# How to compare these two sets of data?

1. Apr 21, 2014

### Chemist125

Hi

I have two sets of data, y as a function of x, for two different experiments.

Y is the dependant variable, an instrument reading, x is the independant variable, a concentration of something in a solution.

Plotting the data in excel gives me 2 curves.

The experiments were essiantially the same, but the results are slightly different, I want to compare how much 'error' there is by comparing how close the data/curves from each experiment lie to eachother. Is there some way I can do this mathematically?

Chem125

2. Apr 21, 2014

### Staff: Mentor

There are multiple tests, which test is appropriate here depends on the data.

3. Apr 21, 2014

### Chemist125

I attached an excel sheet with example data.

What do you think?

Thanks

#### Attached Files:

• ###### data.xlsx
File size:
10.4 KB
Views:
77
4. Apr 22, 2014

### Staff: Mentor

- the differences between the instruments are way smaller than the differences between the concentrations
- there is no repetition test where the same instrument was tested with the same concentration multiple times, therefore it is impossible to determine the accurary of the measurements based on that data alone. That would be really interesting.
- there is no clear systematic difference between the two instruments

The data is not sufficient to draw more specific conclusions, independent of the analysis method.

5. Apr 23, 2014

### Chemist125

That data shows a repetition test with the same concentrations.... ...on the same instrument.

Sorry that I did not make this clear.

6. Apr 23, 2014

### Staff: Mentor

Ah okay.
If we can assume the deviations are independent of the concentration, and there is no systematic difference between the two sets of measurements (unclear, but we don't have data to test this with any relevant precision), you can calculate all differences and find the standard deviation of them. This should be related to measurement uncertainties in some way. You would need more data to tell more.

7. Apr 23, 2014

### Chemist125

Yes, we can make those assumptions.

What do you mean calculate the differences?

8. Apr 23, 2014

### Staff: Mentor

For concentration 1, take "measured value 1" - "measured value 2", repeat for all other concentrations. Something that would be 0 in the ideal case, if I understood your data.

9. Apr 23, 2014

### Chemist125

Oh, I see what you mean. What extra data would I need to tell more?

10. Apr 23, 2014

### Staff: Mentor

More repetitions would certainly help.

Some uncertainties or other issues could come from details of your setup, which is impossible to estimate without knowing the setup.