- Thread starter nelson
- Start date

thanks for your question.

the question is a little bit misleading, though.

About the first part of it: on the one hand you're asking about linear regression. did you mean "ordinary" when you wrote simple? there are quite a couple of different regression models, the most "simple" would be OLS, which stands for Ordinary Least Squares.

Secondly, on the other hand, you are asking about "univariate" statistics. Regression asks whether there is some relationship between variables (either one independent and one dependent which is called "bi-variate", or several independent variables and one dependent which is called multivariate).

When I think of uni-variate statistics, I think of descriptive statistics describing a variable's distribution... (such as mean, variance, frequencies etc).

However, I'm not really familiar with Generalized Linear Models where univariate statistics also plays a role. Maybe you'll like to look into it.

Cheers!

--

I'm very knew to statistics, too - feel free to comment on my posting, I'm looking forward to your suggestions.

I am pretty sure that e.g. multivariate analysis has more than one dependent variable, right? Or do you mixed it up and you meant multiple regression?

for that see here: http://talkstats.com/archive/index.php/t-522.html

Please make sure that you are correct when you answering, people like me trust you.

Is there anybody else who can help us

once again my question, what is the actual difference between univariate analysis and ordinary linear regression, since I have for both one dependent variable, right? Just give me an easy example, I guess I am just confused