Cor Regr Simples
Cor Regr Simples
Cor Regr Simples
1
T5_01Ac1
T5_02Ac1
T5_03Ac1
0.8
T5_04Ac1
T5_05Ac1
0.6
T5_06Ac1
T5_07Ac1
T5_08Ac1 0.4
T5_09Sc1
T5_10Sc1
0.2
T5_11Sc1
T5_12Sc1
T5_13Sc1 0
T5_14Sc1
T5_15Sc1
−0.2
T5_16Sc1
T5_17Em1
T5_18Em1 −0.4
T5_19Em1
T5_20Em1
−0.6
T5_21Em1
T5_22Em1
T5_23Em1 −0.8
T5_24Em1
−1
T5_01Ac1
T5_02Ac1
T5_03Ac1
T5_04Ac1
T5_05Ac1
T5_06Ac1
T5_07Ac1
T5_08Ac1
T5_09Sc1
T5_10Sc1
T5_11Sc1
T5_12Sc1
T5_13Sc1
T5_14Sc1
T5_15Sc1
T5_16Sc1
T5_17Em1
T5_18Em1
T5_19Em1
T5_20Em1
T5_21Em1
T5_22Em1
T5_23Em1
T5_24Em1
Correlation plot
1
Agree1
Consc1
Extra1 0.8
Neuro1
Open1 0.6
Agree2
Consc2
0.4
Extra2
Neuro2
0.2
Open2
CndProb
0
EmoSym
HypAc
−0.2
PeerProb
ProSoc
Locus −0.4
SlfAcd
SlfEmo −0.6
SlfSoc
Grit
−0.8
SE
−1
Agree1
Consc1
Extra1
Neuro1
Open1
Agree2
Consc2
Extra2
Neuro2
Open2
CndProb
EmoSym
HypAc
PeerProb
ProSoc
Locus
SlfAcd
SlfEmo
SlfSoc
Grit
SE
10 20
9 18
8 16
7 14
6 12
RMG
RG
5 10
4 8
3 6
2 4
1 2
0 0
200 300 400 500 600 700 800 40 55 70 85 100 115 130 145 160
GF_GC EPN_EG
Correlations
Table 4.22 Alternative Estimates of effect size. Using the correlation as a scale free estimate of effect
size allows for combining experimental and correlational data in a metric that is directly interpretable
as the effect of a standardized unit change in x leads to r change in standardized y.
4.5.1.1 Spearman ⌥: a Pearson correlation
Regression by.x = Cxy
of ranks
by.x by.x = r x
2 y
x
Cxy
Pearson correlation rxy = x y
rxy
In the first of two major papers
Cohen’s d published
d= in the American
X1 X2
r = ⌥ d2 Journal dof =⌥ Psychology
2r in 1904,
x d +4 1 r 2
Spearman (1904b) reviewed
Hedge’s g
for psychologists
g = X1 sx X2
the efforts
r= ⌥ 2 g
made to define g=
the correlation co-
g +4(d f /N)
e⌅cient by Galton (1888) and Pearson (1895). ⌥ Not only ⇤ did he consider ⌅ the application of
2
the Pearson correlationt -totestranked data, t = 2dbut d f he also 2
r = developed 2 t = 1r drf2 for attenuation
t /(t + d f ) corrections
⇤ 2d f
F-test
and the partial correlation, two subjects 4d 2 d f will ber addressed
F =that = F/(F + d flater. ) F = 1r advantage
The r2 of using
⇤
Chi Square r= 2 /n 2 =r n2
ranked data rather than the raw data is that ln(OR)
it is more robust
ln(OR)
to variations 3.62r
in the extreme
Odds ratio d = 1.81 r= ⌥ ln(OR) = ⌥
scores. For whether a person has an 8,000 or a 6,000 on 1.81 an exam,2 +4
(ln(OR)/1.81) that he or she 1 r2 is the highest
requivalent r with probability p r = requivalent
score makes no difference to the ranks. Consider Y as ten numbers sampled from 1 to 20
and then find the Pearson correlation with Y 2 and eY . Do the same things for the ranks
of these numbers. That is, find the Spearman correlations. As is clear from Figure 4.5, the
important than the squared correlation and it is more appropriate to consider the slope of
Correlação e regressão
http://rpsychologist.com/d3/CI/
http://rpsychologist.com/d3/cohend/
Correlação e Regressão
http://rpsychologist.com/d3/correlation/
Fórmula da correlação
N
∑z xi zyi
r= i=1
N −1
N
⎛ ( x i − x )⎞ ⎛ ( y i − y )⎞
∑
i=1
⎜ s
⎝
⎟⎜ s
⎠ ⎝
⎟
⎠
x y
r=
( N − 1)
https://rpsychologist.com/d3/correlation/
Produto-momento!
å
• Desvios da média em unidades de desvio padrão
z z
• Co-relação: ocorrência simultânea together r= X Y
Y
3
8 X 8
7 7
6 6
5 5
Y
4
Y
3 3
2 2
1 1
0 0
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
X X
http://setosa.io/ev/ordinary-least-squares-
regression/
Conceitos até aqui ...
• Escore z
• Variância/covariância
• Correlação
• Equação da reta: intercept e slope
• VD e VI
• Variância total, Variância residual, variância da regressão
• Proporção de variância explicada
• Graus de liberdade
•F
Manual do jasp: https://jasp-stats.org/jasp-materials/
https://www.jamovi.org
https://rpsychologist.com/d3/CI/
https://rpsychologist.com/d3/cohend/
https://rpsychologist.com/d3/NHST/
https://gallery.shinyapps.io/simple_regression/
https://gallery.shinyapps.io/anova_shiny_rstudio/
https://gallery.shinyapps.io/multi_regression/
Exercício 1
• ex1_ie_bpr_16pf_avdes.sav
• ANOVA
• RMANOVA
• TWO-WAY INDEPENDENT ANOVA
• MIXED FACTOR ANOVA
136 THE SAGE HANDBOOK OF PERSONALITY THEORY AND ASSESSMENT
12
10
10
y1
y2
8
8
6
6
4
4
5 10 15 5 10 15
x1 x2
12
12
10
10
y3
y4
8
8
6
6
4
5 10 15 5 10 15
x3 x4
40 / 119