Help with some statistics!!

[darkstar]

Hi all, just wondering if anybody on the forums is good with stats? I have been given a long list of questions to complete and i am really struggling!
Some examples of some questions:

Test the null hypothesis that there is no difference in travel distance between men and women in population. Use a 0.05 significance level.

Calculate the linear correlation coefficient between income and travel distance.

I have completed a fair few but have reached my limit!

If anybody can help i'll provide further details, it would be really appreciated!

 

[Fnaar]

It's statistically very likely that someone on here can help... but not me, I fear

 

[ianrauk]

Let me get back to you on these....... like never...

 

[gazmercer]

I'm thick so no chance of me helping out.... I don't even understand the questions.

I've gone dizzy now, must sit down

 

[lady_rider]

Presumably you have the data to do it... but...

Test the null hypothesis that there is no difference in travel distance between men and women in population. Use a 0.05 significance level.

Probably a t-test if your data has a normal distribution. If it doesn't, then use mann-whitney.

If you want to test the data for normality before doing the tests, use kolmogorov-smirnov. A p-value of GREATER than 0.05 means you have a normal distribution and can use a t-test (it tests the null hypothesis that 'the dataset is not deviate significantly from a normal distribution). If it's LESS than 0.05, then use mann-whitney.

Calculate the linear correlation coefficient between income and travel distance.

Pearsons correlation statistic if normally distributed, spearman's rank if not

NOTE: correlation does not imply causation. ie if you get a significant correlation between income and travel distance, you cannot say that greater income causes you to travel more, or that travelling more gives you greater income. You can only say that there is a relationship between travel distance and income. If you wanted to test for causation, you'd have to use a regression.

Ahhh... my PhD may have been worthwhile after all...

 

[marinyork]

Darkstar it would help if you posted more as to gauge what level. I don't usually like getting involved in these threads but if you go along to the university library and read schaum's outline of statistics and find the relevant chapters it will give you an elementary undergraduate grasp of the ideas and there are worked examples. If you want some pointers for grad level Maths I can give you some. If you're on a social science course and will have to use SPSS then I recommend that you learn to use R.

 

[darkstar]

Ta for the replies, think i'll get my head in a book, i've just had enough of all these numbers and sums tbh! I'm doing a geography degree, NOT mathematics!
I though the last of this was in Physics A level!

 

[ComedyPilot]

I'm thick so no chance of me helping out.... I don't even understand the questions.

I've gone dizzy now, must sit down

+1

Move over on the sofa gaz, I need a sit down too.

 

[threebikesmcginty]

Test the null hypothesis that there is no difference in travel distance between men and women in population. Use a 0.05 significance level.

Calculate the linear correlation coefficient between income and travel distance.

errrr...

Anyone fancy a pint?