New profile posts

Is this a right description? I found it in a book but I am confused:
"Three main statistical methods are used in the data analysis: descriptive statistics, inferential statistics, and regressions analysis." and
"Inferential statistics tests the relationship between two data sets or two samples, and a hypothesis is usually set for the statistical relationships
between them."
yes exactly, I want random data like hemogloblin, kolesterol and weight, and lets say I have 20 deaths in 100 patient as binary outcome, I want to have random data ( each within specified limits , hemoglobln values should be between 9-16 for example ) that will give me the specified odds ratio after multivariate regression analysis
noetsi
noetsi
Sorry I did not know this was here. Be warned that unlike say spunky and dason I am a data analyst not a statistician. I am not sure what you mean by random data, or setting ranges on certain predictors. If the variable is random you can't limited it or in any case I have not seen that. I do not know a formal way to work back to an odds ratio. But you could manually say that the odds of this is X of the odds of that
Dear noetsi

I want to produce some experimental data

3 variables , each has prespecified odds ratio which are significant and in a predetermined confidence interval

for example after multivariate logistic regression analysis I find A variables odds 1.4 (1.1-16) , B variables 1.6 ( 1.23-1.40) and C variable 2.3 ( 1.4-3.3)

now I want 120 A,B, C variables with a prespecified mean, SD, how can I produce this data
Dear hlsmith

I want to produce some experimental data

3 variables , each has prespecified odds ratio which are significant and in a predetermined confidence interval

for example after multivariate logistic regression analysis I find A variables odds 1.4 (1.1-16) , B variables 1.6 ( 1.23-1.40) and C variable 2.3 ( 1.4-3.3)

now I want 120 A,B, C variables with a prespecified mean, SD, how can I produce this data
Dear Dason

I want to produce some experimental data

3 variables , each has prespecified odds ratio which are significant and in a predetermined confidence interval

for example after multivariate logistic regression analysis I find A variables odds 1.4 (1.1-16) , B variables 1.6 ( 1.23-1.40) and C variable 2.3 ( 1.4-3.3)

now I want 120 A,B, C variables with a prespecified mean, SD, how can I produce this data
Dear Shachar

I want to produce some experimental data

3 variables , each has prespecified odds ratio which are significant and in a predetermined confidence interval

for example after multivariate logistic regression analysis I find A variables odds 1.4 (1.1-16) , B variables 1.6 ( 1.23-1.40) and C variable 2.3 ( 1.4-3.3)

now I want 120 A,B, C variables with a prespecified mean, SD, how can I produce this data
S
Shachar
Hi Spektra.

Not sure I understood. Do you want to create random data that would follow these prerequisites?
Currrntly I'm doing masters thesis on migration using count model.poisson and neg.binomial have no problem but when i try zeroinflated and hurdle model some pridctor variable shows error in r software.....like computationally singular
as.matrix.fit$hessian...
So how I can fix the problem and include the variable in analysis???
Hi Spunky,

I recall you being versed in psychometric methods and was wondering if I could pick your brain over something. I can't seem to send DMs, but if you are able to , could you kindly reach out?

Best,