that is what he did.

  10         Q.    Do you agree that it is a sound approach to

  11    normalize the cost in that manner?

  12         A.    No, I do not.

  13         Q.    I believe you discussed the reasons for that view

  14    yesterday; is that correct?

  15         A.    Yes.

  16         Q.    Were you here during the testimony of Dr. Taylor?

  17         A.    No, I was not.

  18         Q.    Did you have an opportunity to read the testimony

  19    of Dr. Taylor?

  20         A.    No, I did not.

  21         Q.    Let me --

  22         A.    Let me --

  23         Q.    -- pose this as a hypothetical to you.

  24         A.    All right.

  25         Q.    The journal editor whose journal price was




                                                                176


   1    $9,000, compared to $507 for another journal normalized

   2    cost, in order to calculate the cost per article, would you

   3    say that's a wrong approach?

   4         A.    Are you saying the publisher did that?

   5         Q.    Yes.

   6         A.    For the publisher to do it is actually a bit

   7    different from the librarian doing it on the purchase

   8    decision.

   9         Q.    Why is that?

  10         A.    Well, because a publisher is also in some sense

  11    selling those articles through document delivery or

  12    interlibrary loan and has to set copyright fees on a per-

  13    article basis.  So for the publisher to divide it up on a

  14    per-article basis might give that publisher information

  15    about how to set a copyright fee as articles are delivered

  16    through this alternative means.

  17         Q.    But if the editor who did this, the editor of the

  18    $9,000 journal, did it in response to complaints that his

  19    journal's price was very high, and in order to demonstrate

  20    the cost effectiveness of his journal, would you say that's

  21    the wrong approach?

  22         A.    I think it's the wrong approach for a librarian

  23    making a decision to tell a publisher what they should or

  24    shouldn't do.  But as far as librarians go, their

  25    responsibility is to purchase the entire subscription or to




                                                                177


   1    find an alternative access to that information through a

   2    system like interlibrary loan.

   3         Q.    I believe you conceded yesterday that normalizing

   4    price on the basis of a common unit of measurement is

   5    appropriate in a grocery store?

   6         A.    Correct.

   7         Q.    Is it --

   8               THE COURT:  We will take a five-minute recess.

   9               MR. HUVELLE:  Fine.

  10               (Recess)

  11               THE COURT:  Sorry for the interruption.  You may

  12    be seated.

  13               You know, as long as we have interrupted, you

  14    made some reference yesterday to a book you had written or a

  15    study you had done on comparing ownership to lease access;

  16    is that --

  17               THE WITNESS:  Well, to access.  It's called the

  18    access-to-ownership comparison, where ownership means buying

  19    the journal subscription and access is acquiring articles on

  20    an as-needed basis through interlibrary loan.  So the

  21    faculty member comes in, wants a certain --

  22               THE COURT:  Acquiring from, what?  From other

  23    libraries?

  24               THE WITNESS:  From other libraries, from

  25    commercial document delivery sources, other places that have




                                                                178


   1    the journals you're looking for, essentially.  They're

   2    photocopying an article on that journal and sending it on to

   3    you at your library.

   4               THE COURT:  Which creates copyright fees?

   5               THE WITNESS:  Well, there is a copyright fee

   6    associated with that, so --

   7               THE COURT:  I just want sure what the terminology

   8    meant.  Thank you.

   9    BY MR. HUVELLE:

  10         Q.    If I may just pursue that subject for just a

  11    moment.

  12               In connection with your study, you worked out a

  13    formula by which a library should determine if it is cheaper

  14    to borrow an article than to buy the journal?

  15         A.    Correct.

  16         Q.    And you noted that one consequence of using your

  17    formula, if libraries did, would be that the smaller

  18    journals that were used less, that were more expensive,

  19    would be canceled by some libraries?

  20         A.    I don't know if I said "smaller," but journals

  21    that were used less, more expensive, clearly both these

  22    things factor into the formula, the price of the journal and

  23    the amount to which it's used.  So, low-use, high-cost means

  24    you're more likely as a library to cancel that particular

  25    subscription.




                                                                179


   1         Q.    And one of the macro effects that you did not

   2    address in your cost/benefit analysis was the consequence

   3    that, as a result of canceling journals that are high-cost,

   4    low-use, the cost of those journals would actually go up, or

   5    the price of the journals would go up even further?

   6         A.    Right.

   7         Q.    Which would then lead to more libraries

   8    canceling?

   9         A.    Correct.

  10         Q.    And your position was, well, an individual

  11    library can't worry about that?

  12         A.    Correct.

  13         Q.    Back to the grocery store.

  14         A.    OK.

  15         Q.    The appropriateness of unit costs in a grocery

  16    store, in your judgment does that apply to all products in

  17    the grocery store?  We focused a little bit on pasta, but I

  18    want to know if it has more general application.

  19         A.    In each product, there is going to be a container

  20    of a certain size, but also quality differences.  Now,

  21    clearly unit pricing on pasta doesn't allow me to compare it

  22    to -- and there's unit pricing down a few aisles on beans,

  23    but I'm not really making a comparison between pasta and

  24    beans.  There are quality differences.  There are, you know,

  25    significant differences between those products; that the




                                                                180


   1    unit pricing is for a different reason than comparing the

   2    cost per ounce for beans to the cost per ounce of pasta.

   3         Q.    But do you agree with the appropriateness of

   4    displaying unit cost data for beans and fruit juice and ice

   5    cream, different products?

   6         A.    Many of the different products in the grocery

   7    store, yes.  I mean, there are clearly examples in the

   8    grocery store where you don't have unit pricing, you don't

   9    have a price per ounce of a broom, because we're not

  10    interested in the price per ounce of a broom.  You just want

  11    to know how much you're going to pay for the broom.

  12         Q.    As I understood it, the reason you gave why a

  13    common unit price is not appropriate for scientific journals

  14    is that scientific journals are more like a car than they

  15    are like pasta?

  16         A.    Yes, it's tough to draw that comparison, but,

  17    yes, I would agree with that.  They're more like a car in

  18    that there's information, there's something else embedded

  19    within the product than they are like pasta.  Yes, I would

  20    agree with that, although it's -- it's tough to think about

  21    it.

  22         Q.    You don't agree with my statement.  I'm simply

  23    reciting yours.

  24         A.    OK.

  25         Q.    What is the difference between a car and pasta




                                                                181


   1    that is pertinent to the distinction you're making?

   2         A.    A car you're purchasing as a single unit, and

   3    within that there are many quality characteristics of a car

   4    that are influencing your decision to buy, depending on how

   5    many people you want to fit in the car, how fast you want

   6    the car to go, what color it is, etc.  There are many

   7    characteristics to the car, and you purchase it as a car.

   8               Whereas with pasta you're looking at it and

   9    within the grocery store setting making the comparison

  10    perhaps between two different sized boxes of pasta.  Within

  11    each box is relatively the same quality good -- relatively

  12    the same quality good for the decision you're making, but

  13    you want to know the price per ounce of those two boxes,

  14    which might be of different size.

  15         Q.    So are you telling us that the operative

  16    distinction is whether there is variation in quality among

  17    the products?

  18         A.    It's variation in quality, yes, along with the

  19    amount you're purchasing.

  20         Q.    Do you say it's inappropriate as to cars because

  21    when you buy a car you consider quality?

  22         A.    When you buy a car you consider quality, and

  23    you're also, you know, purchasing a car as one solid good.

  24    You could, even in theory, if you had two cars of the exact

  25    same quality, I still don't think you would want to




                                                                182


   1    normalize it on a price per pound basis.

   2         Q.    The cost basis -- a cost comparison is a cost

   3    efficiency analysis?

   4         A.    Right, correct.

   5         Q.    Which compares cost, correct?

   6         A.    Correct.

   7         Q.    It has nothing to do with quality?

   8         A.    Correct.  Doing a cost efficiency analysis makes

   9    the assumption that the quality is very similar, not

  10    identical.

  11         Q.    Is it your testimony that there are no

  12    differences in the quality of pasta or other goods you see

  13    in a grocery store?

  14         A.    There are certainly differences in the quality of

  15    pasta.

  16         Q.    Isn't the whole point of unit pricing in a

  17    grocery store that you see the brand name, the Proctor &

  18    Gamble Tide, and then you see the Safeway detergent X, and

  19    you go in with some assumption as to the differences in

  20    quality, and then the price data on a unit basis allows you

  21    to see whether for one you are paying $1.50 per ounce and

  22    the other you're paying 5 cents an ounce.  It just gives you

  23    the cost comparison, and then you use that in making a

  24    decision; is that correct?

  25         A.    Correct.  And that's actually a very good




                                                                183


   1    example, because as you're using detergent, you're obviously

   2    using it by the ounce, not by the unit you're purchasing.

   3         Q.    So that's a factor in whether or not the unit

   4    cost is appropriate?

   5         A.    In the sense that you're using it by the scoop,

   6    by the ounce.

   7         Q.    If I have these journals and I know the annual

   8    subscription price of these journals, I know the annual

   9    subscription price of these journals, and I know the annual

  10    subscription price of the green journals, and those are all

  11    one year of that journal, and then the orange and white is

  12    another journal, one year.  The librarian knew the price of

  13    all four journals.  Is there any cost price information

  14    about these journals that in your judgment would be helpful

  15    in making acquisition decisions?

  16         A.    Well, just the subscription price, simply that.

  17         Q.    That's the only information?

  18         A.    Well, the only information about price, cost.

  19    You want to know --

  20         Q.    What about --

  21               MR. PLOTZ:  Please let him answer.

  22               MR. HUVELLE:  I'm sorry.

  23         A.    You want to know to what degree these journals

  24    are going to be used by your patrons.  That's the other side

  25    of the coin.




                                                                184


   1         Q.    Let's focus on price.

   2         A.    OK.

   3         Q.    What about telling the librarian what the cost

   4    price per page is of the journals?

   5         A.    I don't think that's useful information.

   6         Q.    Isn't it true that if a librarian knew the price

   7    of this journal and he knew the price of this journal and he

   8    knew the price of those journals, that he wouldn't think

   9    about that, or that what he would do is, he would take into

  10    account the differences in the amount of information he's

  11    getting; isn't that right?  He would normalize in his own

  12    mind?

  13         A.    One would hope that he or she would not do that,

  14    that they would look at the usefulness of what they're

  15    purchasing rather than simply the size of it.  The example I

  16    gave in my deposition earlier is this, you know, these

  17    places that sell books by the foot, for aesthetic reasons.

