Dan Goodman on Nostr: Journal impact factor is a terrible measure, but I do understand that when reading ...
Journal impact factor is a terrible measure, but I do understand that when reading applications it's a useful shorthand way of giving points, but I'm wondering what is it we're really looking for? I'm not interested in citations but I am interested in work done well.
Is there a low effort way to get a signal of this? I'm wondering about some sort of trust model where you can say if I trust the work of X and they trust Y then I trust Y (but less than X). That way, we get a heterogeneous and less exploitable signal, but many potential issues.
For example, you would need trust relationships to be stored either publicly (which could be socially problematic) or put your trust in some service (which creates dependency). And it might lead to various biases (gender, culture, etc).
Is there a low effort way to get a signal of this? I'm wondering about some sort of trust model where you can say if I trust the work of X and they trust Y then I trust Y (but less than X). That way, we get a heterogeneous and less exploitable signal, but many potential issues.
For example, you would need trust relationships to be stored either publicly (which could be socially problematic) or put your trust in some service (which creates dependency). And it might lead to various biases (gender, culture, etc).