1 Topic by Bob Merritt

  • Bob Merritt
  • Veteran
  • Offline
  • From: Dumfries
  • Registered: 11-12-2006
  • Posts: 94

Topic: ID verification categories: a problem with R3 import addin

Recorder 3 has three categories of species ID verification, namely:  “OK”, “needs confirmation” (relates to records for which the original determination is questioned for some reason) and “known incorrect”.  Recorder 6 also has three categories, but they do not correspond exactly with those of Recorder 3, hence the problem when data are imported.  Clearly “OK” is equivalent to “passed verification” in R6 and “known incorrect” is equivalent to “failed verification”.  It seems reasonable to assume that “needs confirmation” is equivalent to “pending verification”, but that is not the case in reality. There is an additional category in R6, namely “not verified”, which refers to instances when no attempt has yet been made to identify the species.

In R6, when data are imported using the R3 import addin, IDs that were flagged with “needs confirmation” in R3 appear incorrectly as “not verified”.  One consequence of this is that they are not shown when you select Data Entry – Display Unchecked Records – Failed/pending verification; only records which failed verification (i.e. “known incorrect” in R3) are listed.

2 Reply by RobLarge (edited by RobLarge 28-02-2013 11:45:06)

  • RobLarge
  • RobLarge
  • Veteran
  • Offline
  • From: Wiltshire
  • Registered: 15-07-2008
  • Posts: 812

Re: ID verification categories: a problem with R3 import addin

I think the R6 verification flag is not set directly by the import from R3. There is an intermediate step, the determination type, visble on the determinations tab of an occurrence. This has six possible values:Correct, Considered correct, Considered incorrect, Incorrect, Requires confirmation, Unconfirmed.

It is my understanding that the imported value from R3 is used to select one of these six values, which in turn are used to set the verification flag as follows: The first two determination types equate to Passed verification, the next three to Failed/pending verification and the final one to Not verified.

This represents an improvement on the R3 system as it enables finer descrimination of the determination type.

It is also incorrect that unverified items do not appear when Display Unchecked is not set. The Checking this refers to is the tickbox which appears to the left of the species-name in the observation hierarchy. This has no effect on determination, it merely indicates that the record has been checked to see that it has been entered correctly. If I remember right, the Checked tickbox is left unchecked for records entered manually, via a recording card or similar, while for imported data it is checked by default. I am not sure how imports from R3 are handled with regard to this flag.

All this is explained (probably better than I can) in the R6 help

Rob Large
Wildlife Sites Officer
Wiltshire & Swindon Biological Records Centre

RobLarge's Website

3 Reply by Bob Merritt (edited by Bob Merritt 28-02-2013 15:05:53)

  • Bob Merritt
  • Veteran
  • Offline
  • From: Dumfries
  • Registered: 11-12-2006
  • Posts: 94

Re: ID verification categories: a problem with R3 import addin

Thanks, Rob, for your explanation.  The problem would be solved, therefore, if the imported value from R3, in this case the one for "needs confirmation", was set to select the "requires confirmation" value from the determination type options available in the determinations tab of an occurrence. Alternatively, one could manually reset the determination type as has been suggested helpfully to me by Charlie Barnes.  It would be better all round if the problem could be fixed at source in the R3 import addin - could this be done please, Sally and Mike?

4 Reply by RobLarge

  • RobLarge
  • RobLarge
  • Veteran
  • Offline
  • From: Wiltshire
  • Registered: 15-07-2008
  • Posts: 812

Re: ID verification categories: a problem with R3 import addin

What determination type has it selected for the ones you have as 'needs confirmation'? It is so long since I did the R3 import, I cannot rememeber what happened to ours (or even if we had any in that category).

Rob Large
Wildlife Sites Officer
Wiltshire & Swindon Biological Records Centre

RobLarge's Website

5 Reply by Bob Merritt (edited by Bob Merritt 28-02-2013 15:24:47)

  • Bob Merritt
  • Veteran
  • Offline
  • From: Dumfries
  • Registered: 11-12-2006
  • Posts: 94

Re: ID verification categories: a problem with R3 import addin

Rob, I have checked four entries and all have a determination type of "observation"

6 Reply by RobLarge

  • RobLarge
  • RobLarge
  • Veteran
  • Offline
  • From: Wiltshire
  • Registered: 15-07-2008
  • Posts: 812

Re: ID verification categories: a problem with R3 import addin

Bob, at the moment I don't have time to check, but I seem to recall this happening with our import and we probably corrected the problem, by a batch update, or some other means.

If you are able you should try to check whether, a) all of your needs confirmation records have imported as Observation and b) any of the other types have also imported as Observation.

I am not sure, but I imagine if one Needs confirmation record has gone in as observation, they probably all have, so b is probably the more important one.

I am intending to run some queries on our db to check out the determination types sometime in the near future, so your post is timely.

Rob Large
Wildlife Sites Officer
Wiltshire & Swindon Biological Records Centre

RobLarge's Website

7 Reply by Bob Merritt

  • Bob Merritt
  • Veteran
  • Offline
  • From: Dumfries
  • Registered: 11-12-2006
  • Posts: 94

Re: ID verification categories: a problem with R3 import addin

Rob, the imported dataset in question has 63374 records of which 1 record was 'known incorrect' (R3-speak), 30 were 'needs confirmation' and the remainder were 'OK'.  I have run a report in R6 on the dataset and returned 63372 records (the missing two are the one 'known incorrect' record and one of the 'needs confirmation' records which I had edited, changing the det type from 'observation' to 'requires confirmation').  Of the reported records, some 58k have a det type of 'observation' and the remainder have a det type of 'validation'.  The common factor is that records which have the same 'sample recorder' and 'obs determiner' all have a det type of 'observation' and all records for which they are different have the det type of 'validation'.

8 Reply by sallyrankin

  • sallyrankin
  • Administrator
  • Offline
  • Registered: 07-04-2006
  • Posts: 371

Re: ID verification categories: a problem with R3 import addin

This is another JNCC add-in to which we have no access to the source code so we can’t modify it. As far as I can recall, it was not modified to handle the changes to the way verification is handled in v6.13 of Recorder 6 hence it doesn’t transfer this field correctly. When Mike Weideli and I did Recorder 3 to Recorder 6 data transfers I used to extract this data from Recorder 3 and Mike used the extract to correct the Recorder 6 database. This was made possible by the key translate table that Mike used to create for the transfers we did. Unfortunately users doing their own transfers will need to correct the data in Recorder 6 manually. As Rob says the setting of the Verified flag is explained in the Recorder 6 Help – search for verification and select Verification of identifications. Also, Data Entry – Display Unchecked records can be used to filter for Unchecked, Zero abundance and Failed/pending verification observations.

Sally Rankin, JNCC Recorder Approved Expert
E-mail: s.rankin@btinternet.com
Telephone: 01491 578633
Mobile: 07941 207687