It seems you have two groups of values (from two experiments with a different parameter value) and you want to find the significance of the data under the assumption that the expected difference between these two groups is zero.

This significance is obtained from the t-test.

However, the result of this test can only be interpreted correctly when some more assumptions of the test make sense in the context of your data. These assumptions are: (i) independence of the values, (ii) normal distribution and (iii) variance homogeneity.

Assumption (i) means that knowing one value should not give you any extra information on any other value. This is given "by design".

Assumption (ii) and (iii) are not very critical, but if they are extremely unreasonable then the interpretation of the results of the t-test can be largely wrong. You should check if the values scatter more or less symmetrically around the means and that the standard deviations are roughly similar. As you are measureing fluorescence intensity, it is likely that the scattering is not symmetric and the standard deviation is larger for larger mean values. In this case I suggest to perform the test on the logarithms of these values (the logs will then have a symmetric distribution with similar standard deviations), the tested hypothesis is then that the expected ratio of intensity values is 1 (that is: the expected log ratio = expected difference in logs is 0).

The t-test is available in Excel and Calc and other spreadsheet programs. You can also find websites where you can run a t-test online.

Dear Magali 

The choice of the test depends on the type of your data ( scale, ordinal,.. ) and the distribution of the data ( normally distributed or not).

In the case of the continuous variable, and normally distributed data, then you can use T-test for two independents samples, but in the case of other assumptions, you can use non-parametric test ( Mann- Whitney test ).

Good Luck

I also have similar challenge to statistically tell the significance of the difference between two numbers e.g.

Parameter Point A Point B

Na (cmol/Kg) 40 70

NO3 (mg/kg) 30 16

etc.

What test/statistics can sort this out.

So for sodium, you want to determine the significant difference (if any) between point A=40 and point B=70?

The mean for point A is 40 divided by 1, which equals 40.

The standard deviation for point A is the observation (40) minus the mean (40) squared which equals 0, divided by the sample size minus one. This is zero divided by zero, and that value is undefined.

One way to think about replication is that it is a way to estimate how often we will get the same result if we repeat an activity multiple times. Without that information there is no statistics. Some might point out confusion tables, but my response is that technique does not seem to apply to Manoah's question.

One more question.

Has anyone simulated sediment in swatcup using SME as objective function. Abbaspour said it is the best OF for sediment. All papers i have read are using NS R2. Help please.

How can i check if variance is significant between two variables?

For my project, I measure placental length & weight collected from mother who were diseased& non diseased & calculate the standard deviation. But after that i'd like to know if there is significant différents in length and weight for placenta that came from diseased mother & non diseased mothers by considering the standard deviation . I know that some statisticals test exist but I don't know which one I should use. any suggestions ?

F-test (when assuming normal distributions) or Lenvene's test.

To test for the significance of a difference between two numbers, assume they are Poisson counts. Input two observed counts in the top two boxes. All the other boxes should have the value zero ('0'), and you will get a P value.

https://www.sgapeio.es/INFORMEST/VICongreso/taller/applets/Sisa/t-test.htm

Dear Friends,

I have three plant species ABC, five treatment 0,5,10,15,20. and four-date series 1,2,3,4-day. I have replicated. How can I test their significant difference? I do not need to compare species ABC. Do you know how to use R or others to test?

Huachun Zou how do we interpret the result of those Poisson analysis? what is the limit of p-value for saying that the two values is significantly different? Thank you in advance.