>Dragan,
>
>> I had them as a two semester class - one probability, one statistics.
>>
>> One thing I learned for sure is that "probability is a science, while
>> statistics is a method".
>
>ROTFL!
>
>The main problem with "Introduction to Statistics" type classes is that they all ask the students to follow a recipe (detrmine which formula to use and plug in the numbers). Very boring. Statistics classes are much more interesting once of you have enough Probability and (mathematical) Analysis to understand the theory behind it all.

Right - I'm a mathematician, I know what you mean.

One trouble with statistics is that you should actually do a bunch of calculations yourself just to get the feeling of what's a reliable (or rather, representative) sample, and what's an unstable bunch of numbers that may give you wildly different results if just a few of them change a little.

And it's so easy to fool people using statistics. A simple "fuel prices have gone up 10%, therefore the bus ticket price goes up 10%" story actually passed many times, even though the fuel price contributed about 20% in the cost - so about 2% would be justified. Or comparing numbers across countries: I remember once Yugoslavia was European champion in number of traffic deaths, even while we had fewer cars and roads per capita. Explanation: definition of "traffic death" meant "dead on the spot or within --- days"; the number of days varied between 7 and 30 in most countries, and was legally set to 365 in Yugoslavia.

So while I usually trust the calculations to be correct, I rarely trust the sample, method, and intent :).