Back in 1984 when I started in this field “RDBMS” I suspect didn’t appear in our MH/WB/therapy literature. However, things have changed and collection of large and complex datasets is not uncommon in these fields now. A RDBMS is just (that’s a very big “just”) software to manage complex databases in which data is not simply “rectangular”, i.e. it’s not just a single data table in which all rows have the same columns.
How we best handle such data has become a huge challenge for our fields, one very much in development.
Details #
Most therapy service data now is complex. We might reasonably be interested in which practitioner saw which client, when and where and we might be interested in data for each of those “facets”:
* practitioner: gender, age, training(s), first qualified, first worked here …
* client: gender, age and so much more, some of it itself likely to fall into more complexities of how best to structure its storage
* when: pretty simple, date, start time, finish time but perhaps more or put it under session
* session: cancelled? (more on that), attended, started when, if not on time, why?, finish time, type: individual (i.e. dyadic), couple (oops, may need to link to two or even more practitioners), family (ditto), group (oops, needs to link to more clients and perhaps more than one practitioner) and much more detail possible
* where: clinic, which?, client’s home, elsewhere? (what about walking/nature therapy?)
* episode: apart from single session therapy, most therapy is structured in episodes of a sequence of sessions, make sure these are linked and this too may have further information.
As you can see from that list, the data simply isn’t rectangular, it’s multidimensional and involves multiple datatables. We can also see that we will need ways to cross-link the crucial bits, e.g. a session can link to more than one client and to more than one practitioner. This is where the “Relational” comes in: different tables of data are related, cross-linked by indices so a session record will have fields linking it to the practitioner(s) and the client(s) but also to the location.
Try also #
Dataset/database
R Project for Statistical Computing (“R”)
Statistical software
Tidyverse
Chapters #
Implicit in Chapter 8 about service level change management.
Online resources #
Not yet!
Dates #
First created 10.iv.24.