  18    The cheapest way to fill your library shelves is to purchase

  19    books by the foot, but that doesn't mean that your patrons

  20    are going to be using those materials, and it would be a

  21    poor collection management decision to simply buy books by

  22    the foot which are sold for aesthetic reasons of filling up

  23    people's shelves in houses.

  24         Q.    Suppose we had four journals, they're all good

  25    enough to get in the ISI database.  In each case there's one




                                                                185


   1    faculty member who says, I'm on the board of that particular

   2    journal so I'd like to have it.  It's embarrassing me for me

   3    to go to board meetings and not have it.

   4               So we know something about the quality.  But the

   5    librarian needs to justify to his boss, we're using the

   6    money wisely, we're getting a good bang for the buck.

   7    Right?

   8               Isn't that what unit pricing is about?  Bang for

   9    the buck?  What you're getting for your money?

  10         A.    I don't know if I'd call it "bang for the buck."

  11    I just strongly disagree with normalizing pricing for

  12    journals on a per-page or per-character basis.

  13         Q.    You're saying that if this journal is a little

  14    bit less expensive than most journals, that the librarian

  15    ought to, in making the decision, just consider this as less

  16    expensive without regard to the fact that that journal

  17    provides, or those journals provide perhaps ten times as

  18    many articles?

  19         A.    I think your question is putting the librarian in

  20    a vacuum without any information about use or possible use

  21    of those journals.  If you only have price and you have

  22    nothing else, boy, that -- I mean, that would be an

  23    impossible decision to make.

  24         Q.    I'm not --

  25         A.    OK.




                                                                186


   1         Q.    The librarian has lots of other information.  Is

   2    it appropriate for the librarian, in addition, in addition

   3    to knowing the annual subscription price, and in addition to

   4    knowing things about quality, to know what the cost per

   5    character is, or the cost per page, or the cost per article,

   6    or some other common measurement that allows a comparison on

   7    a common basis?

   8         A.    No.  The normalization that should occur is the

   9    cost per subscription, how much I'm paying for a

  10    subscription.  That's really the only normalization that

  11    should occur under these circumstances.

  12         Q.    Isn't it true that in the case of the green

  13    journals over there, it's not like a car in either you get

  14    the whole thing or not, that contains several thousand

  15    articles; is that correct?

  16         A.    Correct.

  17         Q.    No one is going to read all the articles?

  18         A.    I don't know if there's anybody that would read

  19    all the articles.  The editor would read all the articles.

  20         Q.    Well, we can even wonder about that.

  21               In the pasta example, it's divisible into meals,

  22    correct?

  23         A.    Correct.

  24         Q.    The university cafeteria can buy pasta on a

  25    per-unit basis and then students come in and they take a




                                                                187


   1    large portion or a small portion, but they do that, right?

   2         A.    Correct.

   3         Q.    And in the case of the journal, the 2,000, 3,000

   4    articles, the librarian buys it, and different students come

   5    in, read one article, two articles.  They treat it not as an

   6    entity but as a divisible product; is that correct?

   7         A.    The students treat it as a divisible product.

   8         Q.    True.

   9         A.    Divisible in the sense that it's their use,

  10    right?  It's a student coming into the library and using

  11    that journal.  You could have 50 students all using the same

  12    article from that journal.

  13         Q.    So in your view, the only -- the only -- price

  14    information is the annual subscription price?

  15         A.    The annual subscription price, yes.  There is

  16    the -- these fixed costs associated with buying a journal,

  17    too, that involve shelving and cataloging, etc., so --

  18         Q.    You mentioned yesterday that Dr. Barschall did

  19    not take that into account?

  20         A.    Correct.

  21         Q.    Should he have?

  22         A.    It's relatively the same for all these journals.

  23         Q.    So there's no reason to?

  24         A.    There's really no reason to, under these

  25    circumstances.  Those costs are about $60 to $70 per journal




                                                                188


   1    subscription, so when you're talking about prices and in the

   2    early '80's when you did this, they were probably less than

   3    $60 or $70.

   4         Q.    But it washes out?

   5         A.    It probably washes out as a small percentage of

   6    the whole.

   7         Q.    If you were a librarian and you looked at this

   8    set of journals, this set of journals and those two sets,

   9    would you make any assumptions about what the subscription

  10    price would be?

  11         A.    I am not a librarian, but, no, I wouldn't make

  12    any assumptions.

  13         Q.    You have no way of telling which one would be

  14    more expensive?

  15         A.    No.  Hopefully you would be able to find that

  16    price information.

  17         Q.    So it would not be surprising to you if this

  18    journal cost more than both of those journals combined?

  19         A.    It wouldn't be particularly surprising, no.

  20         Q.    And it would not be surprising to you if the

  21    price per page of this journal was 67 cents and the price

  22    per page of those journals was 5 cents?

  23         A.    I wouldn't be --

  24         Q.    The same with this?  It wouldn't surprise you?

  25         A.    No.




                                                                189


   1         Q.    But that's not something a librarian ought to

   2    think about in deciding whether or not to subscribe to this

   3    journal or that journal?

   4         A.    They should not think about the price per page,

   5    no.

   6         Q.    But it certainly allows us to say this journal is

   7    much more expensive?

   8         A.    The subscription price might allow us to say that

   9    that journal is much more expensive, yes.

  10         Q.    And this one, which has the same per-page price

  11    as this one, can we say that this journal is more expensive

  12    than those two when the total price is less than those?

  13         A.    I'm sorry.  I'm getting a little confused.  The

  14    per-page price --

  15         Q.    The per-page price is the same?

  16         A.    Same.

  17         Q.    Total price is less of this than this and less

  18    than those?

  19         A.    How can the per page price be the same and the

  20    total price when that's a smaller stack?  I don't

  21    understand.

  22         Q.    The per-page price is the same.

  23         A.    Right.

  24         Q.    The total price of the smaller stack is therefore

  25    less.




                                                                190


   1         A.    Oh, is less, OK.  Yes.

   2         Q.    Would you agree that this journal also is more

   3    expensive than those journals?

   4         A.    I don't know that -- you're asking me to make a

   5    mathematical calculation of multiplying the number of pages,

   6    so --

   7         Q.    If the total subscription price of this journal

   8    is less than those journals --

   9               THE COURT:  You know, I think if somebody reads

  10    this record, it might be helpful to know that what you are

  11    referring to are stacks of blue-covered --

  12               MR. HUVELLE:  Your Honor, that's an excellent

  13    point, and I can try to clarify it, although it would have

  14    been helpful, I guess, if I would tried earlier, but I will

  15    do my best now.

  16               THE COURT:  Sometimes you read a record and --

  17               MR. HUVELLE:  I can imagine.

  18               THE COURT:  It hurts me here, Doctor.

  19         Q.    The journals with the 67-cent-per-page price that

  20    I have been referring to, one is an orange journal; is that

  21    correct?

  22         A.    Correct.

  23         Q.    And that bears the title Ferroelectrics?

  24         A.    Correct.

  25         Q.    And then the other journal that's small in number




                                                                191


   1    but high in price is a white-and-black journal called

   2    Physics and Chemistry of Liquids?

   3         A.    Correct.

   4         Q.    And let me represent that the two larger stacks

   5    are a green stack, which is Physical Review B, and a black-

   6    and-orange stack, which is Journal of Applied Physics.

   7               MR. HUVELLE:  I will represent that for the

   8    record.  And thank you, your Honor, for --

   9               THE COURT:  And the relative heights of the

  10    stacks I think are part of what you can do to make your

  11    question meaningful.

  12         Q.    Would you estimate for us the size of the orange

  13    Ferroelectrics stack?

  14         A.    It looks to be about ten inches high.

  15         Q.    Or a little less?

  16         A.    That's --

  17         Q.    And the Physics and Chemistry of Liquids?

  18         A.    It looks to be two inches high.

  19         Q.    And the green Physical Review B?

  20         A.    That's those two stacks, right?

  21         Q.    The two stacks, each about a foot and a half.

  22         A.    Yes, so about three feet high together.

  23         Q.    And the Journal of Applied Physics is --

  24         A.    Is just that one stack?

  25         Q.    Yes.




                                                                192


   1         A.    So about a foot and a half.

   2         Q.    Do you have any of the exhibits up there from

   3    yesterday?

   4         A.    I have Physics Today, Bulletin of the Physics

   5    Society, a table from my book, and the impact numbers from

   6    the Science Citation Index for '86.

   7         Q.    I would like you to look at Plaintiff's Exhibit

   8    2.

   9               MR. HUVELLE:  Your Honor, I have an extra copy if

  10    you need one.

  11               THE COURT:  I think I have it up here.  Yes.

  12         Q.    Do you have Plaintiff's Exhibit 2?

  13         A.    Yes.

  14         Q.    Can you turn to Table 3.  Is this a ranking of

  15    journals by cost per character?

  16         A.    Yes.

  17         Q.    And it's a three-page list?

  18         A.    Yes.

  19         Q.    Now, in the first page, the journals run from

  20    39 -- .39 to 3.4 cents?

  21         A.    Yes.

  22         Q.    Now, on page 2, they run from 3.4 to 7.4 cents?

  23         A.    Yes.

  24         Q.    Is that correct?

  25               And these are ranked in order of increasing cost?




                                                                193


   1         A.    Yes.

   2         Q.    And the total on the three pages is about 200

   3    journals that Barschall studied?

   4         A.    Yes, about.

   5         Q.    And then on page 3, you go halfway down the page

   6    and you get to the first journal that has a cost of 10 cents

   7    per thousand characters.  Do you see that?

   8         A.    Yes.

   9         Q.    And then you go down a little further and there's

  10    a journal with a cost of 11.9 cents per thousand characters?

  11         A.    Yes.

  12         Q.    Then you jump up to 14 cents per thousand

  13    characters; is that correct?

  14         A.    Yes.

  15         Q.    One journal.

  16               Then you jump again to 16 cents per thousand

  17    characters?

  18         A.    Yes.

  19         Q.    And then we have the bottom 14 journals, the most

  20    expensive -- 14 most expensive journals on a cost per

  21    thousand character?

  22         A.    Yes.

  23         Q.    And of those 14 most expensive journals on a

  24    cost-per-thousand-character basis, 11 are Gordon & Breach

  25    journals; is that correct?




                                                                194


   1         A.    Yes.

   2         Q.    And so the least expensive of the Gordon & Breach

   3    journals ranks 187 out of the 200?

