Samenvatting Onderzoeksmethoden (STATA)

Inhoudsopgave



1 Basisprincipes van econometrie...................................................................................................... 4

1.1 Introductie..................................................................................................................................... 4

1.2 Ordinary Least Squares (OLS)..................................................................................................... 5
1.1.1 Gauss-Markov assumpties..................................................................................................... 6

2 Basisprincipes en categorische variabelen....................................................................................7

1.3 Basis regressiemodel................................................................................................................... 7

1.4 Multiple regression........................................................................................................................ 8
2.1.1 Goodness of fit....................................................................................................................... 8
2.1.2 Rvf-plot................................................................................................................................... 8

1.5 Dummy-variabelen........................................................................................................................ 9
2.1.3 Gebruik van dummy-variabelen.............................................................................................. 9

3 Formele specificatie en enkelvoudige hypothesetest..................................................................10

1.6 Removing the unit scale: standardized coefficients....................................................................10

1.7 Transforming our variables: Logarithmic transformations...........................................................10
3.1.1 Polynomials.......................................................................................................................... 10
3.1.2 Logarithms............................................................................................................................ 11

1.8 Normaliteit-testing....................................................................................................................... 12
3.1.3 Normaliteit residu’s: visuele inspectie...................................................................................12
3.1.4 Normaliteit residu’s: statistische testen.................................................................................12

1.9 Enkelvoudige hypothesis testing................................................................................................. 13
3.1.5 T-test.................................................................................................................................... 13
3.1.6 Confidence intervals............................................................................................................. 14
3.1.7 Nullity testen van een parameter.......................................................................................... 14

4 Testen van hypotheses op meerdere paramaters........................................................................15

1.10 Hypothesetesting op meerdere parameters: T-distribution & F-distribution..............................15
4.1.1 2-parameter Equality............................................................................................................ 15
4.1.2 2-parameter Restriction........................................................................................................ 16
4.1.3 M-parameter Nullity: Joint F-test...........................................................................................16
4.1.4 M-parameter Nullity: Joint F-test – a Special Case...............................................................16
4.1.5 M-parameter Restriction....................................................................................................... 17

1.11 Hypothesetesting Special Cases: Testing Parameter Stability.................................................17
4.1.6 Chow-test............................................................................................................................. 17
4.1.7 Alternatief voor Chow-test: dummy-variabelen.....................................................................18

1.12 Hypothesetesting Special Cases: Formal Specification Error Test using RESET-test..............19

, 4.1.8 Visuele inspectie: Partial Plots.............................................................................................. 19
4.1.9 Statistische inspectie: Ramsey’s RESET test.......................................................................19

1.13 Extreme observaties................................................................................................................. 20
4.1.10 Ex ante: Visuele inspectie................................................................................................... 20
4.1.11 Ex post: Visuele inspectie................................................................................................... 20
4.1.12 Ex post: Dfbeta’s................................................................................................................ 20
4.1.13 Ex post: Standardized Dfbeta’s.......................................................................................... 21
4.1.14 Ex post: Studentized Residuals.......................................................................................... 21
4.1.15 Extreme observaties oplossen............................................................................................ 21

5 Multicollineariteit en modeldiagnose............................................................................................. 22

1.14 Multicollineariteit....................................................................................................................... 22
5.1.1 Perfecte multicollineariteit..................................................................................................... 22
5.1.2 Non-perfecte multicollineariteit............................................................................................. 22
5.1.3 Multicollineariteit detecteren.................................................................................................23
5.1.4 Multicollineariteit oplossen.................................................................................................... 23

1.15 Model specification en diagnostic testing..................................................................................24
5.1.5 Main causes of bias in OLS.................................................................................................. 24
5.1.6 Underfitting........................................................................................................................... 24
5.1.7 Overfitting............................................................................................................................. 24
5.1.8 Variabelen: include or not?................................................................................................... 25

6 Heteroskedasticiteit......................................................................................................................... 26

1.16 Gevolgen van heteroskedasticiteit............................................................................................ 26

1.17 Heteroskedasticiteit detecteren................................................................................................. 27
6.1.1 Visuele inspectie................................................................................................................... 27
6.1.2 Statistische inspectie............................................................................................................ 27

1.18 Heteroskedasticiteit oplossen................................................................................................... 28

7 Autocorrelatie.................................................................................................................................. 29

1.19 Oorzaken van autocorrelatie..................................................................................................... 29

1.20 Gevolgen van autocorrelatie..................................................................................................... 30

1.21 Autocorrelatie detecteren.......................................................................................................... 30
7.1.1 Visuele inspectie................................................................................................................... 30
7.1.2 Statistische inspectie............................................................................................................ 31

1.22 Autcorrelatie oplossen.............................................................................................................. 32
7.1.3 Lagged variables.................................................................................................................. 32

1.23 Autocorrelatie oplossen in STATA............................................................................................ 33

,
, 1 Basisprincipes van econometrie



1.1 Introductie




Omitted variable bias: Onterecht variabelen uit het model halen, kan voor vertekening zorgen


Cross-section: Iets bestuderen op één moment
Panel data: Eén fenomeen meerdere keren observeren overheen de tijd
Pooled data: Meerdere fenomenen bestuderen overheen tijd


Spurious correlations: Correlatie ≠ causaliteit  gebruik gezond verstand hiervoor


β0: Constante term, intercept
β1: Helling, rico
 Als β1 = 450, dan betekent het dat als X met één eenheid stijgt, Y gaat stijgen met 450 eenheden


 Extreme observaties gaan ervoor zorgen dat de regressielijn naar zich toe wordt getrokken


Regressielijn: Een lijn door de puntenwolk, een lijn die het gemiddelde gaat weergeven van de
punten (observaties)


Variabelen: Informatie in de data
 Zoals bijvoorbeeld ‘C’ en ‘I’
Parameters: Ongekende coëfficiënten
 Zoals bijvoorbeeld β0 & β1


Y-variabele: Afhankelijke, endogene, verklaarde variabele
X-variabele: Onafhankelijke, exogene, verklarende, voorspellende variabele


Lagged: Vertraagd


Error-term:
o Storingsterm, residual of residu
o Hetgeen we niet kunnen verklaren, is compleet random