The dependent and independent variables are assumed to follow normal distribution. Furthermore, in the case of multiple regression, this assumption is also applicable to the combination of variables.
If their distributions vary substantially from the normal distribution, the F test, which assesses the overall regression, and the t-test, which assesses statistical significance of the coefficients, may no longer remain valid. These tests assume that the residuals are distributed normally.
Univariate normality can be visually examined through the histogram and the stem & leaf plot (Exhibit 33.27), though this may be problematic for smaller samples. The normal probability plot, Exhibit 33.28, is a more reliable check. This compares the cumulative distribution of the data with that for a normal distribution. If data is normal, it is scattered close to the diagonal which represents the theoretical normal distribution.
Additionally, the Kurtosis statistic tests for “peakedness” or “flatness” of the data, and the skewness statistic is a measure for skewness.
Use the Search Bar to find content on MarketingMind.
In an analytics-driven business environment, this analytics-centred consumer marketing workshop is tailored to the needs of consumer analysts, marketing researchers, brand managers, category managers and seasoned marketing and retailing professionals.
Unlock the Power of Digital Marketing: Join us for an immersive online experience designed to empower you with the skills and knowledge needed to excel in the dynamic world of digital marketing. In just three days, you will transform into a proficient digital marketer, equipped to craft and implement successful online strategies.