Problem 1
With mean = median, the data set is symmetrical. Well it's more accurate to say "roughly symmetrical" due to the phrasing "mean is approximately equal to the median". Using the median as its center is fairly common with many distributions, so there isn't an issue here.
The error is from the phrasing "the quartiles should be used to measure the spread". The quartiles themselves don't measure spread. They are simply single data points and do not tell us the spread. If Ralph said Q3 = 75 is a measure of spread, then he would be incorrect. However, the difference between the third and first quartile will help get the IQR (interquartile range), as described by this equation below
IQR = Q3 - Q1
and this is one tool to measure how spread out data is. The larger the IQR is, the more spread out the data will be.
The IQR is how far of a gap it is from Q1 to Q3. Another tool to measure spread is the standard deviation. Ralph was probably thinking about the IQR when he mentioned the quartiles, but he should be more specific about what he means.
=======================================================
Problem 2
Answer: Skewed left
This is the same as saying "negatively skewed"
---------------
Step-by-step explanation:
The main cluster of values is on the right side, while a small portion of values are to the left. We consider this the left tail and the left tail is pulled longer compared to the right tail. The further left we go, the shorter the bar and eventually we hit outliers. These much smaller outliers compared to the main cluster pull on the mean to make it smaller than it should be.
One example could be that this curve represents test scores. The majority of the class could get in the range of say 70 to 90. Then there's the unfortunate few outliers who scored much lower (possibly 40 through 60); those outlier scores pull down the mean grade of the class overall.
In short, negatively skewed data indicates that mean < median.