Reprint count period

[CarolinOdebrecht]

With regard to the publication periods in which we count the reprints, something was particularly noticeable in the case of the German collection. The sampling document says: period is 1980-2000. Especially in Germany with the reunification the start of the period is somewhat difficult.

I would to discuss whether we would like to extend this period maybe from 1950-2020. With 1950 as a start, we would have a generation (30 years) between the start of reprint count and the end of the periods of ELTeC text collection. The extend the end or the reprint period would make it possible to draw a more accurate picture of the canonicality of German texts.

[CarolinOdebrecht]

We need to set a final sampling parameter until 2019-02.23. The possible adjustments should be compatible with the existing text selection, which does not have to be started from scratch.
So far, we have been working on the basis of the following criteria:
We propose to use the number of times a work is reprinted as an
objective measure of its reception from 1980-2000, using categories like the following:

• low: no reprints at all, (30 % of the novels in a language collection)
• medium: reprinted once,
• high: reprinted more than once. (30 % of the novels in a language collection)

Here are just three basic scenarios, (some of the could be combined, were already discussed):

1. Extend the reprint count period (1950-2020/2019) in order to be more flexible. This will reflect a broad modern perspective on reprints.
2. Adjust the balancing range, e.g. we could reduce the required minimum from 30 % to e.g. 10 %.
3. Remove the criterion.

This is now open for disussion. What do we want to map with this criterion? How can we make this criterion operational in a meaningful way?

(This issue is related to issue #5 )