Standard Error and confidence interval

Hello,

I have a 16x31 matrix, DeltaAnglesAdj, and I calculated the Standard error, columnwise:

std_err_mean = std(DeltaAnglesAdj,[],1)/sqrt(size(DeltaAnglesAdj,1)); (1x31 vector)

Then I calculated the P95:

P95 = tinv(0.975, size(DeltaAnglesAdj,1)-1);

How can I calculate the Confidence interval?

when I applied the method found on previous questions on Mathworks, it gives me error:

mean_data = mean(DeltaAnglesAdj,1); (1x31 vector)

CI95 = mean_data + std_err_mean*[-1 1]*P95;

Error using *

Incorrect dimensions for matrix multiplication. Check that the number of

columns in the first matrix matches the number of rows in the second matrix.

To perform elementwise multiplication, use '.*'.

The error persists also after following the instruction:

CI95 = mean_data + std_err_mean*.[-1 1]*.P95;

↑

Error: Invalid use of operator.

How can I solve this?

Answer 1

[Star Strider]

The actual problem is that the vectors you are working with are row vectors. To create ‘CI95’ as a (2x31) matrix, you need to code it a bit differently.

Try this:

CI95 = mean_data + std_err_mean.*[-1; 1]*P95;

Then, to plot the mean and confidence intervals:

figure
plot((1:31), mean_data,  (1:31), CI95)

.