Process Data

[ben-domingue]

The plan is to have a process data standard that is fairly straightforward. A 'process dataset' will effectively be two tables:

• Table 1: a standard IRW dataset [where the rows represent item responses]
• Table 2: a 'process' dataset where the rows represent process observations (so there may be multiple rows/response). These two tables will be joined by id,item. The process tables will live in a separate dataset on redivis; we'll add documentation to the website on this matter.

[ben-domingue]

The plan will be to:

1. start with one example dataset
2. once we are happy with the standard, add some initial documentation to the website
3. process the rest of the pisa data to this standard.
4. continue work on website modifications to describe process data and formatting of novel process datasets.

[ben-domingue]

For 1, I will share the data directly.