# What is the best way to study Statistics?

By terrych

@terrych (1227)

United States

6 responses

@ahgong (10064)

• Singapore

28 Oct 07

Having trouble with statistics as a subject eh?
Well, you are not alone terrych!
I can only say practice, practice, practice and more practice!
Get someone who is good at the subject to tutor you. He/she should be really patient with you when it comes to dealing with the math.
Do lots of sums and find out why that formula is used to solve that sum instead of another. Understanding how to tackle the various questions is more important than just knowing the formulas.
Once you can start to recognize the different formulas to use to tackle the various sums, then you are on the way to passing the subject!
Good luck to you on your stats exams!

@ahgong (10064)

• Singapore

29 Oct 07

well, like I said before, practice, practice, practice and more practice!
I remember the pain of having to go through the exams for stats. It is horrible. The number of extra hours I have to put into this subject to pass it is really a pain! I am glad it is over!
I did the ten-year series forward and back, ensuring that I know how to do each and every sum twice!
Then I started work on those past year papers of other schools. Some of the high school advance statistics are also pretty tough. If you can get your hands on them, you can practice more!
The only way to overcome this subject is to keep practicing till you know how to tackle the problem with the right formula! ;)
Again, all the best to your exams!

@theprogamer (10539)

• United States

27 Oct 07

Things like Median, Mean, Min and Maximum are easy. As is Standard Deviation (since half the time its given to solve for the statistic in question). But I will admit with all of the language and then the execution its also easy to get lost. Sometimes you want to equate the definitions with examples you can relate to.
Median - number separating the highest half from the lowest half in a sample/population
(5, 10, 15, 20, 25) - in this case it would be 15
I'd relate it to things median age (35) or median household incomes (25-30,000) I don't know the exact total but hopefully it gives you an idea.
Mean - the average of a population/sample
- using the same example it would be 15 again. But remember to add up then divide by the population to get the mean.
Min - minimum number of a population/sample
Max - maximum number of a population/sample
Mode - value that has the largest observations in a population/sample
(5, 5, 10, 15, 20, 25, 5, 5) - so it would be 5 in this case.
Just think of names that are very common (Smith for example) and try tying it to that or whatever you choose.
Skew/skewed - a problem that occurs if the sample/population has a very lopsided variable in the case. A classic example is rich vs poor population or even incomes.
Expected value - the outcome expected in a population/sample or the expected mean.
The common example is a roll of 6 sided dice. With each item in the sample being an outcome divided by the total population. ((1+2+3+4+5+6)/6)= 3.5
So sometimes you don't get logical answer but that is what expected value yields, a mathematical expected value from the probabilities. Another is expected values with children. 0.5 chance for a boy or girl and then you multiply that for each future child. i.e 0.25 chance for a boy or girl or two boys.
This may help out http://www.stat.psu.edu/~resources/ClassNotes/rho_02/index.htm
Variance - the spread of the sample/population
The numbers/outcomes minus the mean value, squared, then added together and divided by the number of sample/population.
Use this as an example and help http://www.mste.uiuc.edu/hill/dstat/variance.html
This is a simpler example http://www.animatedsoftware.com/statglos/sgvarian.htm
Standard deviation - square root of the variance
Try to remember common standard deviations like 5, 1, 0.5
From the last link example http://www.animatedsoftware.com/statglos/sgstdev.htm
Null hypothesis - set up to be negated in favor of an alternate hypothesis
Alternate hypothesis - another hypothesis to deny the null hypothesis
Typically a p-value is found out and if its small then the null can be rejected
P value is the point of an extreme outcome and the probability of getting it.
T statistic - are the means of two populations equal (null hypothesis)
T= Two populations/samples just to try and help you remember
Need more help
This is notes and presentations http://www.stat.psu.edu/~resources/bystat.htm
http://www.graphpad.com/quickcalcs/index.cfm (Value Calculators)
I know you cant use them but at least you can check yourself.
Phew. That was a bit exhausting.
Hopefully no one will mind if I just nap now.
Good luck Terry, I'm rooting for you. Maybe even some my strength or spirit could give you some invisible help or luck.

@theprogamer (10539)

• United States

28 Oct 07

Gah, I'm looking at this again, my eyes are bleeding.
I'm glad it can give you some help Terry and I still hope you get the best grade you can. The worst part is I know I missed several other points.
You can also ask again for help under the student discussion, or in this thread. Or just come get me if you need a second set of eye and experience again.

@kimiwarrior_91 (21)

• Malaysia

4 Nov 07

you need to do more exercise. Statistics is not very difficult to understand. Only standard deviation and variance need more calculation. Most students having trouble to solve problems involving standard deviation and variance. As I wrote just now, you need more exercise because reading is not enough to understand it.

@goodsign (2287)

• Malaysia

27 Oct 07

Hi Terrch,
As last year I had finished up locally, my Master's Degree in IT and one of its module is 'statistic'. One of the web link as my reference previously, was;
http://www.amstat.org/careers/index.cfm?fuseaction=whatisstatistics
I hope my small help will benefits you.
Happy Posting and hve you nice day. Thanks.

@ssh123 (31085)

• India

27 Oct 07

I had studied statistics in my college days when I was studying econometrics. But that statistical analysis was quite simple like mean, median and mode, a bit of hyperbola and parabola, demand and supply equations, index system etc.
My nephew is studying applied mathematics in Michigan University and he tells me that it requires the knowledge of pure mathematics to set problems and solve problems in applied statistics.
I believe one has to have good knowledge of mathematics to study statistics. Mere studying theory one cannot get through the exams I believe.

@terrych (1227)

• United States

27 Oct 07

Yes, I believe you are right, with the theory I am going not where... so I tried to do some exercises like doing all the math, and I was fine until then, but in the last test the teacher just gave as a bunch of SAS output... and I felt absolutely lost... I started panicking, and I do not want that again in the final exam... that is why I am trying to get some advice about how to study STAT...
Thanks for your answer!

@sgarjun (124)

• India

5 Nov 07

I am from India. I took statistics after my High School.
Now I am doing b.stat(hons.) in Indian Statistical Institute.
I have good knowledge in statistics and in mathematics too.
So tell me what topics you are studying in statistics?