class: center, middle, inverse, title-slide # Einführung in die Quantitative Datenanalyse ## Sitzung 8: Maße der zentralen Tendenz und Variabilität ### Proseminar an der Freien Universität Berlin ### 25.06.2017 - Marcus Spittler --- # <span class="red">Inhalt der 8. Sitzung</span> 1. <span class="red">Maße der zentralen Tendenz</span> - Modus - Median - (Arithmetisches) Mittel 2. <span class="blue">Maße der Variabilität</span> - Varianz - Standardabweichung - Variationsbreite - Variationskoeffizient 3. <span class="green">Form von Verteilungen</span> - Symmetrie - Wölbung --- </br></br> .pull-left[ <img src="./img/7hamilton.jpg" alt="Hamilton" style=""> ] .pull-right[ </br></br></br></br></br></br></br></br></br></br></br> Margaret Hamilton (1969) mit dem Code der Apollo 11 Mission ] --- # <span class="red">Übersicht</span> Zulässige Berechnungen ab jeweiligem Skalenniveau: | **Skalenniveau**| **Zentrale Tendenz** | **Variabilität** | |-----------------|--------------------------|---------------------------| | Nominalskala | Modus | Entropie | | Ordinalskala | Median | Summenhäufigkeitsentropie | | ab Intervallsk. | - Arithmetisches Mittel | - Standardabweichung | | | - Harmonisches Mittel | - Varianz | | | - Geometrisches Mittel | - Variationsbreite | --- # <span class="red">Arithmetisches Mittel</span> - Arithmetisches Mittel wird mit `\(\bar{x}\)` (*mean*) bezeichnet - Entspricht dem **Schwerpunkt**/Zentrum der Verteilung - Ab *intervallskaliertem* Merkmal `$$\bar{x} = \frac{1}{n} \sum_{i=1}^{n} x_{i}$$` - Empfindlich gegenüber Ausreißern/Extremwerten ```r mean( c( 800,1100,1500,2500,2800, 3000) ) ``` ``` ## [1] 1950 ``` ```r mean( c( 800,1100,1500,2500,2800, 3000, 80000) ) ``` ``` ## [1] 13100 ``` --- # <span class="red">Arithmetisches Mittel</span> - Eigenschaften des **arithmetischen Mittels** 1. Summe der Abweichungen vom Mittel ergibt immer *null* `$$\sum_{i=1}^{n} (x_{i}-\bar{x}) = 0$$` 2. Summe der quadrierten Abweichungen vom Mittel ergibt immer ein *Minimum*. Diese Eigenschaft macht man sich bei der *Methode der kleinsten Quadrate* / *Least squares* zunutze `$$\sum_{i=1}^{n} (x_{i}-\tilde{x})^2$$` --- # <span class="red">Modus</span> - **Modus/Modalwert** `\(Mo\)` (*Mode*) - Häufigster Wert einer Verteilung - Für jedes Skalenniveau geeignet - Sollten zwei verschiedene Werte die selbe Häufigkeit haben, spricht man von einer *bimodalen Verteilung* ```r icecream <- c("chocolate", "vanilla", "strawberry", "vanilla", "strawberry", "vanilla", "chocolate", "chocolate", "chocolate") table(icecream) ``` ``` ## icecream ## chocolate strawberry vanilla ## 4 2 3 ``` --- # <span class="red">Median</span> - **Median** `\(Md\)` (*median*) - Der Median einer Stichprobe von Werten ist definiert als der Wert, der größer gleich 50% der Werte der Stichprobe ist. - `\(x_{MD} = x_{(0.5)} = min \left \{ x_{i}\, |\, F(x_{i}) \geq 0.5 \right \}\)` - Kennzeichnet die *Mitte* der Stichprobenwerte - Ab *ordinalskaliertem* Merkmal - Wichtige Eigenschaft: Robust gegen Ausreißer --- # <span class="red">Berechnung des Median</span> - Erster