/sci/ - Science & Math

[Anonymous No.5298510]

/sci/, I need your help

It's statistics, and I need to mathematically prove that values that are transformed back from the log scale still hold significance.

I know it's a homework thread, but this is not your generic homework, and if I fail this I am literally fucked. I am a CompSci major.

----------------------

Now, I am comparing two positively skewed distributions.

Why is the backtransformed arithmetic mean from the log scala still the same as the original mean?

How can I interpret the backtransformed mean value difference? And how do I interpret the backtransformed null hypothesis?

Also: Why is all of this independent from the log basis?

How do I go about this guise? Any help is appreciated!

[Anonymous No.5298553]

bamp

[Anonymous No.5298667]

pl0x help guise, I am dying here

[Anonymous No.5298734]

>>5298553

This is actually a pretty ignorant thing to think as a cause for it, considering the studies of attempts at consciously switching them generally just end up with the kids being unhappy with their situation. There are broadly-speaking different interests between the genders with overlap and there's no reason why it should be an exact 50/50 split in every profession to satisfy a political agenda while increasing job-dissatisfaction.

The more equal countries get, like here in Sweden, the gap actually becomes wider. If there isn't a cultural feeling that there's a _need_ to fight for the right to be a certain profession then more people pursue careers they're naturally inclined towards. That's a good thing. Anyone who is inclined towards more technical or social professions not facing any prejudice just because they're not in the majority, but everyone seeking their preferred direction.

As for the original topic I wouldn't be very helpful. Not my strong suit. Just bugs me just blatantly state a correlation as causation when there's even studies showing that it's ass-backwards.

[Anonymous No.5298736]

>>5298510

This is actually a pretty ignorant thing to think as a cause for it, considering the studies of attempts at consciously switching them generally just end up with the kids being unhappy with their situation. There are broadly-speaking different interests between the genders with overlap and there's no reason why it should be an exact 50/50 split in every profession to satisfy a political agenda while increasing job-dissatisfaction.

The more equal countries get, like here in Sweden, the gap actually becomes wider. If there isn't a cultural feeling that there's a _need_ to fight for the right to be a certain profession then more people pursue careers they're naturally inclined towards. That's a good thing. Anyone who is inclined towards more technical or social professions not facing any prejudice just because they're not in the majority, but everyone seeking their preferred direction.

As for the original topic I wouldn't be very helpful. Not my strong suit. Just bugs me when people just blatantly state a correlation as causation when there's even studies showing that it's ass-backwards.

[Anonymous No.5298740]

>>5298553

This is actually a pretty ignorant thing to think as a cause for it, considering the studies of attempts at consciously switching them generally just end up with the kids being unhappy with their situation. There are broadly-speaking different interests between the genders with overlap and there's no reason why it should be an exact 50/50 split in every profession to satisfy a political agenda while increasing job-dissatisfaction.

The more equal countries get, like here in Sweden, the gap actually becomes wider. If there isn't a cultural feeling that there's a _need_ to fight for the right to be a certain profession then more people pursue careers they're naturally inclined towards. That's a good thing. Anyone who is inclined towards more technical or social professions not facing any prejudice just because they're not in the majority, but everyone seeking their preferred direction.

As for the original topic I wouldn't be very helpful. Not my strong suit. Just bugs me when people just blatantly state a correlation as causation when there's even studies showing that it's ass-backwards.

[Anonymous No.5298748]

Statistics? that's weird i'm majoring in computer science as well but with the extra engineering, we do not have to take statistics just trig and up till linear algebra. you're taking a class that you do not need

[Anonymous No.5298761]

>>5298740
Dude, these comics are for lulz. No /sci/entist gives a shit about gender issues or demographics.

I guess I could do it myself if I dug something up on how exactly log transformation in statistics works. The problem is just that I have a whole lot of other SAS related exercises, and I left it for the last day. I barely did the previous exercises because I signed up for too many courses, and if I don't do every exercises from now on, I fail Statistics.

I am gonna work all night, I hope it works out. Turing watches over his programmers. /sci/, I will not fail you.

One last bump and off I go the lectures.

[Anonymous No.5298764]

>>5298748
Postgrad in Biomedical Engineering, Statistics is obligatory.

[Anonymous No.5298854]

>>5298748
In California all engineers/comp sci people take at least one calc based prob/stats course. I majored in civil and we took one course for engineers.

[Anonymous No.5298955]

>>5298510
Bamp for answer.

>>5298854
Totally agree, seeing how the pioneers of computing were mathematicians we Programmers get thought too little math.

[Anonymous No.5298968]

>>5298748
We've done combinatorics, sequences and series, quick refresher on calculus, a little on using an integrating factor for a differential equation.

[Anonymous No.5299272]

One more bump.

Can't be that we haven't got any statisticsfags in all of /sci/?