   4         A.    If there's 200 here, yes.

   5         Q.    Can you turn to Plaintiff's Exhibit 3.

   6         A.    Yes.

   7         Q.    Do you have that?  Can you look at Table 1.

   8         A.    Yes.

   9               THE COURT:  Bear with me just a moment.

  10               MR. HUVELLE:  I have an extra copy.

  11               THE COURT:  I will take your extra copy.  Thank

  12    you.

  13               Exhibit 3, what are you looking at?

  14               MR. HUVELLE:  We are looking at Plaintiff's

  15    Exhibit 3, Table 1.

  16         Q.    You looked at this yesterday, Dr. Kingma; is that

  17    correct?

  18         A.    Yes.

  19         Q.    And you looked at the categories listed on the

  20    left on Table 1?

  21         A.    Yes.

  22         Q.    Dr. Barschall has grouped the journals included

  23    on this page under eight categories?

  24         A.    Yes.

  25         Q.    And you have no reason to dispute that those are




                                                                195


   1    reasonable and appropriate categories?

   2         A.    I have no reason to dispute these categories.

   3    I'm not a physicist, so --

   4         Q.    Do you agree that the top-ranked journal in the

   5    Letters Journals category is an AIP journal?

   6         A.    Ranked by ratio of cost to impact?  Yes.

   7         Q.    The top two are AIP and APS; is that correct?

   8    Under Letters Journals?

   9         A.    Correct.

  10         Q.    And under Review Journals, the top-ranked journal

  11    is an APS journal?

  12         A.    These are actually ranked by cost per thousand

  13    characters.  Yes, correct.

  14         Q.    The top-ranked -- but it's true as to cost, ratio

  15    of cost to impact factor as well, is it not?

  16         A.    Correct.

  17         Q.    And in the Atomics Physics category, the

  18    top-ranked journal by ratio of cost to impact is an APS

  19    journal?

  20         A.    Correct.

  21         Q.    The same is true in the Condensed Matter Physics

  22    category?

  23         A.    Correct.

  24         Q.    The same is true in the Nuclear Physics category?

  25         A.    Correct.




                                                                196


   1         Q.    The same is true in the Particle Physics

   2    category?

   3         A.    Correct.

   4         Q.    The same is -- well, in the case of Applied

   5    Physics, it's an AIP journal that is the top ranked?

   6         A.    Correct.

   7         Q.    And in the Instrumentation category, it's an

   8    AIP-APS journal that is top-ranked?

   9         A.    Correct.

  10         Q.    And the number two and number three journals in

  11    that category are also AIP journals; is that correct?

  12         A.    In instrumentation?

  13         Q.    Yes.

  14         A.    It says IOP.

  15         Q.    Well, if you look at the ratio of cost to

  16    impact --

  17         A.    Oh.

  18         Q.    Is it --

  19         A.    Yes, correct.

  20         Q.    The top three are all AIP journals in that

  21    category?

  22         A.    By cost to impact ranking.

  23         Q.    Right.  Dr. Kingma, if you will indulge me in a

  24    hypothetical or an assumption, the assumption is that the

  25    ratio of cost to impact is an appropriate measure of cost




                                                                197


   1    effectiveness.  Can we proceed with that assumption?

   2               Let me ask you, if we accept that assumption, is

   3    it true that the data on Table 1 would therefore support the

   4    conclusion that the journals of APS and AIP are the best

   5    bargain?

   6         A.    Obviously I don't agree with the assumption.

   7         Q.    I appreciate that.

   8         A.    If you assume that that's the ranking, then

   9    that's what the numbers say.

  10         Q.    Thank you.

  11               Let me go back now to where we were when we ended

  12    at the conclusion of yesterday.  You had identified five

  13    possible criteria that one might use in measuring the

  14    effectiveness of journals.  Do you recall that?

  15         A.    Yes.  I think we talked about them as proxies,

  16    numbers that could serve as proxies for use value of a

  17    journal.

  18         Q.    I would like you to focus on effectiveness.

  19         A.    OK.

  20         Q.    Isn't it true that those are five possible

  21    measures for the effectiveness of journals?

  22         A.    Some good measures, some bad measures, yes; by

  23    definition they are measures.

  24         Q.    But they are all ones that you have identified,

  25    whether good or bad?




                                                                198


   1         A.    Correct.

   2         Q.    And we discussed readership, one of the possible

   3    measures, yesterday.  I would like to go on to the second

   4    one, use.  OK?

   5         A.    OK.

   6         Q.    And can you tell us what you mean by "use"?

   7         A.    "Use" is the number of times somebody is going to

   8    pull it off the shelf, pull a journal, an issue of that

   9    journal off the shelf and use it.  Now, they may use it and

  10    photocopy an article from it and ultimately read that

  11    article several times, but use is simply the pulls off the

  12    shelf.

  13         Q.    And as you used the term, all uses are treated

  14    the same?

  15         A.    You can treat all uses the same.  You can weight

  16    them by who uses it, if a faculty member pulls it off the

  17    shelf versus a student pulling it off the shelf.

  18         Q.    One difficulty of use as a measure is that it is

  19    difficult to tell, or difficult to differentiate between

  20    uses; isn't that true?

  21         A.    It would be possible to construct a use survey

  22    where you are differentiating between uses.  That would

  23    clearly be a more expensive use survey than one in which you

  24    were not differentiating between uses.

  25         Q.    Typically, if a library tries to study use, it




                                                                199


   1    tries to determine the number of times the journal is pulled

   2    off the shelf, correct?

   3         A.    Typically -- can you repeat the question?

   4         Q.    Typically when a library does a use study, it

   5    determines how many times the journal is pulled off the

   6    shelf?

   7         A.    Correct.

   8         Q.    When you do that kind of study, you don't know

   9    whether the person glanced at the article, read two

  10    sentences, and then put it back?

  11         A.    Correct.

  12         Q.    You don't know whether the user reshelved it

  13    before it could be counted?

  14         A.    Correct.  Hopefully in designing your use study

  15    you have taken care to try and prevent users from doing

  16    that.

  17         Q.    And you don't know if, after the journal has been

  18    pulled off the shelf, it has been used by multiple people?

  19         A.    Correct.  Again, hopefully you have designed your

  20    study such that you are reshelving it as quickly as possible

  21    so that there is limited multiple use possible.

  22         Q.    And you don't know whether it is use by a

  23    first-year physics student, who reads the article, becomes

  24    discouraged by the complexity, and switches majors, or

  25    whether it's use by a Nobel laureate who finds that the




                                                                200


   1    article contributes importantly to his own research, do you?

   2         A.    Again, it would depend on your design of the

   3    survey.

   4         Q.    But it would be legitimate for a library to

   5    differentiate between different uses?

   6         A.    If they could, yes.

   7         Q.    It would be legitimate for a library to focus

   8    entirely on uses by the faculty?

   9         A.    I don't think it would be legitimate to focus

  10    entirely on uses by the faculty.  Clearly students use that

  11    library, and their concerns as patrons should be addressed.

  12         Q.    But might it be the case that the library would

  13    decide that, by determining the uses of the faculty, that

  14    that was a reliable basis for making decisions about what

  15    journals to acquire?

  16         A.    The library might have a use study -- again, if

  17    the library is having a use study and it's determining the

  18    use of faculty and students, separating it, differentiating

  19    it by that, then one would want to use both pieces of

  20    information, and I believe any librarian would say that you

  21    would not want to throw out the use of the students in

  22    making a determination.

  23         Q.    But you could devise a study that only gathered

  24    information as to faculty uses?

  25         A.    Possible.




                                                                201


   1         Q.    And if you did so, that would provide helpful

   2    information to the librarian?

   3         A.    Correct.

   4         Q.    And a third measure that you have suggested is

   5    faculty rankings?

   6         A.    Yes.

   7         Q.    And that, by nature, focuses entirely on the

   8    faculty?

   9         A.    Yes.

  10         Q.    And it focuses on their views as to the journals

  11    that contribute most significantly to their work?

  12         A.    The journals they find most useful, correct.

  13         Q.    And do you think that's an appropriate measure?

  14         A.    Yes.  Falling short of having use statistics, I

  15    think use statistics are actually better than that

  16    particular measure.

  17         Q.    But it's appropriate in the sense that it

  18    provides a reliable basis for management decisions by

  19    librarians, faculty rankings?

  20         A.    Correct.  It's not -- again, it's not the first,

  21    best choice, but it's a useful number, falling short of

  22    having use numbers.

  23         Q.    Citation data is a comprehensive faculty ranking,

  24    is it not?

  25         A.    It's a list of how many times each journal is




                                                                202


   1    getting cited by the entire universe of faculty, clearly not

   2    by just the faculty within your library or that are served

   3    by your library.

   4         Q.    It reflects the considered judgment of the most

   5    productive members of the faculty as to the journals and

   6    articles that contribute most significantly to their own

   7    research?

   8         A.    Well, its citations include everyone's citations;

   9    not only the most productive we are talking, but the

  10    faculty.

  11         Q.    I was using it for the faculty members who

  12    actually write articles.

  13         A.    Oh, OK.

  14         Q.    With that clarification, can you respond to the

  15    question?

  16         A.    That for all the -- the faculty that are writing

  17    articles and citing all the entire body of faculty citations

  18    show how many times they are citing things.

  19         Q.    With respect to the five possible measures of

  20    effectiveness, you have stated that different people have

  21    different views as to which of the five measures is the

  22    best, correct?

  23         A.    Correct.

  24         Q.    And indeed, you noted that Professor Barschall

  25    believed that the impact factor was the best measure of




                                                                203


   1    effectiveness; is that correct?

   2         A.    Correct.

   3         Q.    And are you familiar with Professor Barschall's

   4    career?

   5         A.    As a physicist?  No.

   6         Q.    As a physicist.

   7         A.    I am not.

   8         Q.    You know that he had a long and distinguished

   9    career as a physicist?

  10         A.    My understanding is that he did.

  11         Q.    Received many honors for his contributions to the

  12    field of physics?

  13         A.    That's my understanding.

  14         Q.    Would you be inclined to give some deference to

  15    his views as a physicist to what is a reasonable measure of

  16    effectiveness of journals?

  17         A.    No.  If he received those same honors in

  18    economics and information science, I would give deference;

  19    and to the same degree that I don't think he would respect

  20    my opinion on physics too much, or should he.

