![]() |
|
|||
|
I have never taken true stats, but I am a mathematician and taken courses like intro to political analysis and geographic analysis which used stats at a low level.
What exactly does the project require? Do you need a survey? Do you need x amount of published papers? Will you need to use microcase, SPSS, SAS? Define the project a bit more and I think people will be able to help with ideas. Jerrod |
|
||||
|
I have taken STats. What were you looking to do? Simple values, standard deviations, what?
__________________
Papasmurf: "there are no miracle supplements, fix your diet" http://www.bodybuildingforyou.com/forums/showthread.php?t=39994 |
|
||||
|
i tutor stats so maybe i can help. I assume you just need the basics like standard deviation, mean, mode, distribution curves. you can always do search for already done surveys or researchs and just use those datas, or make your own data. Maybe marriage stats, or weight in a certain age group. If you need stuff like regression or damn forgot what its called.. haha.. two datas and how strong correlated they are.. then you can maybe use amount of tv watched a week with some kinda image issue (maybe how big a person thinks they should or how skinny they should be) .. haha i dono what else to say.. hope that helps
__________________
Dog20 for BB4U troll. Vote Yes. |
|
||||
|
I will post up the requirements as soon as I get home..thanks for the help folks.
man i dono, do you know any sites I can get data and stuff from? I tried to google statistics projects and got a lot of garbage. Waht you said is about on queue, and I need a scatter chart. here is the requirements: Last edited by Sean; 06-27-2006 at 04:54 PM. |
|
||||
|
lmao its too long, i just looked thru a bit.. well i dono any site for statistics, but if ur desperate, u can always type something like AIDS statistics in a certain year and I'm sure that will be ready for you. I saw from requirements that u need a study .. and how to pick ur sampler or whatever.. those are easy right? for those, I'd think using your class or survey is the best, like ask them how many siblings they have and the amount of time they have on homework. About choosing a sample, I'd just go with random sampling. forgot what the other one is. And for the study, you can do experimental about working out and supps (ex. no explode). A few days without no explode, work out a routine.. then a couples with no explode, with the same routine.. then measure your heartbeat, or how many minutes it takes for u to rest after workout to fully feel recuperate.. that'd give u treatment and control and all taht stuff
__________________
Dog20 for BB4U troll. Vote Yes. |
|
|||
|
government sites are good for statistics. that and respected organizations
__________________
Training Log: http://www.bodybuildingforyou.com/forums/member-journals-light-weight/43055-thearas-log-why-not.html middleweight, here i come! |
|
|||
|
wow i hated stats...as far as profects go i dont think i ever had to do one...good luck though...its hard to make stats interesting
__________________
I don't care if I die, as long as they bury me in big f*cking box |
|
|||
|
http://factfinder.census.gov/servlet...rvlet?_lang=en
That is the US census. I have used it many times. What stats package are you using? If, say, it's SAS, try searching for SAS data sets. You can then avoid inputting all of it into the package, thus saving time. http://lib.stat.cmu.edu/datasets/ some data sets here. I didn't read your whole projects req. yet, but I did read the first part that said pick a topic and design a way to test it. As someone mentioned before, why not do something on fitness? I went to google and typed in "fitness data sets" and the first link was NY state community health. Some of the sets are fitness, tobacco usage, socio-economic status. Just from that, an idea that pops into my mind would be to compare fitness levels (make sure this is well defined) of children in the rural south vs children in northern cities. You could look at their socio-economic status and the fitness levels of their parents. Children that are in the low income bracket tend to have parents that do not eat as healthy and are usually overweight. The idea that always works for me is to find a theme (like body building), and then think about ways in which you want to measue some particular phenomenon. think about how they could be coorelated. From what I understand from my statistical methods class, all you want to do is prove or disprove and make sure you have a strong argument (most academic research is usually done by taking someones topic and trying to disprove it). Thats my 2 cents what ever it is worth. Jerrod *in addition Remember that at first things may not appear connected but they may in fact be. Heard of the book freakeconomics? The guy wrote a whole book on odd connections. One of his claims is that since abortion has been legal and easier to access, crime has gone down. I once saw a paper on country music and suicide. Last edited by Jerrod; 06-28-2006 at 12:33 PM. |
![]() |
| Thread Tools | |
| Display Modes | |
|
|