Hello Md. Sabbir Hasan,

The answer depends on how much data you have. If your data consists of two cases (one each of the two food items), and you have six measurements taken from each case, I can think of no good statistical method that will give you a multivariate (six measures simultaneously) comparison...you would need replications of the two food items.

If you had, say, 20 replicates of each food type, and all 6 measures for each 40 cases, then a manova could possibly be conducted (grouping variable = food type; measured or dependent variables = the six properties). If you felt obliged to evaluate the properties individually, then a set of anovas or independent t-tests (same hypothesis either way, with only two groups) could be executed. However, the latter approach requires 6 hypothesis tests minimally, whereas the multivariate approach only requires one.

Good luck with your work.

Thanks a lot sir David Morse for your valuable comment.

David Morse is certainly correct.

I would like to point out that multiple univariate analyses address quite a different question than does MANOVA. There is often confusion regarding this. Carl Huberty and I detailed this in a PB paper a while back. If you are interested, it is:

Article Multivariate Analysis Versus Multiple Univariate Analyses

As David said, you need to accumulate some subjects in any case.

My best on your research,

John

John Morris, Definitely, your article is fantastic in this regard,,, so much thankful to you.