Lies, Damned Lies and Statistics.

Stare. Think. Scratch head. Stare. Idea! Research (Wikipedia). R program. Program fails. Scratch head. Rewrite. Fails. Rewrite. Graph!

And repeat.


Dave was going well. His pile of graphs got bigger by the day. Such pretty things, even his supervisor had fallen for them. So, so pretty.

However, after the days turned into weeks, the weeks turned into months and the months turned into years, Dave finally decided perhaps he should actually do some statstics for his statistics PhD, to add to his pretty pretty graphs. Therein lied his mistake.

"What? That can't be right." Nope, when Dave spoke to himself it was almost always a bad sign.

The statistics suggested his results were wrong. He checked again. And again. And again. How could it be? The graphs were so very very pretty.

Defeated, with only five minutes before his meeting with both supervisor and sponsor, why, oh why, had he checked the stats?

Dave now had several options, most of which involved faking his own death, faking his results or coming clean. The fool chose to come clean.

"Hmm.." said his supervisor.

There were several possible explanations as to what happened next. Perhaps the supervisor felt sorry for the student. Perhaps he thought stupid student makes him look bad. Perhaps he was still thinking about the pretty pretty graphs. Or perhaps he wasn't really paying attention.

In any case, the facts were brushed to one side, some other tests were suggested and the conversation continued as if it had never happened.

Ah the beauty of statistics. If at first you don't succeed, just keep changing the test until you find one that agrees with you. And if THAT fails, create a whole new test.

Phew.