Loading ...
Sorry, an error occurred while loading the content.
 

Re: GEOSTATS: philosophical question

Expand Messages
  • Charlie Hendricksen
    Friends, An extremely interesting question, one I ve raised often to colleagues in the Geography Department, and I ve also raised at team meetings working on a
    Message 1 of 5 , Jun 4, 2000
      Friends,

      An extremely interesting question, one I've raised often to
      colleagues in the Geography Department, and I've also raised at team
      meetings working on a large grant whose mission is to study ecological,
      social and human health impacts of environmental contamination at a
      single large site. The grant team also is charged with the
      responsibility of communicating the results to stakeholders.

      With respect to the three problems posed by Richard Hammond, here
      are my opinions:

      1) the client rarely understands how many subjective decisions go into
      each product

      Interpretation of the data that is the source of the presentation also
      has objective criteria which may or may not be correct.

      2) the client rarely cares

      The client may be more interested in the decisions that have to be made
      than in the information that the graphic conveys. In other words, the
      client is willing to accept professional interpretation. The client may
      be ignorant of the professional decisions made in interpretation, or may
      assume that professional standards prevailed. The client cares, but
      doesn't have the time to argue with the experts.

      3) they generally truly believe that the map IS reality, not mearly a
      subjective representation.

      Unfortunately this is far too often the case. Skepticism is often
      overpowered by hope for truth in a sea of uncertainty. The producing
      professional has a clear responsibility to warn the client of the
      uncertainties and also the responsibility to assist the client in
      interpretation of the product.


      There is an answer to all these problems if the producing consultant is
      willing to take the time, and if the client is willing to pay for the
      time (meaning that the proposal has to be clear about the Quality
      control and its costs).

      I have proposed including metadata to be included with the product.
      This metadata would have a section on data quality, a section on the
      design of the product, and above all a section on interpretation
      (caveats for the naive reader in particular). In an informationally
      advanced environment, this metadata would be hyperlinked to the
      product. In the ideal environment, that metadata, as well as the
      product itself, would be subject to shared annotation and discussion.
      The annotation and discussion would also be hyperlinked to the source
      documents.


      An example: In our work, we mapped well contamination data from three
      annual reports. The maps were made with five intervals, each color
      coded to point symbols (dots) for each well report. We (grad student
      and professional staff) decided to create an animated GIF of these three
      images, more for curiosity and to learn the tools than as a serious
      presentation. The animated GIF was spectacular, and was an immediate
      hit because anyone could call up the map with their web browser. The
      presentation appeared to demonstrate an improvement in contamination
      levels, which, if true would be nice. But the data was sampled so
      poorly that any conclusions were ludicrous. Some wells were not sampled
      every year, the wells were sampled on no discernible schedule, and wide
      areas were not sampled at all. In a bureaucratic perversion of science,
      some contamination levels were listed as rounded up values (if a trace
      was detected, make the value 10ppb). In short, there was no way that
      map could be represented as valid science.

      Nevertheless, the enthusiasm for this flawed movie was such that the
      cynics soon came up with a name for a universal strategy to be used to
      convince the naive: "Give 'em the blinkies." Suggestions to hyperlink a
      set of caveats to the map were not adopted. The map remains on the web
      site without comments, ready to mislead any viewer.
      --
      Charlie Hendricksen veritas@...

      "Information technology structures human relationships."

      Eeqmc11@... wrote:
      >
      > Greetings all - my first post
      >
      > After reading the complex problems roaming around this board, I am hesitant
      > to post such a question. However, the more I make maps, graphs and charts
      > (per GIS, etc.), the more I am struck with the awsome repsonsibility of the
      > map (chart, graph, etc.) maker. I find two distinct problems in creation -
      > 1) the client rarely understands how many subjective decisions go into each
      > product and 2) the client rarely cares. I find then a third problem with the
      > intended audience 3) they generally truly believe that the map IS reality,
      > not mearly a subjective representation.
      >
      > since all models share this same problem (i.e., no matter how complete, the
      > model is still a model and is therefore subjected to subjective decision
      > making throughout its lifetime), I was wondering if anyone would care to
      > comment on the implications.
      >
      > Not a very rigorus question, but perhaps some fun!!
      >
      > thanks,
      >
      > Richard Hammond
      > United States Environmental Protection Agency
      >
      > philosophical
      > --
      > *To post a message to the list, send it to ai-geostats@....
      > *As a general service to list users, please remember to post a summary
      > of any useful responses to your questions.
      > *To unsubscribe, send email to majordomo@... with no subject and
      > "unsubscribe ai-geostats" in the message body.
      > DO NOT SEND Subscribe/Unsubscribe requests to the list!
      --
      *To post a message to the list, send it to ai-geostats@....
      *As a general service to list users, please remember to post a summary
      of any useful responses to your questions.
      *To unsubscribe, send email to majordomo@... with no subject and
      "unsubscribe ai-geostats" in the message body.
      DO NOT SEND Subscribe/Unsubscribe requests to the list!
    • Tim Chatterton
      Richard, Many people also fail to realise that we have no access to anything other than subjective impressions of the world and therefore make claims about
      Message 2 of 5 , Jun 4, 2000
        Richard,
        Many people also fail to realise that we have no access to
        anything other than subjective impressions of the world and
        therefore make claims about science being objective or
        about truth. The problems concerning this issue are
        probably far more worrying.

