Hi Ricardo,
Yes. It is possible to create generalisable theory without needing massive datasets. Generalisation comes mainly from sound reasoning rather than data. In fact, in many cases in which there are large data sets, the primary problem is that the useful data are hidden by the rest.
Part of my background is in systems dynamics modelling in which the process is to create an accurate predictive causal theory model initially from observation of structural relationships between factors and then the use of minimal data to calibrate the model to be an effective causal theory representation. Good prediction comes from good study design rather than massive data.
Other examples in the design field of developing generalisable theory without using massive datasets are the work of Neilsen and Norman reported in Alertbox, and research by Professor David Sless.
The main use massive data (in the sense that Ken comments) is to go fishing to try to identify causal explanations in areas about which very little is known. That is appropriate to some situations but not all.
An alternative use of massive data is to create AI models via learning, e.g via neural nets. This is not obviously the making of generalisable theory.
Another use of massive datasets is for data mining of, for example, sales data to identify most appropriate business decisions. In most cases as far as I understand this also doesn't produce generalisable theory.
Ken and I have had this discussion on and offline previously - we disagree on it.
Cheers,
Terry
==
Dr Terence Love
MICA, PMACM, MAISA, FDRS, AMIMechE
Director
Design Out Crime & CPTED Centre
Perth, Western Australia
[log in to unmask]
www.designoutcrime.org
+61 (0)4 3497 5848
==
ORCID 0000-0002-2436-7566
-----Original Message-----
From: [log in to unmask] [mailto:[log in to unmask]] On Behalf Of Ricardo Martins
Sent: Tuesday, 16 January 2018 9:29 PM
To: [log in to unmask]
Cc: Ricardo Martins <[log in to unmask]>
Subject: Re: Expanding the discussion about statistics and design
Thanks Terence.
I guess the original meaning Ken intended for was about the generalization of results. Do you think it is possible to generalize without replicating the first study or without massive data?
Cheers,
Ricardo
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]> Discussion of PhD studies and related research in Design Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]>
Discussion of PhD studies and related research in Design
Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
|