  21         Q.    You have never cited a physics article, have you?

  22         A.    Not that I can recall.

  23         Q.    Dr. Barschall did that many times, did he not?

  24         A.    Yes.

  25         Q.    He knew what was involved in that process?




                                                                204


   1         A.    In citing a physics article, yes.

   2         Q.    And do you know that he had examined the

   3    relationship -- he had examined citation data, had he not?

   4         A.    Yes.

   5         Q.    And he concluded that it was a reliable measure;

   6    is that true?

   7         A.    He concluded that, yes.

   8         Q.    Is it true that you do not know very much about

   9    impact factors?

  10         A.    No.

  11         Q.    Not true?

  12         A.    No.

  13         Q.    In what connections have you worked with impact

  14    factors?

  15         A.    In the connection, obviously, of this case, prior

  16    to this, with my work on the economics of information, that

  17    it would be simply part of the entire universe of

  18    information science that one is aware of.

  19         Q.    Have you ever used impact factors in one of your

  20    own analyses?

  21         A.    No.

  22         Q.    Have you ever had occasion to study impact

  23    factors and how the data is presented and examine it for

  24    different journals?

  25         A.    Have I looked at the impact factor numbers?




                                                                205


   1         Q.    Apart from in connection with this case, have you

   2    had occasion to work with impact factor data?

   3         A.    I've seen the impact factor numbers in the

   4    Science Citation Index prior to this case, but I have not

   5    worked with those numbers.

   6         Q.    What about citation data?  Have you worked with

   7    citation data?

   8         A.    Again, I have seen the citation numbers, but I

   9    have not worked with them in research.  I have also -- am

  10    familiar with research that used citation numbers and impact

  11    factors, but I have not used citation numbers myself or

  12    impact factor numbers.

  13         Q.    You mean the articles relating to citation data?

  14    Is that what you're referring to, what you are familiar

  15    with?

  16         A.    Articles relating to citation data.

  17         Q.    And you looked at those articles in connection

  18    with this case?

  19         A.    Well, I looked at some of those articles prior to

  20    this case, too.

  21         Q.    It's true that even when you collected the

  22    citation data to present to the Tenure Review Committee, you

  23    didn't gather that data yourself, did you?

  24         A.    Correct.

  25         Q.    You sent someone else to do it?




                                                                206


   1         A.    I had a graduate student do it.

   2         Q.    Could you look at Plaintiff's Exhibit 3 again,

   3    the Physics Today article.

   4               In the first paragraph, do you see the statement

   5    at the end of the first paragraph, after Professor Barschall

   6    refers to the cost per thousand characters and then to the

   7    impact factor?  He then states as follows:  "The ratio of

   8    these two measures is perhaps the best indicator of a

   9    journal's cost effectiveness."  Do you see that statement?

  10         A.    Yes.

  11         Q.    Do you disagree with that statement?

  12         A.    Yes.

  13         Q.    Apart from that statement -- strike that.

  14               Does a similar statement appear in the Bulletin

  15    of the American Physical Society?

  16         A.    Yes.  On the top of page 1438, under "Cost per

  17    Impact," the second sentence.

  18         Q.    And there it's just slightly different.  It

  19    reads, "This ratio is perhaps the most significant measure

  20    of the cost effectiveness of the journal"?

  21         A.    Correct.

  22         Q.    Referring again to the same ratio?

  23         A.    Yes.

  24         Q.    And you disagree with that statement?

  25         A.    Yes.




                                                                207


   1         Q.    Am I correct that, apart from those two

   2    statements, there are no statements in either PX 2 or PX 3,

   3    whether express or implied, that in your judgment are false?

   4         A.    I would say that the implication of these

   5    articles is that this information can be used for collection

   6    development, which I think is a false implication.

   7         Q.    Apart from that, there is no statement in either

   8    article that in your judgment is false; is that correct?

   9         A.    (Pause)  I would say that that implication,

  10    actually, was borne out in the last paragraph of the Physics

  11    Today article.

  12         Q.    I'm not sure if you're addressing my question.

  13         A.    Oh, the implication -- I said that the

  14    implication of this having influence on collection

  15    management, it's false that one should use this kind of

  16    procedure for collection development in a library, and

  17    you're saying is there any other statement --

  18         Q.    Any other statement, express or implied, in

  19    either article that in your judgment is false.

  20               MR. PLOTZ:  With respect, I think Dr. Kingma was

  21    answering the question.

  22         A.    I guess what I'm -- in trying to narrow down

  23    that, in turning to the last paragraph here, there are a

  24    couple of statements that I would disagree with Professor

  25    Barschall on.




                                                                208


   1         Q.    What document are we on?

   2         A.    We are in Physics Today, the last paragraph.

   3    It's really just --

   4         Q.    Can we -- you'll have an opportunity --

   5               MR. PLOTZ:  He's in the middle of an answer,

   6    Judge.

   7               MR. HUVELLE:  He's not answering my question,

   8    though.

   9               THE COURT:  Let him complete his answer.  The

  10    question was that, apart from those two statements and the

  11    implications derived from those two statements, is there

  12    anything else in these two articles that you regard as being

  13    false?

  14               THE WITNESS:  I think the implications of Table

  15    2, that one that should compare publishers on this measure,

  16    again, is a false comparison more making collection

  17    development decisions.

  18         Q.    Anything else?

  19         A.    No.  I believe that's it.

  20         Q.    Do you see -- there are no other statements,

  21    express or implied, in either article that in your judgment

  22    are false?  Correct?

  23         A.    Correct.

  24         Q.    You were deposed in this case on April 18?

  25         A.    Correct.




                                                                209


   1         Q.    As of that date, you had spent between 40 and 80

   2    hours working on this matter?

   3         A.    Since April 18?

   4         Q.    No.  Before that.

   5         A.    Oh, before that date.  Correct.

   6         Q.    How many hours have you spent on this matter

   7    since April 18?

   8         A.    Somewhere between 10 and 20.

   9         Q.    When did you resume work on this matter?

  10         A.    Worked briefly on it over the weekend -- it was

  11    probably a couple of days before that -- looking at some of

  12    these issues, and then obviously within trial right here.

  13         Q.    Have you done any analysis in connection with

  14    this matter since April 18?

  15         A.    Analysis of the numbers?

  16         Q.    Any further analysis that goes beyond what you

  17    had done as of April 18?

  18         A.    No.

  19         Q.    Did you review any additional articles or other

  20    written materials that you had not seen prior to April 18?

  21         A.    No.

  22         Q.    With whom have you spoken about this matter since

  23    April 18?

  24         A.    About this matter, the attorneys, Mr. Plotz,

  25    Mr. Lupert, the other attorneys.  I believe that's it.




                                                                210


   1         Q.    Am I correct, then, that you have not learned

   2    anything new regarding this matter except -- since April

   3    18 -- except what was told to you by the attorneys?

   4         A.    Correct.

   5         Q.    I would like to show you a copy of Defendants'

   6    Exhibit CCC.  Could you identify this document?

   7         A.    This is a report I prepared for this case.

   8         Q.    Did you testify on April 18 that it was a full

   9    and accurate statement of the opinions to which you expected

  10    to testify?

  11         A.    Yes.  I did make a couple of corrections on it.

  12         Q.    There were two minor corrections that you made?

  13         A.    Correct.

  14         Q.    Those were on page 7?

  15         A.    Correct, along with a clarification of the

  16    benefit effectiveness efficiency issue that we spoke of at

  17    the deposition.

  18         Q.    Apart from those corrections, this report, you

  19    testified, was a full and accurate statement of the opinions

  20    that you held and about which you expected to testify?

  21         A.    Correct.

  22         Q.    In your testimony yesterday, you asserted that

  23    Professor Barschall erred in comparing journals of different

  24    types; namely, review journals, letters journals, and

  25    archival journals, and also journals in different physics




                                                                211


   1    disciplines.  Do you recall that testimony?

   2         A.    Correct.

   3         Q.    That was one of the major criticisms that you

   4    articulated yesterday?

   5         A.    Correct.

   6         Q.    Can you refer me to the place in your report

   7    where you discuss that major criticism?

   8         A.    On page 5, No. 6, where it talks about variations

   9    in citation rate related to type of publication,

  10    nationality, time period, and size and type of specialty.

  11         Q.    Is there anything about review journals?

  12         A.    No.  It just says "variations in type of

  13    publication."  It doesn't specific -- you're right, it

  14    doesn't specifically mention review journals.

  15         Q.    And that was subparagraph 6?

  16         A.    6 on page 5, right.

  17         Q.    Now, on page 1 of your report, you refer -- and

  18    I'm going back to the correction that you made during your

  19    deposition --

  20         A.    OK.

  21         Q.    You referred to a cost/benefit ratio.  Do you see

  22    that?

  23         A.    Yes.

  24         Q.    You agree now that Dr. Barschall did not purport

  25    to prepare a cost/benefit ratio?




                                                                212


   1         A.    Correct.

   2         Q.    And that that is a different kind of analysis?

   3         A.    Related but different kind, correct.

   4         Q.    Different from cost effectiveness, which is what

   5    he did?

   6         A.    Correct.

   7         Q.    And so you -- all of your criticisms with regard

   8    to the adequacy of the cost/benefit ratio are misplaced; is

   9    that correct?

  10         A.    No.  It's, as I said at the deposition, even if

  11    you substitute the word "effectiveness" for "benefit" here,

  12    all the criticisms remain.  It's just a matter of using the

  13    word "benefit" instead of "effectiveness."

  14         Q.    Why did you use the wrong term there?

  15               MR. PLOTZ:  Objection.

  16         A.    Well --

  17               MR. PLOTZ:  Objection to the characterization,

  18    your Honor.

  19               THE COURT:  Overruled.

  20         A.    Cost/benefit analysis is a term in the common

  21    vernacular of economists and those who work with economists,

  22    and it's actually sort of the encompassing of all the types

  23    of analysis.  Cost-effectiveness is a piece.  If you think

  24    about a Vien diagram with the largest circle on the outside

  25    being cost/benefit analysis, cost-effectiveness analysis




                                                                213


   1    falls within that, and cost-efficiency analysis falls within

   2    that circle.

   3               Cost/benefit, using that term is really more of a

   4    sort of global term, which included many things that it

   5    shouldn't have.  But it's -- I think it's one that's

   6    throughout the sort of economics vernacular.  You know, this

   7    is "echo-speak" in some sense that people understand, using

   8    the term cost/benefit.

