Clinical pearl for stats MCQs: when two data sets have different means or units, only a relative measure of spread is comparable, and that measure is the coefficient of variation.
The CV is computed as $CV = \dfrac{SD}{\text{mean}} \times 100$. Dividing the SD by the mean strips away the units and the scale, so a CV of, say, $15\%$ in one group can be set against $15\%$ in another even if the raw values are wildly different.
The distractors are all absolute. Variance ($a$) is SD squared and carries squared units. Standard deviation ($d$) is spread in the original units, so it cannot be compared across different scales. Standard error of mean ($c$) measures how precisely the sample estimates the population mean, not the spread comparison asked for.
Hence the answer is coefficient of variation, option $b$.
Ref: Fundamentals of Biostatistics, 7th ed., pp. 20-21.