For example, measuring lymphocytes conc. in control gp was 100 ug/ ml while group 1 average was 115 ug/m, performing student t test results were signficantly different. But how to explain this in terms of fold change?
Is expressing the results in terms of fold change a must to prove that there was an improvement (increase) in the group than the control?
The Wikipedia article describes how there is some ambiguity in the term.
https://en.wikipedia.org/wiki/Fold_change
This ResearchGate thread may be of use to you. Note that log2 fold change is common in some fields.
https://www.researchgate.net/post/How_to_calculate_the_log2_fold_change
I think the basic concept of fold is that you would divide the Group average by the Control average.
So, if Control were 100 and Group were 200, that would be 2-fold. (200 / 100). The ambiguity comes in where in common language I would say the increase was 1-fold, but I may be in the minority there. I also find it a little odd to call this a fold change of 2, but my understanding is that's how the term is used.
[Edit: I'm running a little poll of facebook friends. So far the results are evenly split between calling this a 2-fold increase and a 1-fold increase. So, in any case, it's important to note that lay people's interpretation of this language isn't uniform.]
So, my advice: 1) Be clear if you are using log2 fold change or simple fold change. 2) Be careful with your language about change or increase in reporting results, especially if your audience may not be familiar with fold change calculations.
If I understand the concept correctly, for your example:
C = 100
M = 115
M/C
### 1.15
log2(M/C)
### 0.202
A couple of other simple resources:
http://rstudio-pubs-static.s3.amazonaws.com/13988_bb11d85b79b2436280de434988558140.html
https://www.hunker.com/13414935/how-to-calculate-fold
115 ug/ml / 100 ug/ml = 1.15
But note that the t-test is about the difference, 115 ug/ml - 100 ug/ml = 15 ug/ml, and not abot the fold-change. You can use the t-test to test a fold-change by tesing the difference between the logarithms.
Another question:
Is it a must to express the results in terms of fold change to prove that there was an improvement (increase) in the group than the control?
Well, there are different ways to show that the Group mean is higher. You could report the means. You could plot the means. You could state that the Group mean was higher.
There is no prove. The if t-test is "significant", you conclude that you have enough data to interpret the sign of the mean difference between the groups. In that case you would in some way report that sign. The typically recommended way is to write something like "the mean lymphocyte concentration in group 1 was higher than in the controls (t-test: p
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics