I am an engineering student and as part of an undergraduate research, I created a device that measures a certain value. However, there is a commercial device that measures that same value.

Here's what I want: I want to measure if the values my device measures are similar enough to those values of obtained by the commercial device.

So far, I've looked at a t-test. However, most of the tutorials I see have some sort of causality between the two data sets. What I mean by this is that in most tutorials I see, data set A is taken before some sort of intervention, and data set B is taken after the intervention. In my problem, data from my device can be taken independently from data from the commercial device.

Second, the definition of t-test from Wikipedia is that "It is used to determine whether two sets of data are significantly different from each other." However, what I want is to measure if my two data sets are comparably similar to each other.

So basically, here are my questions.

1. Is the causality I mentioned necessary for a t-test?
2. Is a t-test even the proper statistical analysis for this kind of problem?
3. If it is, how can I make it so that it will measure similarity, not difference. If it is not, could you point me to the right direction?

The sample size is around 20-ish, if that is a relevant detail to this question.