Schritt: Anordnung der Daten nach Größe geordnet (Rangreihe) bzw. theoretisch plausibler Rangordnung - bei **ungeradem** `\(n\)` = Median `\(Md = Rangplatz (n+1)/2\)` - bei **quantitativen** Merkmalen und **geradem** `\(n\)` hier definiert als das arithmetische Mittel zwischen oberem und unterem Rangplatz: `\(Md = x_{bar}\)` von `\(x_{Rangplatz:n/2}\)` und `\(x_{Rangplatz:(n/2)+1}\)` - bei **ordinalskalierten** Merkmalen und **geradem** `\(n\)` ist der MD der untere Rangplatz: `\(Md = x_{Rangplatz:n/2}\)` - Die Berechnung des Medians bei **geradem** `\(n\)` ist uneindeutig, hier ist Vorschlag präsentiert, es gibt jedoch verschiedene Methoden. --- # <span class="red">Median für ordinale Merkmale</span> - Beispiel: Wir haben 14 Menschen in einem Fast Food Restaurant bei ihrer Bestelllung beobachtet. Dabei haben wir erhoben, welche Größe das von ihnen bestellte Menü hatte. - Die Größe des Menus haben wir in der Reihenfolge der Bestellungen notiert, z.B.: ```r menus <- c("sehr klein", "groß", "groß", "groß", "sehr groß", "mittel", "sehr klein", usw. ... ) ``` -- - Die Größe der Menus ist ein **ordinalskaliertes** Merkmal. Die Anzahl der beobachteten Personen `\(n\)` ist mit 14 **gerade**. --- # <span class="red">Median für ordinale Merkmale</span> Zuerst legen wir wie gewohnt einen neuen Vektor mit den Daten an ```r menus <- c(rep("sehr klein",3), rep("groß", 5), "mittel", rep("klein",4), "sehr groß") table(menus) ``` ``` ## menus ## groß klein mittel sehr groß sehr klein ## 5 4 1 1 3 ``` Danach "ordnen" wir den Vektor um für R die Reihenfolge festzulegen ```r menus <- ordered(menus, levels = c("sehr klein", "klein", "mittel", "groß", "sehr groß")) table(menus) ``` ``` ## menus ## sehr klein klein mittel groß sehr groß ## 3 4 1 5 1 ``` --- # <span class="red">Median für ordinale Merkmale</span> ```r # Package "DescTools" für die Berechnung des Median # bei ordinalskalierten Variablen library(DescTools) Median(menus) ``` ``` ## [1] klein ## Levels: sehr klein < klein < mittel < groß < sehr groß ``` Der Median unserer Variable `menus` liegt bei "klein". Das heißt, die unteren 50% der Besteller haben ein "kleines" oder "sehr kleines" Menu bestellt. --- # <span class="red">Median bei Intervallskala</span> ```r income <- c( 800,1100,1500,2500,2800, 3000, 80000) summary(income) ``` ``` ## Min. 1st Qu. Median Mean 3rd Qu. Max. ## 800 1300 2500 13100 2900 80000 ``` ```r median(income) ``` ``` ## [1] 2500 ``` ```r mean(income) ``` ``` ## [1] 13100 ``` --- # <span class="blue">Maße der Variabilität</span> - Während Maße der zentralen Tendenz uns Auskunft über die Mitte, bzw. das Zentrum der Werte liefern, informieren uns Maße der Variablität über die **Unterschiedlichkeit** der Werte. --- # <span class="blue">Varianz und Std.Abweichung</span> - **(Stichproben-) Varianz** `\(s^2\)` (*variance*) - Def.: Die Stichprobenvarianz ist die Summe der quadrierten Abweichungen aller Messwerte vom arithmetischen Mittel, dividiert durch `\(n-1\)`. - `\(n-1\)` bezeichnet man als **Freiheitsgrade** - Ab *intervallskaliertem* Merkmal `$$~\sigma^2 = \frac{\displaystyle\sum_{i=1}^{n}(x_i - \bar{x})^2} {n-1}$$` - **Standardabweichung** `\(s\)` (*standard deviation*) - Da die Varianz quadiert ist, ist sie nur schwer inhaltlich interpretierbar `$$~\sigma = \sqrt{\sigma^2}$$` --- ## <span class="blue">Standardabweichung</span> Beispiel für zwei Verteilungen mit dem gleichen Mittelwert, aber unterschiedlichen Standardabweichungen: <img src="./img/7sd2.svg" alt="" style="max-width:100%"> --- # <span class="blue">Varianz Beispiel</span> ![](data:image/png;base64,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)<!-- --> --- # <span class="blue">Varianz Beispiel</span> | ID `\(i\)` | Note `\(x_{i}\)` | 1. Schritt `\(x_{i}-\bar{x}\)` | 2. Schritt `\((x_{i}-\bar{x})^2\)` | |--------|-------------:|----------------------------:|--------------------------------:| | 1 | 3.3| 0.8| 0.64| | 2 | 1.7| -0.8| 0.64| | 3 | 2.0| -0.5| 0.25| | 4 | 4.0| 1.5| 2.25| | 5 | 1.0| -1.5| 2.25| | 6 | 2.0| -0.5| 0.25| | 7 | 3.0| 0.5| 0.25| | 8 | 2.7| 0.2| 0.04| | 9 | 4.0| 1.5| 2.25| --- # <span class="blue">Varianz Beispiel</span> ```r # Mittelwert mean(grades) ``` ``` ## [1] 2.5 ``` ```r # Varianz var(grades) ``` ``` ## [1] 0.8672727 ``` ```r # Standardabweichung sd(grades) ``` ``` ## [1] 0.9312748 ``` --- ## <span class="blue">Standardabweichung</span> Standardabweichung in der Gaussschen Normalverteilung <img src="./img/7gaussd.svg" alt="" style="width:180%"> --- # <span class="blue">Variationsbreite</span> - **Variationsbreite** (*range*) - Differenz aus dem größten und kleinsten Messwert `$$x_{n}-x_{1}$$` ```r range(grades) ``` ``` ## [1] 1 4 ``` ```r range(income) ``` ``` ## [1] 800 80000 ``` --- # <span class="blue">Interquartilabstand</span> - **Interquartilabstand** (*Hinge / IQR*) - Auch Tukey-Angelpunkte - Drückt die Länge jenes Bereichs aus, über den die mittleren 50% der Verteilung streuen. - Berechnet sich analog zum Median. - `\(Q_{1}\)` ist der *untere Angelpunkt* unterhalb dem 25% der Verteilung liegen. `$$IQR = Q_{3}-Q_{1}$$` --- ### Boxplot <img src="./img/5graf-boxplotex.png" alt="Boxplot" style="max-width:75%"> <small>[Weitere Erläuterung](https://flowingdata.com/2008/02/15/how-to-read-and-use-a-box-and-whisker-plot/ )</small> --- #<span class="green">Form von Verteilungen</span> - **Verteilungen** können: - uni- oder bimodal - symmetrisch oder schief (*skewness*) - spezielle Funktionen sein (z.B. Normalverteilung) --- ### Form von Verteilung <img src="./img/7distribution1.png" alt="Verteilung" style="max-width:75%"> <img src="./img/7distribution2.png" alt="Verteilung" style="max-width:75%"> --- ### Form von Verteilung <img src="./img/7distribution3.png" alt="Verteilung" style="max-width:75%"> <img src="./img/7distribution4.png" alt="Verteilung" style="max-width:75%"> --- ### Form von Verteilung <img src="./img/7distribution5.png" alt="Verteilung" style="max-width:75%"> --- ## Vielen Dank für die Aufmerksamkeit <iframe src="https://giphy.com/embed/26ufp2yCvgElWaX9S" width="422" height="480" frameBorder="0" class="giphy-embed" allowFullScreen></iframe>