The three classes of measurement error.

(a) Gross errors ✓. Human blunders — misreading a scale, transposing digits when recording, choosing the wrong range, computing wrongly. They arise from carelessness or inexperience, exactly as the statement says, and they are usually large and obvious.

(b) Systematic errors ✓. These come from the equipment and the environment: friction in a movement, worn bearings, a zero offset, calibration drift, loading of the circuit by the instrument, or the effect of temperature and stray fields. They are divided into instrumental, environmental and observational errors, and their hallmark is that they push every reading the same way by roughly the same amount.

(c) Random errors ✓. What remains after the other two are removed — the small, unpredictable scatter from causes too numerous and too slight to identify individually. They follow a normal distribution about the true value and are treated statistically.

(d) — "Gross and systematic errors are non-correctable." FALSE. This is the one wrong statement, and it inverts the most useful fact in the topic. Both of those classes can be dealt with: gross errors by taking several readings, by careful technique and by a second observer; systematic errors by calibrating against a standard, by applying correction factors, by re-zeroing, or by choosing an instrument whose loading is negligible. It is the random error that cannot be eliminated — only reduced, by averaging n readings, since

\(\sigma_{mean}=\dfrac{\sigma}{\sqrt{n}}\)

So the correct statements are (a), (b) and (c) — option 4.

The practical consequence. Systematic error limits accuracy (closeness to the true value) while random error limits precision (repeatability), which is why calibration and averaging are two different remedies for two different problems.

Hence, the correct answer is (a), (b) and (c) only.