Use the Analytic Hierarchy Process (AHP) to prioritize alternatives or determine which alternative best meets a specified goal.

Use AHP to perform a cost benefit analysis or to compare alternati…

First of all, I wouldn’t count astronomy or other observation “sciences” as hard sciences in the same way as I would count physics or chemistry, for the reasons cited above. I am not the only one to make this distinction.

Astronomy’s “accomplishments” are mainly discoveries of existing empirical facts in the universe. It is the science of physics that allows us to draw conclusions about the principles governing the things discovered, and thus propose theories about how things were in the past. A great deal of their theories depend on the idea of dark matter, which they can’t even really define except that it’s more or less what makes the whole system work. No problem with that, but it’s not the same thing as physics, and you can never be as sure as you are with physics.

Insofar as astronomy purports to “know” about the past, I say, this is silly. The theories are fascinating, but untestable, which is why this is not science, either, though science is used in astronomy. (I also don’t like geology: a bunch of memorization about rocks and then because one rock is this way, you are supposed to believe another rock is the same way 10 billion years ago, though they actually have no idea about the whole situation then… ridiculous).

And I wonder why someone would be called a creationist just because he or she doesn’t think that astronomy or geology are on the par of the hard sciences? I’ve heard this before, and I simply don’t get it. I’m not a creationist, just a someone who takes the reasonable view that we probably don’t know exactly how things happened 10 years ago in a single petri dish, MUCH LESS 10,000, 10 million, or 10 gazillion years ago outside of spacetime (no matter what God said about it, which anyway is not much, like one paragraph).

Secondly, what Heisenberg showed, fascinatingly, is that sometimes tiny particles appear to behave as if they are also paying attention, appearing to change attributes or disappear under observation. (In a “real” scientific experiment, by the way.)

And this notion, I think, can explain so much. It’s almost magical.

On another note….

Taiwan Independence! Tiananmen Square Massacre!

There, let’s see if CA remains unblocked…

The “scientific” requirement for experiments is a red herring. Astronomy is a scientific field with great accomplishments, yet has no experiments to its name. Instead, it horizontally applies the findings of physics, chemistry, and even biology to the experimentless stage of the cosmos.

It does this using the Method of Difference and Method of Similarity, which are common in the social sciences too.

Instead, the problem of the social sciences are that the subjects pay attention. Molecules, rabbits, and stars do not observe the function of a complex adaptive network, rapidly changing nearly every aspect of their behavior to ensure their individual survival. Humans do.

Error bars for methodologies and epistemologies (not the same thing) are of course nice, though I have rarely seen an error bar of positivism or naturalism in a “hard sciences” article. Indeed, including the latter would doubtless draw charges of “creationist!”

Regarding science, I do prefer the hard sciences. I think that science should restrict itself to what it does best, which is test hypotheses in a controlled environment. Even up to biology and psychology, or group psychology, this is to some extent possible, because to a fair degree, we can control the circumstances under which the experiements are being carried out. We can track variables. With economics, because what we measure is so limited, I also can accept ranked indices and highly quantitative analyses.

Historical / socialogical “sciences” are different. The reason I don’t like numbers here is because they give us a false sense of security in an extremely complex situation which is neither controlled, nor controllable, and the variables of which cannot all be tracked. In particular, the sort of index which assigns a particular number to the overall extent of failure of a country’s government / state apparatus, I think, is misleading. The choice of indicators is so subjective, the methodologies for collecting data and quantifying it so untested and prone to bias at all levels, that we would not expect to get a really good result. Here, qualitative analysis serves us much better.

Finally, I should mention that my oft-quoted statement about the vanity of assumptions referred to individuals such as myself and was not intended to be a denial of the possibility of such an index. I think that it’s extremely hard, but could be done well. In that case, however, I would expect to see a lot more written in the article about the methodologies used, and the limitations of such an interpretation (such as we see in real scientific papers). That would give me a lot more faith in the results produced.

You’re exactly right.

What you call “techniques,” I learned in Scope and Methods as “epistemologies” — different ways of knowing.

It looks like Elisabeth is an interpretivist-fundementalist. When she

There’s a reason why some of us wouldn’t attempt to make such a index, knowing how foolish and vain it would be to assume we could quantify chaos and anarchy on a global scale and make comparisons.

she is directly attacking, not just positivism, but science itself.

She has a good point – I’m a fan of looking at context, too — but when she dismisses objective model building she goes too far.

A rational person will select epistemologies based on how they help him achieve his goal, whether it be understanding a situation or enacting some policy. When one confuses an epistemology with the Truth, one blinds oneself to all of the world which extends beyond that epistemology.

PS: tdaxp and ZenPundit are blocked in China, but CA is available. Has the Kaplanesque trio secretly joined the Chinese Communist Party????

Data entry and quantitative data analysis are fun, I agree. But it’s the data collection that counts- with bad data, you can come up with anything you want, and sometimes just a few bad figures will skew all the results.

You are absolutely right. In no way am I saying that quantitative analysis outweighs qualitative, or good old historical analysis. I see models like the AHP as a path to more rigorous analysis, like a highly detailed laundry list that makes sure you didn’t forget anything. Many times such models will produce counter-intuitive results, which is great because then you have more things to investigate that you never thought of before.

There is no such thing as a perfect model (remember Heisenberg) and that’s why I call these types of MCDA “techniques.” By no means are they the be all and end all, but they are one powerful tool in the analyst’s toolbox.

I don’t doubt that they can do a lot of cool things with weighted ratings and matrices, but I do doubt that the people who are playing with the numbers know what they’re doing. I would ask:

- Were the data collecters involved in data analysis?
- Were questions asked about the highest and lowest scores?
- Which group of people was responsible for analyzing the bias in data collection (source, methodology), and was this incorporated into the final analysis?
- What was the methodology for collecting the data, and how did they decide on it? Was it modified throughout the process as unexpected information came up?

For me, it’s not the data analysis that is worrying. I am responsible for data collection- it’s a huge part of my job- and I can tell you that most people think that the analysis is tought. Analysis is fun and formulae. It’s gathering information, finding the right key informants, finding unbiased surveyors, transferring that information correctly (for example, were reports that “the war continues to simmer” counted the same as “the war has exploded across the border”, or “aid workers reported the chance of starvation among vulnerable groups” counted the same as “aid workers reported widespread starvation in the countryside”?

Once you get all the stuff quantified, you’re set. It’s the quantification of chaos and misery that makes me doubtful, especially since the more people try to put a number on it, the further they seem to be from the truth. Think Heisenberg: the closer you get to defining it, the less the thing to be defined seems clear.