Step 1: Understanding the Concept:
This question checks the real working principle of a thermocouple, and asks us to tell it apart from the working principle of other temperature sensors such as RTDs, and from unrelated electrical heating effects such as Joule heating.
Step 2: Key Formula or Approach:
A thermocouple is a self generating sensor: joining two different metals at a junction produces a small EMF whenever that junction sits at a different temperature from the reference end. This is the Seebeck effect, $E = f(\Delta T)$, where $f$ is a material dependent function of the junction temperature difference $\Delta T$ that is generally polynomial, not a fixed proportionality constant.
Step 3: Detailed Explanation:
Go through the four statements using this working principle. Statement (A) says the device is based on the Seebeck Effect, which is exactly the mechanism above, so it is correct. Statement (B) says a change in resistance is measured; that description belongs to a resistance temperature detector or a thermistor, both of which need an external excitation current to sense a resistance change, while a thermocouple needs no excitation and outputs a voltage directly, so (B) is wrong. Statement (C) claims the junction voltage is linearly proportional to the temperature difference; in practice the Seebeck EMF of standard thermocouples follows a multi term polynomial curve over their operating range, which is exactly why reference lookup tables, not a single slope, are used to convert EMF to temperature, so the strict linearity claim in (C) is false. Statement (D) attributes the effect to Joule heating, which is $I^2R$ dissipation from current flowing through a resistor; a thermocouple carries no such heating current to generate its signal, its output comes purely from the thermoelectric junction effect, so (D) is also false.
Step 4: Final Answer:
Only statement (A) is correct: a thermocouple works on the Seebeck Effect.
\[ \boxed{\text{Only (A) is correct}} \]