Need education about the inheritance principle for JSON meta-data

[Peder Isager]

Hi everyone, 

I am currently working with collaborators on implementing Psych-DS for a rather complicated data set consisting of a large number of tables curated from published meta-analyses. There is no one primary .tsv data file, although users could aggregate data from all tables based on their variables of interest. For this data set, we crucially rely on generating a functional meta-data structure in order to facilitate reuse once the data are shared. Issues include: 

• Every table needs to be connected to unique bibliometric information about the paper the table was extracted from.
• We will need an overarching code book for variables shared between several tables, as well as a unique code book for every table that contains the variables unique to that table.
• Variable descriptions for the same variable may some times change between the code books of different tables (e.g. two tables may contain the variable "age", but one simply codes "old" vs "young" while the other gives the number of years exactly).

We have run into the problem that none of us are familiar enough with JSON meta-data to understand how to best construct the various meta-data files, and to ensure that the inheritance principle is adhered to correctly throughout the data set. 

Firstly, does anyone have any suggestions for educational materials (books, online courses, etc.) that would help us gain a deeper understanding of the JSON meta-data curation process? 

Secondly, would it perhaps be useful to write a tutorial document on the meta-data curation process in Psych-DS? This could include details on how to adhere to the inheritance principle, and an explanation the function of well-curated meta-data. The latter is not obvious to most researchers at the moment I think, and a clarification of the benefits of structured meta-data would likely help motivate new users to populate their meta data thoroughly. 

All the best,

Peder M. Isager