        Tim.
        -------------------------------------------
        Tim Chatterton
        School of Environmental Sciences
        University of East Anglia
        Norwich
        Norfolk
        UK
        NR4 7TJ

        Tel:(01603)593115
        Fax:(01603)507719
        E-Mail: T.Chatterton@...




        --
        *To post a message to the list, send it to ai-geostats@....
        *As a general service to list users, please remember to post a summary
        of any useful responses to your questions.
        *To unsubscribe, send email to majordomo@... with no subject and
        "unsubscribe ai-geostats" in the message body.
        DO NOT SEND Subscribe/Unsubscribe requests to the list!
      • Steve Friedman
        Richard, I follow the approach that I m making models, like bridges, airplanes, etc. These just happen to be in paper map form rather than plastic. However,
        Message 3 of 5 , Jun 4, 2000
          Richard,

          I follow the approach that I'm making models, like bridges, airplanes,
          etc. These just happen to be in paper map form rather than
          plastic. However, the analogy seems valid because no one would expect to
          be able to use a model car, plane, or bridge, we intitively understand
          that it is a representation of reality.

          Now for the maps you are creating. True they are models of reality, but
          they have a characteristic that should be integrated directly in the
          legend box which should be inclucded in any map you distribute. It is
          very wise to include a few lines explaining various aspects of the source
          data, such as the scale at which the original data was collected, how that
          scale differs from the map presentation scale and a few other useful bits
          of qualitative information. In particular, you should also have a
          statement regarding the "reliablity" of the map, If you used statistics
          in the development of the final map, how does the map correspond to the
          95% confidence intervals. Additionally, how does the data depicted on
          the map relate to a spatial confidence zone (does that quality of the
          input data vary spatially within a single theme, or across multiple
          themes), do you integrate maps of varying spatial resolution?

          If you include these statements with the map, you identify all of the
          known limitations of the input data, as well as protect yourself by
          acknowledging the limitations of the final product. If end users choose
          to ignore these declarations, what can you say except that it is easy to
          lead a horse to water, but it is much harder to get them to drink.

          Cheers
          Steve Friedman


          On Sun, 4 Jun 2000 Eeqmc11@... wrote:

          > Greetings all - my first post
          >
          > After reading the complex problems roaming around this board, I am hesitant
          > to post such a question. However, the more I make maps, graphs and charts
          > (per GIS, etc.), the more I am struck with the awsome repsonsibility of the
          > map (chart, graph, etc.) maker. I find two distinct problems in creation -
          > 1) the client rarely understands how many subjective decisions go into each
          > product and 2) the client rarely cares. I find then a third problem with the
          > intended audience 3) they generally truly believe that the map IS reality,
          > not mearly a subjective representation.
          >
          > since all models share this same problem (i.e., no matter how complete, the
          > model is still a model and is therefore subjected to subjective decision
          > making throughout its lifetime), I was wondering if anyone would care to
          > comment on the implications.
          >
          > Not a very rigorus question, but perhaps some fun!!
          >
          > thanks,
          >
          > Richard Hammond
          > United States Environmental Protection Agency
          >
          >
          >
          >
          > philosophical
          > --
          > *To post a message to the list, send it to ai-geostats@....
          > *As a general service to list users, please remember to post a summary
          > of any useful responses to your questions.
          > *To unsubscribe, send email to majordomo@... with no subject and
          > "unsubscribe ai-geostats" in the message body.
          > DO NOT SEND Subscribe/Unsubscribe requests to the list!
          >

          --
          *To post a message to the list, send it to ai-geostats@....
          *As a general service to list users, please remember to post a summary
          of any useful responses to your questions.
          *To unsubscribe, send email to majordomo@... with no subject and
          "unsubscribe ai-geostats" in the message body.
          DO NOT SEND Subscribe/Unsubscribe requests to the list!
        • rplbrian@co.washington.wi.us
          Richard, You have hit upon something which has troubled me for sometime.... How do you communicate the reality of the data operating behind a map to the
          Message 4 of 5 , Jun 5, 2000
            Richard,

            You have hit upon something which has troubled me for sometime.... How do you
            communicate the reality of the data operating behind a map to the user/viewer of
            the map? I regularly suggest to users of maps produced by my staff to read a
            Mark Monmonier's "How to Lie with Maps". For the general user of cartographic
            products, it clearly summarizes much of what I try to convey to them in
            conversations.

            Brian Braithwaite
            Washington County
            Real Property Lister/GIS Manager





            --
            *To post a message to the list, send it to ai-geostats@....
            *As a general service to list users, please remember to post a summary
            of any useful responses to your questions.
            *To unsubscribe, send email to majordomo@... with no subject and
            "unsubscribe ai-geostats" in the message body.
            DO NOT SEND Subscribe/Unsubscribe requests to the list!
          Your message has been successfully submitted and would be delivered to recipients shortly.