ramble skepticism

White rhinos are down to a few just because irrational persuasions are effective. Everywhere. Something that might help? Identifying lies.

In case you didn't know 'there are only 3 northern white rhinos left on the planet'. There were 5 since I started counting. One died earlier this year. Another one, last month. Its closer relative, southern white rhino, is relatively safe as of this writing but you never know. Rhinos are ominously popular targets for poachers. Along with Bengal Tiger (skin) and Elephant (teeth), this animal (horns) is one of the first that comes to mind when I hear poaching.

Where exactly do the horns go? Having no direct marketing channel, illegal items are slightly harder to sell (in the usual sense). Although you can create a market out of thin air, most of the cases require external factors. How about a bit of human irrationality ?

One of the reasons why Rhinoceros hunting has recently gathered pace is due to a weird outburst of horn demands from Vietnam. Apparently, there was a guy who was cured from terminal cancer using rhinoceros horns and people weren't able to stop the tendency of 'inferring from a sample size of 1' to outgrow their rationality.

This brings up an important issue. Not that other reasons for dwindling wildlife are not so important, but because this specific reason is prevalent in every other area to an extent you would probably always underestimate.

1. Quacks, Advertisements and Lies

Recently, I got home, started watching TV, and one thing i noticed is that our situation, in India, isn't drastically different from Vietnam. I have seen traditional Chinese medicine usually getting the heat for promoting irrational cultural beliefs (sadly, many involving endangered species). Are we any better?

Natural remedies are inspiring. And to be very clear, I have a space of respect for them (I mean, why not?). But I never quite understand the point of view of many of their propagators here. When engaged in discussion, most of them try to sell one of the following conclusions:

  • Western medicine is trying to destroy our traditional wisdom
  • People in ancient times were much healthier
  • We have a long history with medicine

Honestly, I have stopped paying attention to statements which don't add anything in anyone's plate. So, I ask, tell me what you are proposing? And I get statements like this (not disclosing the source)

vegetarian food reduces domestic violence.

Now, it's not about the people who can cure homosexuality, the one who can tell we are going to die in future, the one who can guarantee 700% reduction in fat or the realists with a conservative 100%. It is about how we see these events and infer. I will be willing to believe the given statement if the person who is promoting this knowledge can answer following questions affirmatively.

2. Few Simple Questions

At school you were taught about chemicals in test tubes, equations to describe motion, and maybe something on photosynthesis—about which more later—but in all likelihood you were taught nothing about death, risk, statistics, and the science of what will kill or cure you.

Ben Goldacre - Bad Science (2008)

Here are a few simple questions (non-exhaustive, please email me to add/remove things) I believe we should be asking to judge the conclusions of a research. Since researches on medicine and likes involve comparison and statistical testings, these are not exactly valid for any research.

2.1. What do the numbers mean?

No, not what you are being told. What the numbers actually mean? Here is a case.

In a test on N (N/2, N/2) people, product A performed 50% better than product B (assume an effect).

What is 50% here?

  • Is there a performance metric which bumps 50% more for A guys (first N/2, say) than for B guys (next N/2)?
  • Is it a mean (max, median) of that metric for groups?
  • Or, is it slightly quirky like "it worked for 50% more guys"?

2.2. Can I reproduce the numbers?

Good research is reproducible research. Given an array of results, access to research documentations and methodologies, a person should be able to get to the same conclusions as published.

Once done, there is a more critical question. This should be up in the order, but most of the times Q1 and Q2 hit it right.

2.3. Is the process fair?

Now since the documented methods are providing the expected results, is the method actually performing a fair test?

For example if you want to compare two toothpastes, you will want to have few basic things like:

  • Two randomly divided groups
  • Doubly blinded experiment
  • No confounding variables etc.

Also, is the research using sufficient sample size? Are the numbers compared properly using established statistical tests? etc.

If you see the previous statement (about domestic violence) as a research outcome (which is the only reason you should be believing it), then you can immediately identify what might be wrong in the experiment (spoiler: everything). When you are communicating via a proper scientific channel, you are automatically bound to go ahead only when the results are conclusive. But when statements like this start to frame the basis of judgments, you know the channel of knowledge generation and communication is severely damaged. I have heard so many people (due to their own personal reasons) around me agreeing on similar issues like 'media torturing and twisting facts'. Yet, shockingly small number of people criticize these incidences of quackery with similar enthusiasm.

The idea is to try and give all the information to help others to judge the value of your contribution; not just the information that leads to judgment in one particular direction or another.

Richard P. Feynman

Many argue that it's important to use dumbed down logic to reach a mass audience. But, there is a difference between dumbed down and simplified. One misleads people in thinking they know the actual thing while the other make them realize what they don't know.

The major problems with the propagators are probably not the numerous questions raised, but the ones unanswered because their claims are unfalsifiable. This approach is not a sane way to do any Science. Most of the claimed success is due to the confounding variables covering the obvious observations like eating healthy helps. Others left are weeded out in a huge placebo storm.

Many a times, you can't avoid irrationality. But, sometimes you can't afford. There is a huge responsibility on you when you are touching dead serious topics involving lives, deaths and crowd emotions. We can't afford statements like "AIDS can be cured through yoga", but while we can, let's ask if these statements have enough evidence to answer the questions.

Before finishing, I would strongly recommend reading Bad Science by Ben Goldacre (the quotes in this post are directly taken from there), specially the chapter 'The Doctor Will Sue You Now' if you are on a tight schedule. The story in the chapter explains the exact conundrum here with amazing precision and is downright serious.

I wrote this post because of the following few reasons, none of which involve initiating a hate debate against any of the specific ideas mentioned.

  • I felt like I should.
  • I still see many of my friends, relatives and myself reacting to senseless persuasions. So at its worst, this post can simply work as a sanity reminder for me.
  • Maybe I am just ashamed of what our species has done to rhinos.

And sadly, the rhinoceros count is not the only thing thats going down.

  • Thanks Gopal Sharma for suggesting ammendments to the list and other fixes.