[Corpora-List] Corpora Digest, Vol 52, Issue 3 [was blind reviewing]

[Alon Lischinsky]

On 2011/10/3 Laurence Anthony <anthony0122 at gmail.com> wrote:

> As you say,  "it's not easy to decide whether [a blind practice] is
> preferable".  However, in these cases, don't we normally look to the
> lessons we get from history? Throughout history, society has
> implemented various methods to avoid biases in decision making
> systems, and 'blind' practices seem to be much more prevalent now.

In some respects, probably. For most of history, blindness was simply
not practicable because of the small world phenomenon.

However, it seems to me that blind procedures tend to be favoured to
protect individuals in situations where no single one exerts
significant power, as in your example of voting systems. In this case,
individual-level bias is of little consequence. The goal of the blind
procedure is mainly to protect individuals from ex-post retaliation.

However, where individual decision-makers have significant power and
their clients are not usually in a position to retaliate (as is the
case with reviewers and authors), transparency and public oversight
can offer more substantial advantages than blinding. You mention peer
juries as an example; they are a somewhat peculiar one, inasmuch as
jury decisions are supposed to be unanimous, and hung juries result in
mistrial, but more important is that they don't much resemble peer
reviews. The reviewing process is, if anything, more similar to a
bench trial, where a single person is in charge of all decisions, and
no checks exist to limit their discretionary judgement. And in all
forms of bench trial I know, magistrates are required to identify
themselves, provisions are made for their eventual recusal, and
tallies of the votes in collegiate bodies (i.e., SCOTUS) identify by
name who subscribed to each opinion.

I don't think blind review is always a bad idea; far from it, evidence
such as that quoted by Adam suggests it can be quite beneficial in
many cases. However, it also has disadvantages, and I don't think they
should be ignored just because it's what everyone else is doing.

A.

_______________________________________________
UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora