How many alkaloids levels were used?

Glycoalkaloids levels means their concentration affected by inoculation. In my study glycoalkaloids concentration were slightly increased by inoculation. However, i do not know the exact statistical test for finding association b/w the glycoalkaloids concentration and disease

The scale type and range of disease severity are important to decide statistical methods. I think you don't research the effect of alkaloids on the severity. You want to determine correlation between disease severity and alkaloids levels.

If all assumptions are satisfied, canonical correlation analysis is suitable to determine correlation between disease severity and alkaloids levels.

Why dont you just go with regression?

Dear shankar please help me in analyzing the data, i have attached the detailed data

dear Venu Perla, thanks for your suggestion, however, i am extremely poor in statistics. i am attaching my data if you plz analyze i will be thankful to you and your help will be acknowledged

Abdul........from table number 1. it looks like to me that there might not be significant differences between the water inoculation and Pi inoculation. Did u perform any test to find out if they are significantly different?

i performed LSD which revealed insignificant differences for control and inoculated values. however, both in table 1 and 2 significant differences were recorded for days after inoculation. thanks for your interest

Hi Abdul, i am really interested in what u are doing. Besides that I also work with Phytophthora infestans and i love statistics.

Abdul I am sorry to say that table 1 does not make sense to me because if we look it logically it says that either u inoculate your plant with water or Pi the glycoalkaloids production is the same. And I dont know how u measured that amount.

And its certainly true that day0 and day 12 would have significant differences because day 12 is the sum of everything.

I dont what is your hypothesis but if your hypothesis was that " there is significant differences in production of glycoalkaloids when infected with Pi " then i am sorry to say that your hypothesis is false. Because i dont see differences in production of glycoalkaloids

also now talking about table 2 your replication values also doesnt seem statistically different , so you can pool them and do analysis. But u need to be sure what u want to do with your data.

can u please define your research hypothesis clearly so that i might help u?

I am also a learning student like you............and this happens in research.

feel free to write...........

Hello Abdul,

first I think you should describe your study and to know your chances. Done this you can define which "best test" to use in their sample.

But here are some tips that may help you.

First take your samples. How many groups are there? You have a "study group" and a "control group"?

Before anything: Make an analysis of the data to know if your study has a homogeneous sample to determine if your tests are "parametric or nonparametric". (Use the Shapiro-Wilk test for this check)

Then you have to know how many moments you will follow up your study to get your results. If you use only two moments in a group using a t-test or another test or two-sample nonparametric times, for example, a chi-square test.

Or if you have in your groups more of a continuous analysis, the ANOVA or similar nonparametric.

Finally, use the biostatistics for their results is very simple, but you have to know before anything else you want to analyze, ok?

If you need help, tell me. I hope I have contributed something.

Hugs, Livia Valentin

Hi Livia,

Abdul sent me a copy of data sets where i can see that there is no significant differences between his control and treatment........... have a look

Dear Shankar, great!!

I'm seeing with carefully the data bank and the posteriorly I send my opinion about what better the test he can be used in your study... I will send for you my opinion for you too, than should be discuss about this. What do you think?

Best regards,

Livia

Dear Livia,

i always loved doing statistics. Sure we can have a discussion on this. Also Livia can you please have a look at my reply that i gave to Abdul. I think the problem is not how to deal with data ............the problem i see is, there seem to be no correlation with glycoalkaloids and disease which is unusual (i think)........

from table number 1, even if it either inoculated with water or pathogen the values are not significantly different.

and table number 2 for just looks like plotting the disease progress over time.

The hypotheses really needs to be clear

Best regards

Shankar

Thanks Shankar, my hypothesis was to assume that disease severity will affect the Glycoalkaloids concentrations. but when tested, disease severity did not affect the glycoalkaloids. and please have a look again on the attached details

Dear Abdul Majeed,

According the data that you have presented, there is no difference between the case and control groups regarding their trend of change in the measured concentrations of Glycoalkaloid (P=0.889). Your question actually needs more complicated data analysis using time-series models. Because you want to compare two groups of data regarding their change over time. In addition, both trend charts are not very different visually (see the attachment).

I hope that this would be helpful.

Cheers

Younes, Jahangiri-Noudeh M.D.

thank you dear Noudeh, your explanation is really helpful

Dear Abdul, I assume glycoalkaloid levels are measured on a continuous scale but on what scale are you measuring the disease severity?. If that is also measured on a continuous scale then Pearson's correlation coefficient test can be applied for the data that you have collected separately on 3,6,9 and 12 days.

If the disease is measured on a categorical scale Spearman correlation coefficient test or Point biserial coefficient test may be applied. The correlation coefficient obtained reflects the direction of relationship and as well the strength of relationship between the variables.

If you plan to find out the relationship between different time periods(3,6,9 and 12days) repeated measures ANOVA can be used.

Dear Shankar,

follow the statistical calculations of your article. Now left you analyze these data!! Your sample is worth to undertake a Shapiro-Wilk test, but the number of subjects for each group allowed the realization of an ANOVA.

Good luck!

Best regards,

Livia