turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- Rate comparisons

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

09-30-2011 03:06 PM

I have samples that I have treated with different conditions and want to know if these conditions have impacted the rate of change over time. Does anyone have a suggestion for how best to do this comparison? I have looked at the different rates in a bivariate plot to calculate the slope, but is there a way to calculate the slope and then compare them to each other with out all the manual transcription of numbers? It would be nice to have confidence intervals around those slopes and see if a graphical representation how they overlap.

Any ideas?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

10-01-2011 09:02 PM

We need more information about your design to offer suggestions..

Are the samples measured repeatedly over time or is the measurement destructive (so each sample x time x condition is independent)?

If repeated, are the repeats over time or over both time and condition?

What are the units of the outcome meaure (binary, continuous)?

If continuous, do you assume normality?

What is the sample size?

I'm not sure what you mean by "manual transcription of numbers". SAS is generally designed to work against the sample level data.

Doc Muhlbaier

Duke

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

10-02-2011 07:44 AM

Hi Doc Muhlbaier,

Thank you for spending some time to look at my post. The same sample is measured over the various time points so it is not destructive. The repeats are done with replicate samples. There are five conditions and four replicates of each sample measured over a five day period. The measurement is continuous and I am assuming normality, but I will try to look at that during data analysis. The sample size is limited though; having only four replicates of each condition so I am not sure how significant that test will be.

Thanks for you insight,

Kevin