Statistics - Do they lie?

This afternoon, I brought a few students to Tanglin Sec for some prize presentation ceremony. The keynote address was "What the numbers say?" by A/P Chua Tin Chiu. Really a simple yet eye-opening talk about statistics.

He quoted some eye-catching headlines from newsapers,

1) "There are 5 times more widows than widowers after the age of 50, thus most husbands are tagged to death by their wives"
2) "You have 40% chance of living up to 100 years old if you are the first born"
3) "Students surveyed there are much more guys than ladies who sped past the red light"

Then he asked: Are you sure about the conclusion statements?
For example, for the headline labelled (3), he asked, "did the students count the number of guys who did not speed past the redlight? Do they happen to be more than the number of ladies who did not speed past? If there are more males than females on the roads to begin with, then the number of guys who speed past should be naturally more, isn't it?
For the headline labelled 2, he said, "The survey has been done in a village in china, where the study was done only to people who live up to 100. However, the number of first born will always be more than the number of 2nd born, who will always be more than the number of 3rd born etc (haha...for obvious reasons right? You need a first born before you can have a 2nd born) and so the number of first born who live up to 100 should also be more than the rest, isnt it?
For the headline, labelled (1), the statistics might be true, but the reason given might not be valid. Some other possible reasons might include women living longer lives than men, older men usually marry younger women as well as older men usually remarry as well.

A/P Chua quoted more newspaper reports with regards to statistics in newspaper, and asked, "Why do you think percentages have been quoted in areas where direct figures might be more accurate, and vice versa? Isn't it the way the media (or the government) wanting to bring across a particular idea?

Another wonderful headline that students will love to quote if true,

"Study shows More tuition = poorer grades"

Well, the study was concluded based on a huge pool of students who wrote down their grades, as well as the number of hours they had for private tutoring. But the BIGGEST question is, "How will the students who have private tuition fare, if they do not have private tuition? Isnt it true that those who need private tuition are generally those who may not cope that well, and so extra help to be given?"

To me, statistics can be summed up by a quote from Benjamin Disraeli, one of England's finest prime ministers, "There are 3 types of lies - lies, damned lies and statistics."