   9         Q.    Did you analyze, in your report, did you analyze

  10    his work with reference to the standards for cost-

  11    effectiveness studies or cost/benefit studies?

  12         A.    As I said, since "effectiveness" falls within the

  13    rubric of "benefit," it in some sense would be both,

  14    although I didn't analyze it with respect to him estimating

  15    the dollar value of a journal subscription.

  16         Q.    But in some senses it would be different?

  17         A.    Right.  The difference is inserting the term

  18    "dollar."

  19         Q.    The difference is also in terms of the

  20    comprehensiveness of the analysis of cost effectiveness?

  21         A.    Correct.

  22         Q.    My question is, when you judged the article for

  23    purposes of the report, did you use the wrong term or the

  24    wrong analysis?

  25         A.    It would have been more correct for me to use the




                                                                214


   1    effectiveness term, the cost-effectiveness term.

   2         Q.    But do you agree that you also used the wrong

   3    analysis in judging his report?

   4         A.    No.

   5         Q.    Can you look at the second paragraph on page 2?

   6         A.    Yes.

   7         Q.    Do the statements in that paragraph apply if

   8    you're judging his work in terms of cost-effectiveness?

   9         A.    No.  That statement refers to cost/benefit,

  10    right.

  11         Q.    So that's an indication that in fact you were

  12    judging his work by -- in terms of the wrong analysis.  You

  13    were applying the wrong economic construct to his work.

  14         A.    Just for that -- those single two senses, you can

  15    say that, but for the rest of the report, no.

  16         Q.    Are you saying that, for most of the report, you

  17    analyzed it in terms of the cost effectiveness study, but

  18    for those, that one paragraph, you deviated and analyzed it

  19    in terms of cost/benefit?

  20         A.    Well, as I said before, cost-effectiveness --

  21    cost/benefit is sort of the big circle and within that is

  22    cost effectiveness.  So analyzing it by the standards of

  23    cost-effectiveness by definition fall within a cost/benefit

  24    analysis.

  25         Q.    But a perfectly proper cost effectiveness




                                                                215


   1    analysis may -- indeed will -- fall short of a proper

   2    cost/benefit analysis --

   3         A.    Right.

   4         Q.    -- necessarily?

   5         A.    Yes.

   6         Q.    And there is no doubt in your mind that Professor

   7    Barschall had done a cost-effectiveness study?

   8         A.    Correct.

   9         Q.    And also that major components of it were a cost

  10    efficiency study?

  11         A.    By definition, you can't do a cost effectiveness

  12    without cost efficiency.

  13         Q.    So paragraph 2 is a mistake on your part?

  14         A.    No, it's -- I mean, the -- both sentences in

  15    paragraph 2 are correct.

  16         Q.    If you're talking about a cost/benefit study.

  17         A.    Correct.

  18         Q.    Which we are not.

  19         A.    Correct.

  20         Q.    Paragraph 3 on that page, am I correct in

  21    understanding that the principal concern expressed there

  22    with respect to Barschall's cost analysis is that a journal

  23    might add pages and pages of meaningless numbers and

  24    therefore lower its cost per impact; is that what you're

  25    saying?




                                                                216


   1         A.    Correct.

   2         Q.    Do you think that Professor Barschall assumed

   3    that his colleagues in the field of physics would not add

   4    pages and pages of meaningless data in order to affect the

   5    cost-per-impact ratio?

   6         A.    I don't pretend to know what he assumed.

   7         Q.    It would have been reasonable of him to make that

   8    assumption?

   9         A.    It's not always his colleagues in the field,

  10    though.  It's also the publishers that are making that

  11    decision about what is included with an article, whether or

  12    not data might be included or additional pages might be

  13    included.

  14         Q.    Do you believe that the points you make in this

  15    paragraph are a serious criticism of his work?

  16         A.    Yes.

  17         Q.    Can you look at page 3 of your report, the first

  18    full paragraph consisting of two sentences.

  19         A.    Yes.

  20         Q.    Do the statements in that paragraph have any

  21    relation to a cost effectiveness study?

  22         A.    It relates to benefit, so in this case you can

  23    substitute the word "effectiveness" for "benefit" and the

  24    statements are true.

  25         Q.    But do the statements have any relationship to




                                                                217


   1    your analysis of Dr. Barschall's cost-effectiveness study?

   2         A.    Yes, because I believe effectiveness is more

   3    appropriately measured by demand or use than by impact or

   4    citation.

   5         Q.    Paragraph 4 -- I'm sorry, page 4.  You criticize,

   6    starting on page 4, citation data; is that correct?

   7         A.    Correct.  I paraphrased from an article by

   8    MacRoberts and MacRoberts.

   9         Q.    You copied the seven points that MacRoberts and

  10    MacRoberts made?

  11         A.    Right.

  12         Q.    You didn't add anything new?

  13         A.    Correct.

  14         Q.    Is it true that many of the criticisms that

  15    MacRoberts and MacRoberts had suggested pertain only if

  16    you're using citation data to evaluate individual authors,

  17    not if you're using them to evaluate journals; is that

  18    correct?

  19         A.    I don't know if that's correct, because you can't

  20    make statements about the degree to which these problems

  21    influence different journals, or at least I'm not familiar

  22    with any research that has been done that have looked at the

  23    differences in citation or the error rates in citations and

  24    the possible differences between different journal titles.

  25         Q.    Well, one of the identified flaws is that an




                                                                218


   1    author may not include his middle initial, and therefore in

   2    terms of citation data the article may not be attributed to

   3    the right person.  That doesn't have anything to do with the

   4    assessment of journals, does it?

   5         A.    Can you point to that on this report?

   6         Q.    Page 6, Subpart B.

   7         A.    Correct.

   8         Q.    The same is true of A and C?

   9         A.    Correct.

  10               THE COURT:  How much longer will you be on

  11    cross-examination?

  12               MR. HUVELLE:  It will be a while.

  13               THE COURT:  An hour?

  14               MR. HUVELLE:  Maybe.

  15               THE COURT:  Half an hour?

  16               MR. HUVELLE:  Between that.

  17               THE COURT:  We will take a recess.

  18               MR. HUVELLE:  Thank you.

  19               THE COURT:  Off the record.

  20               (Discussion off the record)

  21               (Recess)

  22    BY MR. HUVELLE:

  23         Q.    Dr. Kingma, just a couple more questions on your

  24    report.  Can you turn to page 7.

  25               At the end of the first paragraph, with regard to




                                                                219


   1    impact factors, you state, "This truncation of the year of

   2    publication and the years of citations biases the measure of

   3    impact."  Do you see that?

   4         A.    Yes.

   5         Q.    The use of an impact factor precisely records the

   6    data for the period that it purports to cover; is that true?

   7         A.    Correct.

   8         Q.    And in the next paragraph, you make a couple of

   9    errors in your description of the impact factor, in the

  10    same -- in the following paragraph.  Do you see that?

  11         A.    Yes.  That's with my corrections, when my

  12    corrections came up.

  13         Q.    Did you misunderstand how the impact factor was

  14    calculated when you wrote this?

  15         A.    No.

  16         Q.    The changes would be in the first sentence, the

  17    year 1986 should read 1984 and 1985?

  18         A.    Right.  It's correctly described in the paragraph

  19    above that.

  20         Q.    And then the next sentence, again, the year 1986

  21    should read 1984 and 1985?

  22         A.    Correct.

  23         Q.    And in the next sentence, "a single year" should

  24    be "two years?"

  25               I'm sorry. "Two years" should be "one year."




                                                                220


   1         A.    Right.

   2         Q.    And "one year" should be "two years."

   3         A.    Correct.

   4         Q.    Page 9, you talk about the problems in counting

   5    pages.

   6         A.    Correct.

   7         Q.    Let me just ask you to look at a copy of Physical

   8    Review B.

   9               This has been marked as Exhibit XXXXX, is that

  10    correct?

  11         A.    Correct.

  12         Q.    And this is a journal published by the American

  13    Physical Society?  I will represent that to you.  And it

  14    says so on the front page.  And it's one of the journals

  15    included in Dr. Barschall's study.

  16               The concern you had about blank pages does not

  17    apply to this journal, does it?

  18         A.    It would appear with the issue in my hand that

  19    there aren't any completely blank pages that carry page

  20    numbers.  There are obviously half-blank pages, where one

  21    article ends and the next article begins on the next page.

  22    For example, on page 4699, half the page is blank, whereas

  23    half the page carries the end of the article.

  24         Q.    But it's not the case that there are full blank

  25    pages?




                                                                221


   1         A.    That are numbered.  There is a blank page but

   2    it's not numbered, near the end.

   3         Q.    Let me show you, for example, Ferroelectrics,

   4    Exhibit VVV, just one example there.  There's a blank page

   5    that's numbered.

   6         A.    Correct.

   7         Q.    So the journals that would be advantaged by this

   8    methodology would be one such as Ferroelectrics that number

   9    blank pages, correct?

  10         A.    The journal -- it's correct that the journals

  11    that would be advantaged by this methodology are ones that

  12    carry blank pages.  I don't know if these two issues are

  13    representative of Ferroelectrics and Physical Review B.

  14         Q.    Well, I'm not going to ask you to -- we'll let it

  15    stand there.

  16               But in any event, no matter whose advantage, it's

  17    not a very important point in terms of the scope of

  18    Dr. Barschall's study, is it?

  19         A.    It depends on how many pages were blank or how

  20    many pages were recognized.  It's also how many pages carry

  21    Roman numerals.

  22         Q.    Can you turn to page 10 of your report, Character

  23    Counting Bias.

  24               You refer to Dr. Barschall's methodology of

  25    counting a page which contains figures or tables as if it




                                                                222


   1    were completely filled with text.  Do you see that?

   2         A.    Yes.

   3         Q.    And you state that this methodology inherently

   4    discounts the value of figures, tables, and display

   5    mathematics; is that correct?

   6         A.    Correct.

   7         Q.    But your statement is not correct, is it?

   8         A.    Pardon me?

   9         Q.    Your statement is not correct?

  10         A.    That this bias is the cost per impact measure

  11    against journals?

  12         Q.    That it inherently discounts the value of

  13    figures, tables and displays.  What it does is treat a page

  14    containing figures, tables and displays as equivalent to a

  15    page of text?

  16         A.    It treats them as equivalent in size, even though

  17    they might have different production costs.

  18         Q.    But production costs should have nothing to do

  19    with Dr. Barschall's analysis?

