[Babase] Re: Predation Coding Scheme

Lacey Maryott lacey.maryott at duke.edu
Tue Mar 11 09:12:31 EDT 2008



Susan Alberts wrote:
> Thanks karl, see below,
>
>> On 03/10/2008 10:41:41 PM, Susan Alberts wrote:
>>>> Regards data entry error.  Don't forget the "have two different
>>>> people enter the data and then compare" method of error detection.
>>
>>> And, of course, we will have two people proofing, with a switch in 
>>> roles to make sure that the enterer is not re-duplicating her 
>>> original errors.
>>
>> The dual-entry/compare can _be_ the proofing.
>
> Yes, when we do dual entry, that substitutes for proofing for us. But 
> in the case of transcribing hand-written, prose notes into coded, 
> digitizable format, i think that dual entry may be prohibitive and 
> proofing much faster and less labor intensive.  We will continue to 
> think about it and see how dual entry goes for a small part of it,, as 
> I mentioned earlier. I like your idea of comparing that with proofing 
> to see which is more time consuming. Lacey, let us know your thoughts.
>
> Susan
>
I am definitely open to trying both and seeing which is the most 
time-effective.  Sorry I wasn't able to respond from home, my internet 
at home is down right now :-\... working on it.

Lacey
>>   It all depends on how
>> through you want to be.  I don't have extensive experience, but
>> the 2nd entry can, depending, be as quick an operation as the
>> proofing.  Then you let the computer do the actual checking
>> during the compare step.  The thing about having a human
>> do the proofing is that seeing work done can pre-dispose
>> the proofer to interpret the original data, or just the
>> transcription, in an erroneous fashion.  Not necessarily
>> in their own data entry work (if the proofer/enterers are swapping
>> roles), but in the actual proofing itself.  The nice thing
>> about dual-entry is that both data entry people start from
>> a "clean slate" and have no pre-conceptions about what the
>> data should look like.
>>
>> You might want to benchmark the various approaches (comparing time,
>> error rates, with a control without proofing or double-entry,
>> statistically validating the results, publishing a paper on
>> "Data Entry Approaches to Recording Primate Predation Data",
>> etc.  ;-)
>>
>> Karl <kop at meme.com>
>> Free Software:  "You don't pay back, you pay forward."
>>                  -- Robert A. Heinlein
>>
>> _______________________________________________
>> Babase mailing list
>> Babase at www.eco.princeton.edu
>> http://www.eco.princeton.edu/mailman/listinfo/babase
>
>

-- 
Lacey Maryott
Alberts Lab
Department of Biology
Duke University
ph: 919-660-7306
fax: 919-660-7293
Lacey.Maryott at duke.edu 



More information about the Babase mailing list