The Effect
Huntington-Klein, Nick. The
effect: An introduction to research design and causality. Chapman and Hall/CRC, 2021.
Hoofdstukken 1, 2, 3, 4 en 13 uitgelegd vanaf scratch
Handig voor de pre master aan de:
Inhoudsopgave
Inleiding............................................................................................................ 4
Hoofdstuk 1: Onderzoek ontwerpen....................................................................5
1.1 De grote vraag: hoe werkt de wereld?........................................................................5
1.2 Empirisch onderzoek..................................................................................................... 5
1.3 Waarom een onderzoeksontwerp nodig is.....................................................................5
1.4 Wat behandelt The Effect?.............................................................................................6
Cheat sheet hoofdstuk 1...................................................................................................... 6
Hoofdstuk 2: Onderzoeksvragen.........................................................................7
2.1 Wat is een onderzoeksvraag?........................................................................................ 7
Beantwoordbaar zijn......................................................................................................... 7
Bijdragen aan begrip van de wereld.................................................................................7
2.2 Van theorie naar hypothese..........................................................................................7
2.3 Waarom met een vraag beginnen?................................................................................8
Wat datamining wel goed doet......................................................................................... 8
Wat datamining slecht doet.............................................................................................. 8
False positives.................................................................................................................. 8
2.4 Bronnen voor onderzoeksvragen...................................................................................9
2.5 Checklist: is je vraag goed?........................................................................................... 9
Cheat sheet hoofdstuk 2...................................................................................................... 9
Pagina 1
, The Effect, samengevat
Hoofdstuk 3: Variabelen beschrijven.................................................................10
3.1 Soorten variabelen...................................................................................................... 10
Continue variabelen....................................................................................................... 10
Telvariabelen (count)..................................................................................................... 10
Ordinale variabelen........................................................................................................ 10
Categorische variabelen.................................................................................................10
Kwalitatieve variabelen..................................................................................................10
3.2 Distributie: hoe vaak komt elke waarde voor?.............................................................10
Voor categorische en ordinale variabelen......................................................................11
Voor continue variabelen................................................................................................11
3.3 Kerngetallen: gemiddelde, mediaan, percentielen......................................................11
Gemiddelde.................................................................................................................... 11
Mediaan.......................................................................................................................... 11
Percentielen................................................................................................................... 11
3.4 Spreiding: bereik, IQR, variantie, standaarddeviatie....................................................11
3.5 Skewness en log transformatie....................................................................................12
3.6 De normale verdeling.................................................................................................. 12
3.7 Symbolen die je vaak ziet............................................................................................ 12
Cheat sheet hoofdstuk 3.................................................................................................... 13
Hoofdstuk 4: Relaties beschrijven.....................................................................14
4.1 Wat is een relatie?....................................................................................................... 14
Scatterplot...................................................................................................................... 14
4.2 Conditionele distributies.............................................................................................. 14
4.3 Conditioneel gemiddelde............................................................................................. 14
4.4 Lijn fitten en OLS......................................................................................................... 15
Ordinary Least Squares (OLS).........................................................................................15
Covariantie en correlatie................................................................................................ 15
Niet altijd recht............................................................................................................... 16
4.5 Conditional conditional means: controleren voor een variabele..................................16
IJsjes en korte broeken................................................................................................... 16
Cheat sheet hoofdstuk 4.................................................................................................... 17
Hoofdstuk 13: Regressie...................................................................................18
13.1 De basis van regressie............................................................................................... 18
Snelle herhaling uit hoofdstuk 4.....................................................................................18
13.1.1 Foutterm en residu.............................................................................................. 18
13.1.2 Aannames en sampling variation........................................................................18
13.1.3 Hypothesetoetsing..............................................................................................19
Belangrijke nuance: significant is niet hetzelfde als belangrijk.......................................20
13.1.4 Regressietabellen lezen.......................................................................................20
Andere model statistieken..............................................................................................20
13.2 Verder met regressie................................................................................................. 20
13.2.1 Discrete variabelen: dummies.............................................................................20
Categorisch met meer dan twee niveaus.......................................................................21
13.2.2 Polynomen........................................................................................................... 21
Interpretatie van een polynoommodel...........................................................................21
Pagina 2
, The Effect, samengevat
Hoeveel polynomiale termen?........................................................................................22
13.2.3 Variabele transformaties.....................................................................................22
13.2.4 Interactietermen.................................................................................................. 22
Voorbeeld met binaire moderator...................................................................................23
Waarschuwingen bij interacties......................................................................................23
13.2.5 Niet lineaire regressie (logit, probit)....................................................................23
13.3 Standaardfouten kloppen vaak niet...........................................................................23
13.3.1 Heteroskedasticiteit............................................................................................. 24
13.3.2 Autocorrelatie...................................................................................................... 24
Geclusterde standaardfouten.........................................................................................24
Bootstrap standaardfouten.............................................................................................24
13.4 Aanvullende zorgen................................................................................................... 25
13.4.1 Steekproefgewichten........................................................................................... 25
13.4.2 Collineariteit........................................................................................................ 25
13.4.3 Meetfouten.......................................................................................................... 25
13.4.4 Penalized regression (kort)..................................................................................26
Cheat sheet hoofdstuk 13.................................................................................................. 26
Pagina 3