Read this paper if you want to learn logistic regression

Fernandes, Antônio Alves Tôrres; Figueiredo Filho, Dalson Britto; Rocha, Enivaldo Carvalho da; Nascimento, Willber da Silva

doi:10.1590/1678-987320287406en

Antônio Alves Tôrres Fernandes Antônio Alves Tôrres Fernandes (antonio.alvestorres@ufpe.br) is a master’s student in Political Science in the PPGCP/UFPE, an undergraduate student in Economics (UCB/DF), and a member of the Research Methods in Political Science Group (DCP/UFPE).

Postgraduate Program in Political Science, Federal University of Pernambuco, Recife, PE, Brazil.

https://orcid.org/0000-0002-0249-517X

Dalson Britto Figueiredo FilhoDalson Britto Figueiredo Filho (dalson.figueiredofo@ufpe.br) is professor at Political Science Graduate Program at Federal University of Pernambuco and author of the book “Quantitative methods in Political Science”, Editora InterSaberes.

Postgraduate Program in Political Science, Federal University of Pernambuco, Recife, PE, Brazil.

https://orcid.org/0000-0001-6982-2262

Enivaldo Carvalho da Rocha Enivaldo Carvalho da Rocha (enivaldocrocha@gmail.com) is a retired full professor at Political Science Graduate Program at Federal University of Pernambuco.

Postgraduate Program in Political Science, Federal University of Pernambuco, Recife, PE, Brazil.

https://orcid.org/0000-0003-1717-523X

Willber da Silva Nascimento Willber da Silva Nascimento (nascimentowillber@gmail.com) holds a PhD in Political Science from UFPE and works as postdoctoral researcher at PPGCP/UFPE/FACEPE.

Postgraduate Program in Political Science, Federal University of Pernambuco, Recife, PE, Brazil.

https://orcid.org/0000-0002-2257-8108

Antônio Alves Tôrres Fernandes (antonio.alvestorres@ufpe.br) is a master’s student in Political Science in the PPGCP/UFPE, an undergraduate student in Economics (UCB/DF), and a member of the Research Methods in Political Science Group (DCP/UFPE).

Dalson Britto Figueiredo Filho (dalson.figueiredofo@ufpe.br) is professor at Political Science Graduate Program at Federal University of Pernambuco and author of the book “Quantitative methods in Political Science”, Editora InterSaberes.

Enivaldo Carvalho da Rocha (enivaldocrocha@gmail.com) is a retired full professor at Political Science Graduate Program at Federal University of Pernambuco.

Willber da Silva Nascimento (nascimentowillber@gmail.com) holds a PhD in Political Science from UFPE and works as postdoctoral researcher at PPGCP/UFPE/FACEPE.

		Disease
Age Group	N	Yes	No	Yes (%)
20-29	10	1	9	0.1
30-34	15	2	13	0.13
35-39	12	3	9	0.25
40-44	15	5	10	0.33
45-49	13	6	7	0.46
50-54	8	5	3	0.63
55-59	17	13	4	0.76
60-69	10	8	2	0.8
Total	100	43	57

Stage	Description
1^st	Identify the dependent variable
2^nd	Note the technical requirements
3^rd	Estimate and fit the model
4^th	Interpret the results
5^th	Validate the results

Variables	Description
Sex (Control)	Dummy: Female (0); Male (1)
Age (Control)	Continuous: age at election.
Education (Control)	Categorical ordinal: Read and write (0); Elementary School incomplete (1); Elementary School complete (2); High School incomplete (3); High School complete (4); Tertiary education incomplete (5); Tertiary Education (6).
Poverty (Control)	Continuous: percentage of poor people in the state.
Ideology (Control)	Categorical: Left (0); Center (1); Right (2).
Vote Increase 2006 (Control)	Dummy: Increased (1); Lowered (0).
Change (Control)	Dummy: Changed parties (1); Did not (0).
Pork (Control)	Continuous: success rate of execution of parliamentary amendments.
Seats per state (Control)	Continuous: number of seats for each state at the Chamber of Deputies.
Expenditures (Control)	Continuous: campaign expenditures
Scandal (IV)	Dummy: Involved in a scandal (1); Not involved in a scandal (0).
Reelection (DV)	Dummy: Reelected (1); Not-reelected (0).

Reelected	N	%
Yes	273	60.53
No	178	39.47
Total	451	100.0

Involved in a scandal	Reelected		Total
	Yes	No
Yes	10 (17.86)	46 (82.14)	56 (100.0)
No	263 (66.58)	132 (33.42)	395 (100.0)
Total	273 (60.53)	178 (39.47)	451 (100.0)

χ²	gl	Sig
6.832	8	0.555

χ²	gl	Sig
56.356	11	0.000

	β	Standard error	Z(Wald)	Sig.	Exp(β)	(exp(β)-1) x 100
(Intercept)	0.552	1.568	0.352	0.725	1.737	73.734
Poverty	1.171	1.419	0.825	0.409	3.224	222.386
Male	-0.005	0.560	-0.009	0.993	0.995	-0.484
Age	-0.014	0.017	-0.830	0.406	0.986	-1.409
Education	-0.060	0.161	-0.370	0.712	0.942	-5.789
Ideology	-0.125	0.224	-0.561	0.575	0.882	-11.782
Vote Increase	0.908	0.341	2.663	0.008	2.480	148.030
Change	0.078	0.382	0.205	0.838	1.081	8.136
Parlamentary amendments	-0.272	0.639	-0.425	0.671	0.762	-23.785
Candidate/seats	-0.005	0.009	-0.516	0.606	0.995	-0.469
Campaign spending	0.000	0.000	3.920	0.000	1.000	0.000
Scandal	-1.677	0.528	-3.176	0.001	0.187	-81.299

-2log likelihood null	-2log likelihood	Cox & Snell R²	Nagelkerke R²	BIC
3,057,559	237,4225	0.229	0.308	301,891

		Predicted		Total
		Not reelected	Reelected
Real	Not reelected	23.50	17.51	41.01
	Reelected	10.60	48.39	58.99
	Total	34.10	65.90	100.00

Sentence	N	%
Death penalty	50	34
Life in prison	97	66
Total	147	100.0

Sentence	Black	Non-black	Total
Death penalty	28	22	50
Life in prison	45	52	97
Total	73	74	147

[1] Antônio Alves Tôrres Fernandes (antonio.alvestorres@ufpe.br) is a master’s student in Political Science in the PPGCP/UFPE, an undergraduate student in Economics (UCB/DF), and a member of the Research Methods in Political Science Group (DCP/UFPE).

Brasil

Brasil

Read this paper if you want to learn logistic regression

ABSTRACT

Introduction:

Materials and Methods:

Results:

Discussion: