Data collection - RDFBones/FrNonMetricTraits GitHub Wiki

During the data collection phase, non-metric traits are recorded on the material that the researcher had selected before and the results are stored as categorical measurement data of the type PresenceDatum. Although FrNonMetricTraits comprises 14 different cranial and mandibular non-metric traits, not all of them must be examined during an investigation - the completeness or taphonomic state of a skull may impede some of them.

/Graphs_FrNonMetricTraits/FrNonMetricTraits_DataCollection.JPG

The data collection process basically starts with a specimen which was selected by the researcher during the specimen collection process. Due to restrictions on the 'has specified input' property, only a Specimen.IncaBone can be input of Assay.IncaBone. This ensures that the measurements are always performed on suitable material.

According to the OBI ontology, an assay is:

A planned process with the objective to produce information about the material entity that is the evaluant, by physically examining it or its proxies.

In case of FrNonMetricTraits, the documentation of non-metric traits is regarded as assay and the evaluant is always an instance of the respective specimen class. Thus, FrNonMetricTraits has in total 14 assays. Each assay has always exactly one specimen as input and exactly one presence datum as output. The assay specifications contain definitions of the relevant non-metric traits. See study design for further information.

Both the assays and specimens of FrNonMetricTraits are linked to NonMetricTrait, an instance of the class 'Evaluant role'. Basically, the evaluant role defines the aim or purpose of an investigation. This purpose could be realized by another investigation as well. For example, another extension which uses a different set of non-metric traits would also realize NonMetricTrait. Thus, the evaluant roles can be used to classify investigations by querying the ontology.

The collected data are stored as instances of subclasses of PresenceDatum, which are all subclasses of 'Categorical measurement datum'. Each instance of PresenceDatum has exactly one LabelForPresence: either 'not observable' or absent or present. It is the LabelForPresence which states if a non-metric trait, for example an Inca Bone, was observable and if so, if an Inca Bone is present or absent. Usually, 'not observable' may not be used since all material which is not suitable for the relevant assay should be sorted out during the specimen collection process. 'not observable' was included in FrNonMetricTraits since it is part of the data entry form used at Biological Anthropology Freiburg.

In case of some non-metric traits (see Anthropological background), the trait may be present in different characteristics. For example, you may observe a complete, partite or partial Inca Bone. Instead of creating an assay class for each of these possibilities, there is just Assay.IncaBone but the researcher must add (in addition to the LabelForPresence) an integer via the 'has characteristic' property to specify which characteristic is present. Hence, in case of a partial Inca Bone, you add the label present and the integer 2 to the PresenceDatum.IncaBone. What integer codes for which characteristic is explained in the class description of PresenceDatum.IncaBone, for instance.