3.11
Milieu de gamme
Une estimation quantitative relativement simple à calculer de la tendance centrale d’un ensemble de données est son milieu de gamme, qui est défini co…
Le milieu de gamme est l’une des mesures de la tendance centrale. C’est la valeur à mi-chemin entre les deux valeurs extrêmes et est généralement défini comme la moyenne arithmétique des valeurs de données maximale et minimale.
Dans cet exemple de jeu de données sur le temps de sommeil des nourrissons, le milieu peut être calculé en additionnant le nombre maximum et minimum d’heures et en divisant la somme par deux.
Bien que le milieu de gamme soit relativement facile à calculer, il est rarement utilisé en statistiques car il ignore toutes les valeurs de données intermédiaires et manque de robustesse dans la mesure.
Le médium est également sensible aux valeurs extrêmes. Dans cet exemple, une modification du nombre maximal ou minimal d’heures de sommeil peut modifier le milieu de gamme. De plus, le milieu de gamme ne peut pas être utilisé pour des données catégorielles telles que les rangs ou les étiquettes.
Le milieu de gamme est complémentaire à la plage ou à la différence entre les valeurs maximale et minimale. Par exemple, en connaissant la valeur moyenne et la plage de données, on peut calculer les valeurs maximale et minimale dans cet ensemble de données.
View the full transcript and gain access to JoVE Core videos
Q1: How is the midrange calculated from a dataset?
The midrange is calculated by adding the maximum and minimum values in a dataset and dividing the sum by two. This arithmetic mean of the extreme values provides a quick quantitative estimate of central tendency. For example, if infant sleep times range from 8 to 16 hours, the midrange would be (8 + 16) / 2 = 12 hours.
Q2: Why is the midrange rarely used in statistical analysis?
The midrange ignores all intermediate data values and lacks robustness in measurement. It is highly sensitive to extreme values, meaning a single change in the maximum or minimum can significantly alter the result. Additionally, the midrange cannot be applied to categorical data such as ranks or labels, limiting its practical utility.
Q3: What is the relationship between midrange and data range?
The midrange is complementary to the range, which is the difference between maximum and minimum values. By knowing both the midrange and the data range, you can compute the actual maximum and minimum values in a dataset. The midrange is essentially half of the data set's range.
Q4: How does the midrange compare to other measures of central tendency?
Like the arithmetic mean, the midrange is sensitive to extreme values and outliers. However, unlike the arithmetic mean, it ignores all intermediate values and does not account for the distribution of data points. This makes the midrange less reliable for representing overall central tendency in most datasets.
Q5: When is the midrange useful as a statistical measure?
The midrange is useful in relatively fluctuation-free datasets where it can easily provide a quick estimate of central tendency. Its simplicity makes it valuable for rapid calculations when data lacks significant outliers or extreme variations. However, for most real-world applications with variable data, more robust measures are preferred.
Q6: Can the midrange be used with all types of data?
No, the midrange cannot be used with categorical data such as ranks or labels. It is exclusively a quantitative measure designed for numerical datasets. This limitation further restricts its application compared to other central tendency measures that may accommodate different data types.
Q7: What makes the midrange prone to outliers?
The midrange depends entirely on the maximum and minimum values, making it extremely vulnerable to outliers. A single extreme value can substantially shift the midrange, even if all other data points cluster closely together. This sensitivity to extreme values is why the midrange is considered less reliable for representing typical data behavior.