Tuesday, February 24, 2009

Cherry-picking evidence

Now I can let go of many things - but not misrepresentation when it comes to movies. So here is a pointless piece from Washington Post that argues that movies with "faith and values do much better than movies that overtly attack traditional faith and values." And here is the evidence in defense:
As we will see during Sunday's Academy Awards, last year was no exception. Six of the most successful movies of the year -- "Wall-E," "Iron Man," "Indiana Jones and the Kingdom of the Crystal Skull," "Prince Caspian," "Gran Torino," and "The Curious Case of Benjamin Button" -- contained strong redemptive content with positive Christian references.
...
Not only did moviegoers prefer heroic movies with very strong moral virtues, they also rejected movies with anti-Christian, secular, nihilistic, and atheist content like "Religulous," "Adam Resurrected," "Save Me," "Wanted," "Hounddog," "Bloodline," "Hamlet 2," "The Love Guru," "Stop-Loss," and "Saw V."

The difference between the domestic box office averages for these two categories was $59.9 million per movie for the positive movies versus only $10.4 million per movie for the negative ones.

This is what we call cherry-picking the data. I'm going to ignore whether the "positive" movies mentioned above have Christian references or not - or whether anti-war films, like Stop-loss are really anti-Christian. Instead, let me just focus on the cherry-picking of data:

If we go by box-office numbers, then the biggest movie of 2008 was, by far, The Dark Knight, and it has just crossed $1 billion mark worldwide. It is odd that while the whole article is about the box-office success of "positive Christian" movies, the movie that pokes a hole in the theory is conveniently ignored. In fact, by looking at the top-grossing films of the past 3 years (Spider-Man 3 - 2007; Pirates of the Caribbean: Dead Man's Chest- 2006; Star Wars: Episode III - 2005), Hollywood should decide to definitely ignore movies with "positive Christian values".

Hey! Leave the movies alone.
Read the full article here (there are more idiotic claims in there).

Another example of cherry-picking data is this claim from a review by Ziauddin Sardar (the same claim is also made in the second part of BBC radio program, Islam & Science). Talking about science in the Muslim world, he writes:
The nexus between science and the abuse of political power is much more than a coincidence, Masood argues. Even today, science is in comparatively better shape in countries with dictatorial regimes, such as Egypt, Pakistan and Iran.
Ok - so I'm assuming that they (Sardar and Masood) are only talking about Muslim countries (otherwise the claim is obviously false). This statement may in some ways be true - but there are clear and important counter-examples that need to be mentioned. For example, Malaysia is one of the most scientifically advanced Muslim countries - but it is not a dictatorship. Ditto for Turkey. In terms of science & technology, these two are giants in the Muslim world. Plus, the Iranian example is complicated. Iran does have regular elections (fairer than most in the Middle East) and its science has improved significantly since the Revolution - which overthrew a dictator. So where does Iran fit in? I think this particular claim linking progress of science with dictatorships in the Islamic world is waaay too simplistic - it fits well for Pakistan - but it omits crucial counterexamples (including numerous dictatorships where science is not ding so well...).