Comments on: Popular predators

By: Conrad Barwa

Conrad Barwa — Sat, 07 Mar 2009 15:31:26 +0000

Fair enough Nitin, but that is a one sentence qualification right at the end. The title of the original post and the subtitle, editorialise that you endorse the findings of this survey. At least one aspect, that of civilian deaths caused by aerial attacks, I find them hard to accept. Other investigators who have looked at this have concluded that aerial attacks are responsbible for a large number of civilians casualties, over the last 3 years in Afghanisatan they have accounted for about 50% of total civilian deaths caused by NATO forces:

http://www.hinduonnet.com/fline/fl2521/stories/20081024252100400.htm

I am not saying that the civilians in the NWFP and FTA support the Taliban or like their presence but to say that predator attacks that frequently kill civilians is “popular” amongst them seems to be to be inaccurate and morally dubious.

By: Nitin

Nitin — Sat, 07 Mar 2009 02:14:00 +0000

Conrad,

I don’t know anything about AIRRA or Farhat Taj. Hence the conditional wording in the last paragraph.

Udayan has a point: the survey must be seen in the face of purely unsubstantiated claims passed off as conventional wisdom by Pakistani politicians.

By: Conrad Barwa

Conrad Barwa — Fri, 06 Mar 2009 11:27:18 +0000

@ Udayan
>>As compared to what?

Sorry, I am not sure what you are referring to here.

>>We either have to believe the Gilani dude based on the evidence comprising of different types of handwaving. Or imperfect social science.

Eh, we don’t have to believe Gilani at all; what I am saying is that the study is too flawed to draw any firm conclusions from, which is what Nitin’s post seeks to do. Social science is always imperfect – that is why it is regarded as distinct from the other sciences, this should not be an excuse for sloppy work. The methodology of this paper is very weak and doesn’t even include a margin for error. Questionnaire surveys are notoriously difficult to design and can be easily skewed to show a bias. I haven’t seen the work in question but before one accepts its findings which fly in the face of the rest of the available evidence and consensus something stronger than this is needed.

Anybody involved in social science research would tell you that many of the potential problems can be limited or reduced to make it much less imperfect. Much more so than in this case.

>>Not sure about you, but I will trust the Farhat dude more on this, even considering that the AIRRA might be financed by Mr Langley

I simply said that I find Farhat Taj’s figure for the respondents he claims to have contacted and come across as an ethnographic researcher. I am happy to stand by this claim and I don’t think anyone who has done this kind of research will disagree with me. You are quite entitled to believe Farhat Taj, I am just pointing out that I don’t think the evidence provided supports this.

I have not made any assertion about AIRRA being financed by an interested party or being used as a covert propaganda outfit – personally I am more than willing to take their assertions as an independent party at face value. My sole disagreement comes not from any imagined political agenda they have but their methodology and evidence.

By: Udayan

Udayan — Fri, 06 Mar 2009 10:48:47 +0000

@Conrad,

As compared to what?

We either have to believe the Gilani dude based on the evidence comprising of different types of handwaving. Or imperfect social science.

Not sure about you, but I will trust the Farhat dude more on this, even considering that the AIRRA might be financed by Mr Langley.

By: Keshav

Keshav — Thu, 05 Mar 2009 20:35:23 +0000

So far, all the Google results about the AIRRA lead to a guy named Khadim Hussein who apparently does most of the writing for the organization.

Of all the sites, I saw his message was consistent across news sources, even Pakistani ones. His research might be flawed but he seems sincere.

By: Conrad Barwa

Conrad Barwa — Thu, 05 Mar 2009 19:36:27 +0000

Interesting survey but I have to say the results should not be treated with much confidence. NWFP and FTA have turned into de facto warzones so the people you meet and the answers they give need to be carefully sifted before we jump to any conclusions. About 20% of those surveyed refused to give any answers and this itself should speak volumes. Existing data that have been collected on predator strikes both inside Pakistan and in Afghanistan indicate that “collateral damage” is high and that American intelligence on the ground is incredibly poor. Given that these are the same people that launched a rappel assault by Army Rangers on a UN building in Mogadishu beleiving it to be Mohammed Aideed’s hideout and botched the attempt to capture Al-Qaeda militants in Tora Bora instead being handed some patsies who were shipped off to Guantanamo Bay while the real terrorists were airlifted out to Pakistan, I am not entirely surprised. US troops have been critcised by their more professional peers such as the British and two British Army chiefs of staff have recorded the poor intelligence of American troops on the ground as a major operational obstacle.

Also as someone who has done ethnograpic research for more than a year in India – outside a conflict zone, I find it difficult to beleive that Farhat Taj can claim to have access to over 2000 respondents. This is an incredible figure. Ethnographic research is very hard to do, because you have to earn people’s trust and get them to tell you what they really think and not what they simply beleive you want to hear. Having 10% of that number would be a tall task, the figure cited, simply strikes me as fantastical.

Citing this study as proof of anything, is an unwise and hasty move imo.

By: Corporate Serf

Corporate Serf — Thu, 05 Mar 2009 19:16:13 +0000

How reliable are the AIRRA? If they are perceived as “city slickers” the responses might be skewed. If they are perceived as seeking a very specific answer, perhaps favourable to a powerful foreign power, the responses might actually reflect that.

Also, is there a selection bias in people who agreed to speak to the organization volunteers? (I think this might be true, the question is, how big is the effect)

The The News piece you link to is short on the details of the study (perhaps there is a scholarly paper in the works, but my experience of the quality of papers in this sort of area leave me cynical); and the author is actually a member of AIRRA conducting the study.

I don’t doubt the sincerity of the study, but biases could tip the final decision into the other camp, given the close number in most categories.