network profile
| bio | website | |
|---|---|---|
| location | ||
| age | ||
| visits | member for | 11 months |
| seen | May 31 '12 at 4:09 | |
| stats | profile views | 0 |
|
May 30 |
comment |
How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. Thanks a lot! One last question: So what to do if the standardized residual distribution is non-normal? |
|
May 30 |
comment |
How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. I see, so it is enough to check the error distribution, it helps a lot!! I was confused esp. because of some sources like the first: "Simple linear regression allows us to look at the linear relationship between one normally distributed interval predictor and one normally distributed interval outcome variable"... It's confusing... |
|
May 30 |
comment |
How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. Another webpage where they suggest that I should check for normality of variables, though they also talk about the residual plot: dss.princeton.edu/online_help/analysis/regression_intro.htm |
|
May 30 |
comment |
How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. Maybe for introductory textbooks they write that IV/DV should be normal, because then they don't have to explain how to check the residual distribution? I saw on some webpage: "If you are doing a regression analysis, then the assumption is that your residuals are normally distributed. One way to make it very likely to have normal residuals is to have a dependent variable that is normally distributed and predictors that are all normally distributed, however this is not necessary for your residuals to be normally distributed" Can this be the reason? |
|
May 30 |
comment |
How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. Another source that says that non-normal DV/IV have an effect (it does not say that you cannot do it) on the results: utexas.edu/courses/schwab/sw388r7/Tutorials/… |
|
May 30 |
comment |
How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. The page: ats.ucla.edu/stat/spss/whatstat/whatstat.htm |
|
May 30 |
comment |
How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. Sorry, I'll check the printed titles later... what I have now is a web page about what test to choose saying: "Simple linear regression allows us to look at the linear relationship between one normally distributed interval predictor and one normally distributed interval outcome variable." Do they refer to the fact that the variables should be taken from a normally distributed population? (but the data itself does not have to be normal?) I thought that there was only assumption about the residual normality at first, but reading about these I became confused... |
|
May 30 |
comment |
How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. BTW About the no-assumption-about-normal-distribution of IV/DV, I found contradictory things. I've read in several statistics books that normal distribution of the IV/DV is very important, and skewness can result in faulty readings. If I check the residual distribution and its normal then I'm OK? |
|
May 30 |
comment |
How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. Due to the presence of 0s and 1s. I guess it is pretty skewed to put into e.g. an OLS model... what model should I use that have no assumption about distribution? I guess non-parametric tests have no assumptions, but OLS do. What regression model do you talk about that have no assumption about distributions? I do not want to transform away the spikes, that's why I asked. I wonder what to do with a variable like this? And even if I transform log has problems with 0s... |
|
May 30 |
asked | How to deal with a variable that ranges from 0 to 1 and the distribution has two spikes at these values with normal-like distribution in the middle. |
