WM_OR.qmd

---
title: "Wissenschaftliche Methodik"
subtitle: "Repetitorium (Zusatzveranstaltung)"
format: 
  revealjs:
    theme: fom-dls.scss
    logo: DLS-Logo.png
    footer: "WM | Karsten Lübke"
    chalkboard: true
---

## Zur Erinnerung

📛 Schalten Sie bitte die Kamera an.

✏ Arbeiten Sie aktiv mit.

🤷 Stellen Sie Fragen.

💪 [https://tweedback.de/zk48/](https://tweedback.de/zk48/)


## Freitag Abend?


::: {.columns}

:::: {.column width="50%"}
**A**:

<iframe src="https://giphy.com/embed/KzDqC8LvVC4lshCcGK" width="480" height="271" frameBorder="0" class="giphy-embed" allowFullScreen></iframe>

::::: aside
[via GIPHY](https://giphy.com/gifs/floridageorgialine-smooth-florida-georgia-line-KzDqC8LvVC4lshCcGK)
:::::
::::



:::: {.column width="50%"}
**B**:

<iframe src="https://giphy.com/embed/ZOrTflIqFmP0k" width="480" height="271" frameBorder="0" class="giphy-embed" allowFullScreen></iframe>

::::: aside
[via GIPHY](https://giphy.com/gifs/just-sayin-ZOrTflIqFmP0k)
:::::
::::
:::

## UN Ziel 10: Reduced inequalities

::: center

<iframe src="https://upgrader.gapminder.org/q/71/embed?hasBorder=true&hasOpenSansFont=false" title="81% of people get this question wrong" width="100%" height="500" style="border:none;"></iframe>
:::

## Heutiges Thema &#127979;

- Wiederholung

::: {.center}
<iframe src="https://giphy.com/embed/1iTH1WIUjM0VATSw" width="480" height="270" frameBorder="0" class="giphy-embed" allowFullScreen></iframe>



Übung macht die Meister:in!
:::

::: {.aside}
[via GIPHY](https://giphy.com/gifs/rockymovie-movie-rocky-sylvester-stallone-1iTH1WIUjM0VATSw)
:::

## Wiederholung: Inferenz

- Der uns interessierende Wert in der **(Ziel-)Population** wird [(wahrer) Wert des Parameter]{.blue} genannt: $\color{blue}{\beta, \mu, \pi, \ldots}$

- Der auf Basis einer **Stichprobe** berechnete Wert wird [Schätzwert]{.green} genannt. $\color{green}{\hat{\beta}, \hat{\mu}=\bar{y}, \hat{\pi}=p,\ldots}$


$$\color{green}{Schätzwert} = \color{blue}{Wert\,Parameter} + Verzerrung + Rauschen$$

::: {.aside}
[Bueno de Mesquita und Fowler (2021)](https://press.princeton.edu/books/paperback/9780691214351/thinking-clearly-with-data): Estimate = Estimand + Bias + Noise
:::

## Punktschätzung -- Unverzerrt

::: {.columns}

:::: {.column width="50%"}
![](img/Meme_Unsicherheit.jpg){fig-align="center" width="75%"}
::::

:::: {.column width="50%"}
![](img/Estimate.gif){fig-align="center" width="100%"}
::::
:::

## Punktschätzung -- Verzerrt

::: {.columns}

:::: {.column width="50%"}
<br>
![](img/Meme_Bias.jpg){fig-align="center" width="95%"}
::::

:::: {.column width="50%"}
![](img/Estimate-Biased.gif){fig-align="center" width="100%"}
::::
:::

## Bootstrapping

![](img/Schema_Bootstrap.png){fig-align="center" width="85%"}

::: {.aside}
Quelle: Allen Downey: [tinyurl.com/inference3](https://tinyurl.com/inference3)
:::

## Konfidenzintervall

::: {.columns}

:::: {.column width="50%"}
![](img/Meme_Konfidenzintervall.jpg){fig-align="center" width="75%"}
::::

:::: {.column width="50%"}
![](img/Confidence.gif){fig-align="center" width="100%"}
::::
:::

## Simulation $H_0$

![](img/Schema_Simulation.png){fig-align="center" width="85%"}

::: {.aside}
Quelle: Allen Downey: [tinyurl.com/inference3](https://tinyurl.com/inference3)
:::

## p-Wert (I/II)

::: {.columns}

:::: {.column width="50%"}
<br>
![](img/Meme_pWert.jpg){fig-align="center" width="75%"}
::::

:::: {.column width="50%"}
![](img/P0.gif){fig-align="center" width="100%"}
::::
:::

## p-Wert (II/II)

::: {.columns}

:::: {.column width="50%"}
![](img/Meme_H0verwerfen.jpg){fig-align="center" width="75%"}
::::

:::: {.column width="50%"}
![](img/P1.gif){fig-align="center" width="100%"}
::::
:::

## Fehlerarten

::: {.columns}

:::: {.column width="50%"}
![](img/Meme_Fehlerarten.jpg){fig-align="center" width="75%"}
::::

:::: {.column width="50%"}
![](img/Meme_H0akzeptieren.jpg){fig-align="center" width="80%"}
::::
:::