danalwyn: (Default)
From VOX, via Chris Blattman, a conclusion that will surprise nobody who has done much work in Academia (or probably anywhere else for that matter), but is interesting in quantifying what we all knew.

The prior research on corporate boards of directors suggests there is a critical mass of female directors on the board which is necessary before fundamental changes occur in board operations. Similarly, we test whether this was true in academia and find that a critical share of the board of trustees of an academic institution of 25% must be reached before the gender composition of the board influences the speed with which an institution diversifies its faculty across gender lines.

Some other interesting statistics in the cited papers, which have a fairly large sample size.
danalwyn: (Default)
Dear Post-Modern Feminists,

We keep hearing about this wonderful thing referred to as the Feminist Approach to Physics. We were hoping that you could clarify what it is and what it consists of.

We have looked at several sources, but always found the principals to be outlined in vague, unquantified, and normally specious language. There are several arguments for an "integrated, multi-community, intuition-based" approach, without any accompanying description of what this approach would look like, or how it would affect the construction of an experiment. The fundamental point that seems to be raised, to identify the "biases" of a researcher seems to either be a product of an improper reading of a presentation, or a lack of understanding of the process itself. If this refers to experimental biases, they are clearly listed in the paper references, or available upon request. Any failure of you to reproduce these results should result in a counter-publication, and a great deal of dialogue. However, if you are referring to the ideological biases of the researcher, these should not enter. According to the framework of Science, any person from any ideological background should be able to reproduce the same results at any time. Failure by any party should result in the invalidation of the theory, after an exercise of due process (collaborating results may be required).

Those specifics presented mostly deal with either the presentation of history, or the field of education, both of which are in need of overhaul, but have very little influence over the current conduct of experiment, which seems to be the target. Some of the experimental suggestions seems confusing, such as the term "multi-causal", apparently a great favorite, but one that is ill-defined at best. What does it mean for an experimental measurement to be "multi-causal"? It is already the culmination of a great many effects - all of which have to be measured independently. We cannot determine the meaning of this statement. Other suggestions, to replace "unnatural and invasive" particle experiments with more passive observation have failed to produce any experimental design that would be of current use. Most importantly, it is unclear about how accounting for social biases will work in with experimental designs. Do we use only chips made in certain countries? Considering that we only report the results of very specific results, it seems hard to do anything else.

Maybe the theorists know what you're talking about, but the experimentalists are confused. Could you please provide a properly formatted paper on new experimental techniques and statistical theorems, alone with a series of examples of properly done experiments?

We will be eagerly awaiting your reply,


danalwyn: (Default)

August 2017

2728 293031  


RSS Atom

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags