Modelling is great purely as exploratory investigation / iterative design / understanding how to satisfy constraints etc, but for those cases where it makes sense to preregister some principles or procedures you'll follow when fitting a model, does anyone have a nice example?
-
-
Replying to @VandekerckhoveJ
Thanks Joachim. Another perspective on this is that for those who have described the model in a previous publication and now describe an iteration, that's a kind of preregistration, in that people will look askance if one makes certain kinds of changes. In my field, I see that
1 reply 0 retweets 0 likes -
Replying to @ceptional @VandekerckhoveJ
a lot with
@bradpwyble's models, and I wonder what has been his experience e.g. in the review process. It still seems it would be good to try to write down what aspects trying to hold fixed before embarking on a new modeling exercise, but easy for me to say, I'm not a modeller.2 replies 0 retweets 0 likes -
Replying to @ceptional @VandekerckhoveJ
As for holding things fixed before embarking on a modelling exercise, the case for this is clear when doing model-fitting (to limit overfitting), but when making a new model, which is a fundamentally creative process, I don't see the value of locking in ideas. Can you explain?
1 reply 0 retweets 1 like -
Replying to @bradpwyble @VandekerckhoveJ
With a model you've previously published that you now want to apply to the results of new experiments, I'm thinking that before modifying the model, you might want to declare which parts of the model are core and should remain unchanged and which sorts of processes you consider
1 reply 0 retweets 2 likes -
not core to the model and fair game to modify or extend in various ways so that the model can deal with the new task used in an experiment. In other words there might be parts of the model that aren't parametrized but you consider variable in that they're specific to e.g task.
1 reply 0 retweets 1 like
Yes, I think in modelling this is standard practise. This is how I was trained to do modelling and how I have seen others train and be trained.
-
-
Yes agreed. This is often what is meant by the distinction of free vs. fixed parameters.
0 replies 0 retweets 1 likeThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.