Skip to end of metadata
Go to start of metadata

First introduced 11/8/2018 by John Faundeen:


I was asked the following question:

"Just asking a metadata content question, about the Process Step specifically. It's usually a looong block of text in the metadata. If possible, we would like to replace that text block with a DOI link to a journal article that has the same information and more detail."

My first inclination is no, the actual metadata needs to be present, but thought maybe y'all have some comments on this one.

  • No labels


  1. From Matt Cannister: I've seen situations where the author put a short statement into the process steps field, such as "Samples processed and recorded based on Smith (1998)." And then they included the Smith 1998 citation as a cross reference in the metadata. 

    In my opinion all the information related to the product should be self contained in the product's own metadata, but that might not always be reasonable or possible.

  2. From Kim Kloecker: I think keeping the information in the metadata record rather than an external link to follow makes it more likely it will be looked at by data users. Why would the journal article have more detail? Couldn't the metadata author copy/paste and include it all? Does it matter that it is a "looong block of text"? That's an interesting one- usually I see the journal article having less detail (page limits) and the metadata more room for details.

  3. From VeeAnn Cross: My suggestion would be to have either a cross-reference for the journal article. 
    Summarize the information in the process step, and then indicate for more information Smith and others, XXXX (cross-reference). 

    Also, for long blocks of text, formatting can help. A blank line or a > on a line by itself will break things up so it's not a run-on paragraph.

  4. From Janelda Biagas:  I'm in favor of writing the metadata so that the form does the heavy lifting of communicating the process steps succinctly.  That includes a cross-reference to the journal article as well.  From the end-users perspective, if I were interested in more information, I'd follow through with reading the journal.

    Also for consideration, are there situations where the data will be released without an associated journal article?

  5. From Ray Obuch: ISO supports embedded dictionary and other reference info. Links are also used , however they introduce another level of maintenance keeping links and linked info current.

  6. In the past we would a Readme.txt file in the product distribution which wasn't reference in metadata etc. it was a general overview of the product.

    There may be other data sets not directly reference or used in a publication and are standalone. I think there needs to be adequate description of the process that may 

    require (more that not) additional documentation that may be better handled in a Readme file or another external note.