The mode is clear (the value with the highest frequency), so compute the mean, the mode and the standard deviation, then apply $Sk_p = (\bar{x} - Z)/\sigma$.

Setting up the table. With $N = \sum f = 3 + 7 + 12 + 6 + 2 = 30$:

Mean. $\bar{x} = \frac{\sum fx}{N} = \frac{435}{30} = 14.5.$

Mode. The greatest frequency, $12$, occurs at $x = 15$, so the mode $Z = 15$.

Standard deviation. $\sigma = \sqrt{\frac{\sum fx^2}{N} - \bar{x}^2} = \sqrt{\frac{7125}{30} - (14.5)^2} = \sqrt{237.5 - 210.25} = \sqrt{27.25} \approx 5.22.$

Coefficient. $Sk_p = \frac{\bar{x} - Z}{\sigma} = \frac{14.5 - 15}{5.22} = \frac{-0.5}{5.22} \approx -0.096.$ The value is negative but very small, so the distribution is only slightly negatively skewed — almost symmetric. …

The commonest errors are dividing $\sum fx^2$ by $N$ after subtracting $\bar{x}^2$ (the order is: divide first, then subtract the squared mean) and taking the mode as t …