nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

analyzing average +/- st dev data

unknown

If don't have raw data only average values and st dev can I subject this data to ANOVA?

I am trying to analyze some cell culture data and I have DNA and hydroxyproline data that derived from destructive testing. So I have two data sets for a given condition (DNA raw data and hydroxyproline raw data), obtained from different cell cultures (treated the same though). I took averages and st deviations and used error propagation proceedures to derive a hydroxyproline per DNA value (average and st dev).

Can I use JMP to perform ANOVA on data derived from this average valuue and error propagation calculations?

Accepted answers

All comments

unknown

Forgetting JMP for a minute, I'm not sure it makes sense to do ANOVA on this data. I specifically would like to know more about "error propagation proceedures [sic] to derive a hydroxyproline per DNA value (average and st dev)". I specifically want to know about the error propagation part, not the hydroxyproline per DNA part.

If you are doing what I think you are doing, then an ANOVA would be testing the results of your error propagation proceedures [sic] and not really testing differences between means of the two populations.

Message was edited by: Paige

unknown

As the DNA test and hydroxyproline test are destructive I have DNA data for some specimens and hydroxyproline data for others. I can calculate AVG +/- St Dev from this for each measurement.

Then to get ratio AVG(hydroxyproline) / AVG(DNA) +/- %error.

%error = sqrt( [%error(hydroxyproline)^2] + [%error(DNA)^2]).

I just converted the %error or coefficient of variance back into St Dev giving average hydroxyproline to DNA ratio with st dev. This is very common technique and can be found in any basic undergrad analytical chemistry book (I'm sure in others too but I always pull that one off the shelf)

unknown

Why don't you simply do an ANOVA on the data from the measurements; instead of turning the data into averages and standard deviations and ratios and proprogations of errors?

rc_hertzy

I often encounter publications that only report means and std errors, so the option of using raw data does not exist. When I doubt their statistical conclusions I would like to run my own analyses. Can JMP use such data in ANOVA and in testing for equal variances?

unknown

I don't think JMP can do this. I think this is a pencil and paper exercise.

You can probably use a JMP data table as a spreadsheet and use column formulas to do the calculations. In addition to mean and SE you will need the sample sizes.

mpb

Also, from their definitions some of the tests for equal variance depend on individual data values and can't be calculated from the summary statistics.