  20         A.    Well, the price is part of his analysis.

  21         Q.    And the price is the price and it doesn't matter

  22    what the production cost is, does it?

  23         A.    Oh, it -- part of his analysis, he separated out

  24    translation journals for those production costs, so --

  25         Q.    Isn't it true that translation journals are




                                                                223


   1    journals that rarely have impact factors?

   2         A.    To my knowledge that's true.

   3         Q.    So he could not have done a cost-to-impact factor

   4    ratio for translation journals, or he would have to do it

   5    for such a limited sample that it wouldn't do much good?

   6         A.    I'm not sure that's the reason he gave for

   7    separating out translation journals.

   8         Q.    But it is a fact that he -- he did not have the

   9    data to do the full analysis with respect to translation

  10    journals?

  11         A.    I don't know if he had the data or not for

  12    translation journals or how many translation journals he had

  13    data on.

  14         Q.    But in any event, for purposes of Dr. Barschall's

  15    analysis, what's important to the librarian is the price,

  16    not how much it costs the publisher to compile certain

  17    pages?

  18         A.    The price is an important component of the

  19    analysis for a librarian.

  20         Q.    It's price and effectiveness?

  21               MR. PLOTZ:  He hadn't finished, your Honor.

  22         A.    The tables and display mathematics, one would

  23    assume, have bearing on the effectiveness, have some

  24    influence on the quality of information within there.  Now,

  25    if you remove it -- as we said before, this is any cost




                                                                224


   1    effectiveness, cost/benefit, cost efficiency analysis -- not

   2    cost efficiency.  Cost effectiveness, cost/benefit analysis

   3    has two parts to it, the cost part and the effectiveness

   4    part, and while you can take table and display mathematics

   5    and pull it out of the cost part, it winds up in the

   6    effectiveness part.

   7         Q.    This is not pulling it out of the cost side, is

   8    it?

   9         A.    His methodology is, by making an equivalence

  10    between characters and a table of the same size.

  11         Q.    It's assigning a value to mathematical tables?

  12         A.    Correct.

  13         Q.    Now, you may think mathematical tables have a

  14    value that's higher than text or lower than text, but I

  15    don't know -- do you have a way of assigning a value to

  16    mathematical tables in comparison to text?

  17         A.    He used a formula of one to one, basically.

  18         Q.    Right.

  19         A.    And he could have used a formula of 100 to 1

  20    or --

  21         Q.    1.2 to one.

  22         A.    -- 1.2 to one.  It's just an arbitrary pick.

  23         Q.    But you don't know if he discounts the value of

  24    tables.  All you know is he treats it as equivalent to text.

  25         A.    He treats it as equivalent to text, correct.




                                                                225


   1         Q.    And can you look at Phys. Rev. B.  Isn't it true

   2    that that journal, for example, has many tables, display

   3    mathematics, and figures?

   4         A.    Correct.

   5         Q.    You haven't looked at the journals in the study

   6    so you don't know whether or not it's true that all of them

   7    have such figures?

   8         A.    Correct.

   9         Q.    With regard to the including of journals, physics

  10    journals in different disciplines, review journals and

  11    letters journals, did you say that that is like comparing

  12    apples to oranges?

  13         A.    Correct.

  14         Q.    What did you mean by that?

  15         A.    Well, there are different -- they are different

  16    types of journals and carry with them different citations

  17    based on the type of journal it is.

  18         Q.    And you know that because you have compared them?

  19         A.    Because I have seen the numbers on them in

  20    reading the literature, correct.

  21         Q.    Not all the --

  22         A.    I think -- I think -- I'm sorry.  But I think

  23    that even Professor Barschall points to this in one of his

  24    articles.

  25         Q.    In fact, that's where you learned it?




                                                                226


   1         A.    No, I wouldn't say that I learned it from there,

   2    but it's one source of --

   3         Q.    Have you ever read a review journal in physics?

   4         A.    No, I have never read physics journals.

   5         Q.    Do you have any personal knowledge as to the

   6    citation rates for these different kinds of journals?

   7         A.    From economics, not from physics.

   8         Q.    In your readings of economics, they discussed the

   9    citation rates of review journals in physics?

  10         A.    No.  My readings of the citation rates of letters

  11    journals and economics and original research journals and

  12    economics.

  13         Q.    Which you have done in connection with this case?

  14         A.    No, which was prior to this case.

  15         Q.    Is it true that not all review journals listed by

  16    Dr. Barschall have higher impact factors than archival

  17    journals?

  18         A.    I don't know.  I would have to look at the table.

  19    Do you want me to look at the tables?

  20         Q.    Yes.  Table 1 of PX 3.

  21         A.    Correct.  There are some review journals with

  22    lower impact numbers than some original research journals.

  23    I do not know -- I have to admit I do not know, under the

  24    review journals, what specialty or subspecialty these

  25    particular review journals might be reviewing.




                                                                227


   1         Q.    I believe you noted yesterday that Review of

   2    Modern Physics has an impact ratio of -- impact factor of

   3    27; is that correct?

   4         A.    For 1986, correct.

   5         Q.    Right.  And that's more than three times higher

   6    than any other review journal?

   7         A.    I believe so, correct.

   8         Q.    Do you think that's a fact that a librarian ought

   9    to know in connection with acquisition decisions, that

  10    Review of Modern Physics has a very high impact factor?

  11         A.    No.

  12         Q.    Doesn't that number tell the librarian that the

  13    articles in the Review of Modern Physics are relied upon

  14    with extremely high frequency by scientists?

  15         A.    It tells the librarian something about how

  16    physicists as a whole cite articles in the past two years

  17    from Review of Modern Physics, but that librarian has to be

  18    more concerned about how her or his particular set of

  19    patrons might use the Review of Modern Physics, and it might

  20    be that they use it a lot, but it also might be that they

  21    have better use of another journal.

  22         Q.    It suggests that if they -- even if they are

  23    reading other articles, that an article -- a journal that

  24    will be frequently cited in other journals is Review of

  25    Modern Physics; it tells you that, doesn't it?




                                                                228


   1         A.    Correct.

   2         Q.    It tells you that Review of Modern Physics is the

   3    Michael Jordan of physical journals, doesn't it?

   4         A.    I don't know if I would make that comparison.

   5         Q.    Well, 27 impact factor, the next one 7 in the

   6    same category; isn't that like scoring 63 points in a

   7    playoff game?

   8         A.    It certainly is more than three times higher than

   9    the next journal.

  10         Q.    You can say you can't compare Michael Jordan to

  11    other basketball players, can't you?  People say that, but

  12    you can?

  13         A.    They are using it in a different context,

  14    obviously.

  15         Q.    You can make a comparison.  The comparison is,

  16    he's better.  Right?

  17         A.    Correct.

  18         Q.    And that's what this data tells you, as to

  19    physical -- Review of Modern Physics, that, as compared to

  20    other physical -- physics journals, it is cited much more

  21    frequently, and a librarian might want to know that; is that

  22    correct?

  23         A.    I guess that's where the analysis sort of

  24    crumbles, and that we can say Michael Jordan is better in

  25    terms of all basketball players and we can say the Review of




                                                                229


   1    Modern Physics is more cited in terms of all these physics

   2    journals.  But can we say that the Review of Modern Physics

   3    is the best journal for our patrons in our library?  That's

   4    a different question.  Can we say that Michael Jordan is the

   5    best basketball player for our basketball team?  I don't

   6    know.  The length of magic might be might be very broad

   7    compared to any other player.

   8               THE COURT:  You said that the impact factor of 27

   9    was not relevant to the librarian whose concern should be

  10    the needs of the patrons of the --

  11               THE WITNESS:  Their library.

  12               THE COURT:  -- particular library.

  13               THE WITNESS:  Right.

  14               THE COURT:  Well, if we are dealing, for example,

  15    with a large university, isn't it reasonable to assume that

  16    the interests of the users of that library will bear some

  17    relationship to the world of physicists as a whole?

  18               THE WITNESS:  It depends on what specialties that

  19    university might support.

  20               THE COURT:  If you are dealing with a library for

  21    a commercial entity which is engaged in some very refined

  22    form of physical research, that might not be relevant, but

  23    for let's just say university libraries, would it not be

  24    relevant for a librarian in a university library to know

  25    that the Review of Modern Physics is more cited than any




                                                                230


   1    other such review?

   2               THE WITNESS:  Let me go back to the article by

   3    Bensman which showed that there wasn't a correlation or a

   4    very low correlation between impact factor and faculty

   5    ranking of journals.  And clearly within any discipline,

   6    there might be some large general journals that are

   7    important, but even universities are becoming very

   8    specialized, each and every department, and with those

   9    specialties there is typically a core of niche journals that

  10    support each specialty.

  11               THE COURT:  Without meaning to disparage

  12    academia, for which I have great respect, isn't one of the

  13    needs which a researcher has to have sources that can be

  14    cited for the particular propositions which are discussed in

  15    the article?  So isn't the citation itself a form of use of

  16    the publication?

  17               THE WITNESS:  Right.  You would have to use it in

  18    some sense to cite it, although sometimes people might cite

  19    things just because they read them in other things.

  20               THE COURT:  Sometimes you pad citations, like --

  21               THE WITNESS:  Right.

  22               THE COURT:  -- law clerks do when they do draft

  23    opinions.

  24               THE WITNESS:  The trick -- the trick is, though,

  25    that --




                                                                231


   1               THE COURT:  But you seem to be making a

   2    distinction between citation and use.

   3               THE WITNESS:  They are two different things.

   4               THE COURT:  Well, use is the greater, but isn't

   5    citation a form of use?

   6               THE WITNESS:  Yes.  What citations recognize,

   7    too, is a global measure of how much everybody cites it.

   8    Use within your library is a local measure of how much

   9    people -- the patrons you support are going to use a

  10    particular journal.

  11               THE COURT:  Let me just -- with respect to

  12    general libraries such as in universities, is it reasonable

  13    to assume that the interests or needs of a person who is to

  14    utilize the library will bear a relationship to the

  15    interests and needs of physicists at large?

  16               THE WITNESS:  Actually, universities over the

  17    last two decades have become more and more specialized, so,

  18    no, it's not -- you can't make that comparison anymore.

  19    Universities' libraries just simply can't buy everything

  20    that they want, and universities can't support all

  21    specialties that exist.  So many of them tend to limit that

  22    and decide to collect a set of faculty that have a certain

  23    niche themselves.

