Is it necessay to go into depth about what a specific QQ plot tells us, for example the starting point is to reject the hypothesis of normality based on ANY shape that isn't a straight line. But are we required to be able to know which QQ plot represents which kind of skewness and the thinkness of tail?
Hi, perhaps a past exam question might help show the level of detail required: April 2011 question asked: 3 A model has been developed for a financial time series. (i) Describe the construction of a QQ plot to judge the goodness of fit of the model. [3] (ii) Explain how to interpret a QQ plot. [3] The interpretation in (ii) required some high-level description of what to look for (ie a straight line vs āSā-shaped or arched, indicating that one of the distributions has a greater degree of skewness, or that one of the distributions has a higher degree of kurtosis)