Rantlet: Literature Review

Apologies if you're seeing a triple posting; it seems like blogger is misbehaving. They really want me to switch over entirely to LJ, don't they?

I am currently slogging through the literature search section of my thesis, and there's a chunk that is so frustrating that I have to rant about it.

I'm covering some unpublished research by a scientist at a school in the Midwest; the experiments themselves are basically good--they are on internally insulated basement walls, which is the topic of my thesis. It is a large body of research with multiple experiments, so it can't be ignored. Some of the experimental protocols are a bit annoying and show inadequate understanding of what is physically occurring.

But moving into results: the author did not bother to cull data into specific plots which show relevant points. Instead, she dumps complete sets of data into close-to-unreadable graphs--I guess if its unpublished, you don't need to do that step of filtering it down the six or eight graphs that actually back up your conclusions. It would have beeen far more useful, actually, if she had just dumped raw data as CSVs on the web.

The most painful part, however, is the interpretation of the data. Although this person is obviously smart, she has no background in the building science field, otherwise she would not be trying to explain things in terms that she is making up. Worse, some of her science is just plain, dead, obviously wrong; there are phenomena that have long been known in the literature that she seems clueless about. Some of her explanations feel as jarring and antiquated as going to a physics or combustion engineering conference, and hearing results being explained in terms of the ether or phlogiston. Agh.

I think this is a clear problem with non-published/peer reviewed papers.

What makes this all painful is that there are some useful results to cull out of the information she is presenting, but it takes some serious digging to tease out the useful bits (Okay, which way did that wall panel face? Dig dig dig. Okay, that makes sense now. What was the outdoor temperature then? Hrm... that's not in the report... weatherunderground.com search, dig dig dig.)

One thing everyone should find scary is that person is working on that state's building code recommendations. She is actually responsible, it seems, for an insulation detail that has failed all over the state. I wonder if people will connect the dots soon.

I'm trying to be a bit restrained in my criticism while reviewing this literature, but it's pretty friggin' hard. I do have some worries that this might turn into a pissing match if she reads my work, even though I'm confident that I'm right, and I would have the backing of just about everyone I respect in the field.

Well, back to writing.