  24               At the university at Albany, for example, the

  25    Economics Department has a handful of faculty that




                                                                232


   1    specialize in public finance and a handful of faculty that

   2    specialize in econometrics, but we don't have -- we don't

   3    have a lot of economists that do agricultural economics or

   4    consumer economics.  We have one that does labor economics

   5    but not more than one.

   6               So we -- university departments are becoming more

   7    and more specialized.

   8               THE COURT:  But how about at the undergraduate

   9    level?  Is the nature of these publications such that the

  10    undergraduate level is not so relevant?

  11               THE WITNESS:  The undergraduate level is not as

  12    relevant because of -- you know, the journal publications

  13    themselves report research that is digestible to faculty and

  14    graduate students.  Some undergraduates certainly use those

  15    journals.  But it's, you know, it's -- I would say

  16    undergraduates don't go off and do research and do research

  17    to the degree necessary to have academic journals constantly

  18    at their disposal.

  19               THE COURT:  But they write papers?

  20               THE WITNESS:  They do write papers.  They do

  21    write papers, yes.

  22               THE COURT:  Which gets them in the library, one

  23    hopes?

  24               THE WITNESS:  Yes, one hopes.

  25               MR. HUVELLE:  I just have a couple of minutes.




                                                                233


   1    BY MR. HUVELLE:

   2         Q.    The point at which you took issue with me, I

   3    believe once again, is you said that the librarian is going

   4    to look at it differently.

   5         A.    Look at the impact number differently, correct.

   6         Q.    And am I correct in understanding from your

   7    testimony that you have emphasized the need of the librarian

   8    to look at the local community?

   9         A.    Correct.

  10         Q.    And I'm going to use a word that's sometimes used

  11    pejoratively, but I think it's the precise word.  Your

  12    perspective, or the perspective you advocate for the

  13    librarian, is a parochial one?

  14         A.    Meaning --

  15         Q.    Local.

  16         A.    -- a local one, yes.

  17         Q.    And of course libraries, even public libraries,

  18    often buy books based upon The New York Times best seller

  19    lists or a more global perspective of what's developing in

  20    the world of literature or whatnot, correct?

  21         A.    Correct.

  22         Q.    And Dr. Barschall was a respected physicist at

  23    the University of Wisconsin, which is a major center for

  24    studies in physics?  He was a member of the National Academy

  25    of Sciences, correct?




                                                                234


   1         A.    Correct.

   2         Q.    And he thought it was important that librarians

   3    know the global perspective of what journals were important

   4    to the scientists who are writing articles and developing

   5    knowledge in the sciences; is that correct?

   6         A.    He thought that was important, correct.

   7         Q.    And journals historically have played the key

   8    role in transmitting new knowledge from generation to

   9    generation in terms of science?

  10         A.    Correct.

  11               MR. HUVELLE:  No further questions.

  12               THE COURT:  Redirect?

  13               MR. PLOTZ:  Yes, your Honor.

  14    REDIRECT EXAMINATION

  15    BY MR. PLOTZ:

  16         Q.    Dr. Kingma, you have been asked a lot of

  17    questions about impact factor, a lot of questions about

  18    cost.

  19               With respect to the market for library journals,

  20    you were asked whether this was an example of an imperfect

  21    market or market with imperfect information?

  22         A.    Correct.  I was asked.

  23         Q.    What does it mean for a market to have imperfect

  24    information?

  25         A.    Imperfect information means that either the




                                                                235


   1    seller or buyer does not have complete information about the

   2    transaction or good.  Usually we talk about it as the

   3    consumer not having complete information about the product

   4    that a seller is offering.

   5         Q.    And just speaking generically, what is a solution

   6    for a market which has imperfect information?

   7         A.    Generically, to provide accurate information

   8    about the quality, typically the quality of the product.

   9         Q.    Is it your view that the ratio of cost per

  10    character over impact factor provides additional information

  11    that helps solve the problem of an imperfect market?

  12         A.    No.  The appropriate term here is having correct

  13    information.  Adding correct information to a market

  14    improves the efficiency of that market.  Likewise, adding

  15    misinformation to that market may actually push that market

  16    away from efficiency, in economic terms.  I don't consider

  17    this cost per character per impact measure to be correct

  18    information for managing -- making collection development

  19    decisions within libraries.

  20               THE COURT:  I take it that applies as well to the

  21    decision by the author, at least in Exhibit 3, the Physics

  22    Today, the pitch, if I may use that colloquial term, is to

  23    authors rather than librarians.  In the last -- I don't have

  24    it.  Isn't the last paragraph an exhortation to authors

  25    rather than librarians?




                                                                236


   1               THE WITNESS:  Yes.

   2         Q.    Do you know, by the way, whether the journals

   3    which have lower ratios of cost per impact have the capacity

   4    to publish additional articles that authors might submit?

   5         A.    I don't know.

   6         Q.    Do you know whether anyone before Dr. Barschall

   7    performed a cost comparison survey based on a ratio of cost

   8    per character over impact factor?

   9               MR. HUVELLE:  Your Honor, I think we are outside

  10    the scope of the cross.

  11               THE COURT:  No, I will allow it.

  12         A.    I don't know.

  13         Q.    You are not aware of --

  14         A.    I am not aware of other articles, prior to this.

  15         Q.    You testified on cross, with respect to the cost

  16    side of the ratio, that a normalization of cost did not

  17    provide useful information, correct?

  18         A.    A normalization based on cost per page per

  19    character per article analysis, correct.

  20         Q.    And why is that?

  21         A.    It's simply not the purchasing decision that the

  22    librarian is making.  A librarian has to be concerned about

  23    how much they are paying, and that is paying for the

  24    subscription as a whole, and how much use their patrons are

  25    going to receive from that.  And clearly that's use for the




                                                                237


   1    subscription as a whole.

   2         Q.    Now, with respect, for example, to normalizing

   3    price per page -- you were asked some questions about

   4    that -- why doesn't that provide useful information to a

   5    librarian making an acquisition decision?

   6         A.    I don't think normalizing by price per page

   7    provides useful information to a librarian.  It's not --

   8    it's not the relevant decision for a librarian who is

   9    purchasing a subscription as a whole.

  10         Q.    In terms of the subscription to a journal -- a

  11    journal contains information, doesn't it?

  12         A.    Correct.

  13         Q.    And in determining the value of a journal, is one

  14    of the things that one must look at the quality of the

  15    information that's in the journal?

  16         A.    Correct.

  17               THE COURT:  Well, are you assuming that all of

  18    the editorial material in all of these journals are subject

  19    to a process of peer review?

  20               THE WITNESS:  No, certainly not -- I mean, the

  21    articles typically are subject to peer review but not

  22    always.  You know, I -- again, I don't know within every

  23    physics journal that appears in this which ones were subject

  24    to peer review and which were not.  And clearly there's some

  25    information that was not subject to peer review, like




                                                                238


   1    editorials and such.

   2               THE COURT:  Whether a particular journal has peer

   3    review or not, is that something which a librarian would

   4    consider in making a subscription?

   5               THE WITNESS:  Personally, I think that those

   6    journals with peer review are going to be of higher quality.

   7    But, again, a librarian has to be concerned about what their

   8    patrons use.  And if that means their patrons would rather

   9    have a journal that's not peer-reviewed but it's in some

  10    specialty that they really need, then they shouldn't make a

  11    decision based on the fact that there is another journal

  12    that is peer-reviewed.

  13               THE COURT:  Let's assume that you have two

  14    journals, both of which are subject to peer review by

  15    referees of comparable knowledge or standing in the field

  16    and who apply the same criteria.  If you indulge in that

  17    assumption, is the number of pages per subscription dollar

  18    of greater relevance?

  19               THE WITNESS:  I don't see in which case they are

  20    ever of relevance.  I mean, it's how much you're going to

  21    pay for the subscription and how much use you're going to

  22    get out of it.

  23               THE COURT:  Isn't --

  24               THE WITNESS:  There are some --

  25               THE COURT:  Isn't the use to be made -- doesn't




                                                                239


   1    it have some relationship to the quality of the editorial

   2    matter?

   3               THE WITNESS:  One could assume that more quality

   4    peer review would result in more quality articles and that

   5    might result in more use.  But again, within, you know,

   6    within the local decision making, does that result in more

   7    use for my patrons?

   8               THE COURT:  Similar to the criticism that is made

   9    of some political figures who are said not to exercise any

  10    leadership but merely to follow popular polls, you are

  11    assuming that the librarian's role is simply to cater to an

  12    existing demand and that the librarian has no input as to

  13    the quality?

  14               THE WITNESS:  I truly believe that the

  15    librarian's role is to cater to, as you say it, existing

  16    demand and expected future demand, because they are

  17    entrusted with university dollars in order to purchase

  18    journal subscriptions that are going to be used by the

  19    clientele at that university.  And I think that is -- it's

  20    important for them in terms of the management decisions.

  21               THE COURT:  Let me make sure I understand what

  22    you're saying.  You're saying, if we have two journals

  23    subject to the same peer review by equally conscientious and

  24    knowledgeable referees, and for the same subscription price

  25    one has 50 pages and the other has 500 pages, that that




                                                                240


   1    price differential is not something which the librarian

   2    should consider?

   3               THE WITNESS:  The price per page, is that what

   4    you're saying?

   5               THE COURT:  Yes, the price per page or

   6    subscription price.  Yes, price per page.

   7               THE WITNESS:  Well, the subscription price is

   8    important.  The subscription price is important.  That's

   9    key.  I just disagree with normalizing it on a per-page,

  10    per-character basis.  It's the value of the information, and

  11    the information can't be divided up.

  12               THE COURT:  Well, but my hypothetical assumes

  13    that there is the same peer review by the same standards in

  14    both publications, both journals.  Doesn't that give some

  15    comfort that the quality level will be comparable?

  16               THE WITNESS:  If I -- if you can stretch your

  17    hypothetical one more assumption and say these are both on

  18    the same topic, OK, two journals on the same topic --

  19               THE COURT:  The same topic.

  20               THE WITNESS:  Same topic, same everything, same

  21    quality of referees.  Then I would say, actually, that the

  22    economics of that situation means that the one journal won't

  23    be sold.  I mean, if you're saying that it's the exact same

  24    topic, exact same referees, etc., ultimately the market will

  25    decide as to which of these journals people will subscribe




                                                                241


   1    to.  And it's a decision in some sense of, you know, how

   2    faculty will use -- hopefully how faculty will use those

   3    journals.

