Need help as a biostatistics moron! Any help appreciated!

[kafandiyev]

Hello I am completely clueless in statistics and I need help for a problem...

The problem is about two study groups regarding the lesion size at the lip and the time of admission since the onset of the lesion.

I divided the group into two categories: those who have T1(0-2cm) size lesions (20 paitents) and those who have T2(2-4cm) size lesions (8 patients).

I wrote down the size of the lesion and corresponding time of the admission to the hospital for each patient in each group (T1 and T2).

According to average calculations
1- the mean size of the lesions in the T1 group is 1.16 cm2 and the average admission time is 6.2 months. For the T2 group the average size is 5.06 and the mean admission time is 9.9 months.

It is obvious that those who admit late have bigger tumors.

My question is:
what is the best test to calculate the reliability of this correlation?
What program can use for that? I have only excel...
And what are the names of statistical calculations I have to make?

Here are the two groups:
T1:
Admission time Size of the lesion
6 0.25
7 0.5
12 0.5
3 0.5
6 1
6 1
6 1
7 1
4 1
3 1
4 1
6 1
15 1.5
5 1.5
12 1.5
4 1.5
2 1.5
12 2
3 2
1 2
averages
6.2 1.1625

T2:
5 2.5
12 3
2 3
12 6
9 6
9 4
6 8
24 8
averages
9.875 5.0625


I will appreciate any help...

Thank you very much in advance

 

[giordano]

I wouldn't divide in two groups. If you have for each canditate an admission time and a lesion size you can do (with Excel) a graph with time as the x-axis and size as y-axis. You can easily fit a linear regression and get R^2 (percentage of variance explained or coefficient of determination). With your number I get R^2 of 0.23 and a P-Value of 0.012 (you can do that all with excel). Obviously, there is a relation. What I do not understand: how you can find out the time of admission? But this what I told you is only a part of the statistics. You should do a residual analysis, you can look at the graph and at the residual plot for outliers and variances homogenity, etc.

But first, I would read a textbook on statistics. I would suggest Glantz: Primer of Biostatitics. Or Faraway: Linear Models with R.