Any statisticians i...
 

  You don't need to be an 'investor' to invest in Singletrack: 6 days left: 95% of target - Find out more

[Closed] Any statisticians in the house?

7 Posts
6 Users
0 Reactions
57 Views
Posts: 0
Free Member
Topic starter
 

Gents,

I am doing a bit of coursework and i'm having a bit of an issue with a question.

I am doing a section on income equality ie. fraction of population vs fraction of income. I need to get some data points for an equally distributed population, ie. 50% of people earn 50% of income / 45 degree line.

I am told that the mean (hence the median given distribution) is £528/week and that the standard deviation is £281/week.

How can i go about plotting this line? I am assuming that I need to use Z=X-Mu/Standard Deviation (£281) to find a Z value in order to find X.

May be barking up the wrong tree - any professionals in the house? robowns at hotmail dot com if you are willing to school an idiot.

Many thanks,

Rob


 
Posted : 24/08/2016 8:07 pm
Posts: 0
Free Member
 

On STW?

Everyone will earn well above the average, but will pretend not to...


 
Posted : 24/08/2016 8:08 pm
Posts: 2081
Free Member
 

Probably one for the chat forum.


 
Posted : 24/08/2016 8:11 pm
Posts: 17209
Full Member
 

Sorry I'm not following what you are asking. Do you want to calculated fractions from a distribution with mean 528 and standard deviation 281? A cumulative normal function can help with this.

What do you mean by "equally distributed" same CV (stdev/mean)? A Z-score is one tool to characterise how far from the mean one is.


 
Posted : 24/08/2016 8:11 pm
Posts: 0
Free Member
Topic starter
 

Equally distributed meaning that 528 is exactly 50% of income/population. I need to show income in 5% increments. I need to show the solution so I assume dividing by 50 and times 45 is too simple.


 
Posted : 24/08/2016 9:46 pm
Posts: 1277
Free Member
 

I don't understand the 45 degree line.
It sounds to me like you have been set an even distribution problem. So plotting a graph with income as the x axis and % of pop as the y axis then you will get a symmetrical bell curve.

I'd tackle this my putting in the mean and sd in [b]named[/b] cells at the top of an excel workbook, then have a table below with deviation %, income, % of population as the column, deviation will be for -50%,-45%,...0%,...+50%, other two columns worked out by formula referencing the named cells, make a graph from columns 2 and 3.


 
Posted : 25/08/2016 7:14 am
Posts: 17209
Full Member
 

In that case you want the normal distribution function. Create a column from zero to 1 in steps of 0.05 increments. Then n the column next to this use norminv(A_,528,281). Copy this down and it will give you the income for each percentile from 0 to 100 percent, centred on your distribution.


 
Posted : 25/08/2016 8:00 am
Posts: 943
Free Member
 

similar to TiRed but slightly different.

If you want to plot a normal distribution of the actual incomes (x)

in Excel,

column A: list numbers from 0 to say 1000, in steps of 10
column B: =NORMDIST(A1, 528, 281, False)

copy down, then plot as a line or XY scatter

For the cumulative plot:
column C: =NORMDIST(A1, 528, 281, True) and copy down.


 
Posted : 25/08/2016 8:55 am

6 DAYS LEFT
We are currently at 95% of our target!