   4               THE COURT:  It's also a function of how

   5    knowledgeable the market is.

   6               THE WITNESS:  True.  The market needs accurate

   7    information to make those decisions.

   8    BY MR. PLOTZ:

   9         Q.    Dr. Kingma, on the cross-examination, there was

  10    an analogy of -- the grocery store analogy of pasta and

  11    beans.  Is considering journals in different topics

  12    something like considering a choice between pasta and beans

  13    in a grocery store?

  14         A.    I would say that's analogous.  It might even be a

  15    stronger difference among the journals, though.

  16         Q.    Why is that?

  17         A.    I mean -- well, pasta and beans are both sort of

  18    consumable items that, if I'm looking for a side dish that

  19    evening, I might eat either pasta or beans.  But if I'm a

  20    faculty member in one specialty, journals from the other

  21    specialty might be of no use or value to me at all, only the

  22    journals within my specialty.

  23         Q.    Would it make sense for a library to buy an

  24    inexpensive journal that's seldom used?

  25         A.    No.




                                                                242


   1         Q.    Would it make sense to buy an expensive journal

   2    that's seldom used?

   3         A.    No.

   4         Q.    Would it make sense to buy an expensive journal

   5    that is frequently used?

   6         A.    Yes, if there's enough use to justify that

   7    subscription price, yes.

   8         Q.    Just to complete the box, would it make sense to

   9    buy an inexpensive journal that's frequently used?

  10         A.    Yes.

  11         Q.    Now, in terms of, you said, in response to a

  12    question from Judge Sand a few minutes ago, that your view

  13    is that the librarian's function is to respond to the

  14    current needs of its patrons and the expected future needs,

  15    correct?

  16         A.    Correct.

  17         Q.    Who will predict what those expected future needs

  18    or trends are?

  19         A.    Typically the librarian is going to be making the

  20    decision about --

  21         Q.    Based on what information?

  22         A.    Based on the information they have available.

  23    Hopefully they have information on use.  Hopefully they have

  24    information on requests for different journal titles like

  25    interlibrary loan, or hopefully they have talked to the




                                                                243


   1    faculty to determine what journals would be the best buys

   2    for their library.

   3               THE COURT:  Wouldn't the key factor be to see

   4    what the curriculum will be for the near future?

   5               THE WITNESS:  I think that would be -- yes, I

   6    mean, that's extremely important.  Hopefully that will come

   7    out in the expressed wishes of the faculty.

   8               THE COURT:  My limited contact with academia in

   9    law school is that there is an inquiry made by the librarian

  10    with respect to what material the faculty members would like

  11    to see in the library in the light of what it is that they

  12    will be teaching.

  13               THE WITNESS:  Right.

  14               THE COURT:  Isn't that standard?

  15               THE WITNESS:  I don't know if that's standard at

  16    every library.  That's an important thing that should be

  17    done, correct.

  18    BY MR. PLOTZ:

  19         Q.    I want to turn for the moment to the denominator

  20    of the ratio impact factor.

  21               Citation analysis -- rather, citation count is

  22    not the same thing as impact factor, right?

  23         A.    Correct.

  24         Q.    In assessing whether or not either citation count

  25    or impact factor reliably measures the effectiveness or the




                                                                244


   1    benefit of a journal, one must consider what citations

   2    measure, right?

   3         A.    Right.

   4         Q.    And if there are flaws in terms of what it is --

   5    or the value of what a citation measures, would that have an

   6    adverse impact on the validity of citation count?

   7               MR. HUVELLE:  I object to the leading nature of

   8    this question.

   9               THE COURT:  Overruled.

  10         A.    Errors in citation or impact would affect using

  11    it, correct.

  12         Q.    Is that true in a cost/benefit analysis?

  13         A.    Correct.  I mean, given the errors that are known

  14    to exist in citation analysis, those errors will produce

  15    errors within subsequent analysis that relies on them,

  16    pretty much by definition.

  17         Q.    So they would produce errors in a cost

  18    effectiveness analysis?

  19         A.    Correct.

  20         Q.    Now, one of the -- I want to direct your

  21    attention to one of the articles that you cited in your

  22    report, an article by Moed and Vreins.  What did Moed and

  23    Vreins --

  24               THE COURT:  Now, aren't you really going entirely

  25    beyond the scope?




                                                                245


   1               MR. PLOTZ:  No.  Actually, this is going to be

   2    directly relevant to the cross with respect to whether or

   3    not the flaws in citation analysis are directed to

   4    individual authors or to journals as a whole.

   5               THE COURT:  You may continue.

   6    BY MR. PLOTZ:

   7         Q.    Just to put it in context, though, what issue

   8    were Moed and Vreins looking at?

   9         A.    Moed and Vreins were looking at clerical errors

  10    principally, by authors in citing materials.  They took a

  11    set of journal titles and then looked for whatever errors

  12    might have been produced by the authors in those citations.

  13         Q.    Were there any differences across the journals

  14    they looked at in the types of errors that they found?

  15         A.    They weren't going to find any errors in terms of

  16    journal titles, because of the way they collected the data,

  17    by establishing a set of journal titles.

  18         Q.    What errors were they looking at?  What types of

  19    errors?

  20         A.    They were looking at errors such as misspelling

  21    of names, misspelling of articles, etc.  And they found

  22    among those that there was roughly a 9-percent error rate

  23    among citations.

  24         Q.    Were there any differences among the journals

  25    they looked at in terms of the types of errors that were




                                                                246


   1    predominant in each journal?

   2               THE COURT:  When you say "a 9-percent error

   3    rate," do you mean an error of sufficient magnitude so that

   4    the correct reference was not credited with the citation?

   5               THE WITNESS:  In errors -- an error, period,

   6    meaning misspelling of a word, misspelling of a name.

   7               THE COURT:  With the consequence that the

   8    citation was not credited in the analysis?

   9               THE WITNESS:  They weren't looking at the

  10    ultimate citation database.  They were just looking at a

  11    collection of articles.  So you can't say it goes to the

  12    next step, with the consequence.

  13               THE COURT:  And they were saying?

  14               THE WITNESS:  That they were errors in the way

  15    authors put their citations together; they make mistakes.

  16               THE COURT:  For our purposes, the significant

  17    question is whether the nature of the errors was such as to

  18    skew the analysis?

  19               THE WITNESS:  Correct.

  20               THE COURT:  Were these errors of such a nature?

  21    If someone misspelled your name, puts an E at the end

  22    instead of an A, but in the analysis it's clear that it's

  23    your book which is being cited so that they, I guess, would

  24    call it a hit in today's language, then the error is of no

  25    significance?




                                                                247


   1               THE WITNESS:  Right.  I don't know the answer to

   2    that, your previous question.  That is what it comes down

   3    to.

   4    BY MR. PLOTZ:

   5         Q.    Leaving clerical errors aside, is there a

   6    subjective component to an author's decision to use a

   7    citation?

   8         A.    Clearly, there is subjectivity based on -- as was

   9    mentioned before, positive and negative cites carry the same

  10    weight.  Even though one is -- might not be favorable and

  11    the other is favorable, in the final analysis they both

  12    count as a hit.  There may be, you know, minor cites of some

  13    related research which is counted equally to a major piece

  14    of work that had major influence on your research.  But

  15    those, again, are both counted as equal hits.

  16         Q.    In connection with the question Judge Sand asked

  17    you a few moments ago, the relationship of citations and

  18    use, are there uses of journals beyond citation?

  19         A.    Yes, there are.

  20         Q.    What kind of uses would there be?

  21         A.    In reading the literature, knowing what is out

  22    there, looking for things that might be related to your

  23    research, there are, in the physical sciences, there are

  24    clearly a lot of uses within the corporate sector that don't

  25    result in citations.  This is well known.




                                                                248


   1               There are, you know, obviously, a body of uses

   2    beyond simply those that you are going to cite.

   3               THE COURT:  How much longer do you think redirect

   4    will be?

   5               MR. PLOTZ:  Longer than five minutes.

   6               THE COURT:  We are adjourned until 2 o'clock.

   7               (Luncheon recess)

   8

   9

  10

  11

  12

  13

  14

  15

  16

  17

  18

  19

  20

  21

  22

  23

  24

  25




                                                                249


   1                 A F T E R N O O N    S E S S I O N

   2                             2:05 p.m.

   3    BRUCE KINGMA,

   4               Resumed, and testified further as follows:

   5               THE COURT:  Good afternoon.  You may be seated.

   6    REDIRECT EXAMINATION (Resumed)

   7    BY MR. PLOTZ:

   8         Q.    Dr. Kingma, on cross you were asked a whole lot

   9    of questions concerning five theoretical measures of

  10    effectiveness of journals, do you recall that?

  11         A.    Correct.

  12         Q.    And just so we have them in front of us, what are

  13    those five theoretical measures?

  14         A.    The five theoretical measures are readership, or

  15    readings; use; survey of faculty opinions; citations; and

  16    impact.

  17         Q.    And is that list you just gave us in any

  18    particular order of importance?

  19         A.    Yes.  I regard it to be listed as best to worst,

  20    best being readings or use of a journal in terms of

  21    measuring effectiveness.

  22         Q.    Are any of those five measures in your view

  23    invalid measures of effectiveness of journals?

  24         A.    Yes.  I consider the impact factor to be an

  25    invalid measure of effectiveness.  It has serious problems




                                                                250


   1    inherent in it that don't make it a worthwhile measure.

   2         Q.    And what, briefly, are those problems?

   3         A.    The problems that we discussed before, of its

   4    dramatic variability from year to year, of it makes it

   5    difficult to compare cross subspecialties, the different

   6    kinds of types of journals, the inherent problems in

   7    citation analysis, and one more, the fact that it's a global

   8    measure rather than a local measure.

   9         Q.    In terms of its year-to-year fluctuations, you

  10    were asked some questions on cross relating to the review of

  11    modern physics impact factor as reported in the Barschall

  12    study, which was 27, correct?

  13         A.    Correct.

  14         Q.    Do you know what happened to that impact factor

  15    in the next year, 1987?

  16         A.    In 1987, actually there was a 40-percent decline

  17    in the impact factor down to a little more than 16.

  18         Q.    Now, what is it about citation count that you

  19    rank at above impact factor in your list of theoretical

  20    measures?

  21         A.    Well, citation count gives a more complete

  22    picture of a journal.  In some sense, if you wanted eve