3.12
根据数据集计算出的集中趋势度量可能无法揭示有关数据的内在分布。如果将数据集绘制成图,则平均值和中位数可能会有所不同,而且图可能会在集中趋势的一侧有更多的值。这样的数据集则被称为偏向于这一侧。
图中某一侧的尾部越长,倾斜程度就越大。数据集中数值的偏度表明了,集中趋势的度量有些粗糙,忽略了更加细微的细节…
均值、中位数和众数之间的比较可提供有关数据分布情况的信息。
在此示例图表中,图的左侧是右侧的镜像。这种情况被称为数据的对称分布或正态分布。
在这种正态分布图中,均值、中位数和众数位于由虚线指示的相同位置。
假设图形的左侧和右侧不相同,则会导致分布出现偏态。此时,均值、中位数和众数不相等,并反映出数据集中不同的数值特征。
偏度表示异常值的存在。例如,在本例中,异常值位于图形的右侧。
偏度常被用于投资决策。投资模型收益的偏度反映了该投资是会产生频繁的小幅盈利和少数巨大的亏损,还是频繁亏损但偶尔获得较大收益。
View the full transcript and gain access to JoVE Core videos
Q1: What is the difference between a symmetrical and skewed distribution?
In a symmetrical distribution, the left and right sides of the graph mirror each other, and the mean, median, and mode occupy the same position. In a skewed distribution, the left and right sides are unequal, causing the mean, median, and mode to differ and reflect different values in the data set. Skewness reveals asymmetry that central tendency measures alone cannot capture.
Q2: How does skewness indicate the presence of outliers in data?
Skewness reveals whether outliers exist and on which side of the distribution they cluster. When outliers are present on one side, they pull the mean away from the median and mode, creating an asymmetric tail. The longer the tail extends, the more skewed the distribution and the more extreme the outliers.
Q3: Why do mean, median, and mode values differ in skewed data sets?
In skewed distributions, outliers on one side pull the mean toward the tail, while the median and mode remain closer to the bulk of the data. This separation occurs because the mean is sensitive to extreme values, whereas the median and mode reflect the central clustering of observations. The asymmetry prevents these measures from aligning.
Q4: How is skewness used in investment decision-making?
Skewness in investment returns indicates whether an investment model produces frequent smaller gains with few huge losses, or frequent losses with occasional large wins. Analyzing return distribution skewness helps investors understand the probability and magnitude of gains versus losses, enabling more informed portfolio decisions based on risk tolerance and expected outcomes.
Q5: What does income distribution skewness reveal about economic inequality?
Income distribution skewness exposes inequality that average income alone cannot show. When a few wealthy individuals earn substantially more while the majority earns far less, the distribution becomes skewed, with a long tail on the high-income side. This skewness demonstrates that central tendency measures provide incomplete economic insight.
Q6: Why are measures of central tendency considered crude for skewed data?
Measures of central tendency like the mean and median miss finer distributional details in skewed data sets. They fail to reveal whether values cluster on one side or how extreme outliers are. Skewness analysis provides the additional context needed to understand data asymmetry and distribution shape that central tendency alone cannot convey.
Q7: What does the length of a distribution's tail indicate about skewness?
The longer the tail of a distribution extends on one side, the greater the skewness. A pronounced tail indicates more extreme outliers pulling the mean away from the median and mode. Tail length directly correlates with the degree of asymmetry in the data set and the strength of skewness present.