odev 5

R Markdown

Sabit etkiler (fixed effect veya first differencing) ve ya within estimator

Sabit etkiler regresyon modelinin kesişiminin bireyler veya gruplar arasında serbestçe değişmesine izin verilen istatistiksel bir regresyon modelidir. Zaman içinde değişmeyen bireye özgü nitelikleri kontrol etmek için genellikle panel verilere uygulanır

panal veri seti

Panel verileri zaman içinde eşit aralıklarla bir araya getirilen ve kronolojik olarak sıralanan birden çok kişiden elde edilen niceliklerin bir koleksiyonudur Bireysel gruplara örnek olarak bireysel kişiler ülkeler ve şirketler verilebilir.

panal data veri aşağıda

library(wooldridge)
library(stargazer)

## 
## Please cite as:

##  Hlavac, Marek (2022). stargazer: Well-Formatted Regression and Summary Statistics Tables.

##  R package version 5.2.3. https://CRAN.R-project.org/package=stargazer

library(rmarkdown)
require(plm)

## Zorunlu paket yükleniyor: plm

data("crime3")
paged_table(crime3)

time ve individual gore index

crime.panal <- pdata.frame(crime3, index = c("district", "year"))
pdim(crime3)

## Balanced Panel: n = 53, T = 2, N = 106

PLM model

fixedeffect <- plm(log(crime) ~ avgclr, data = crime.panal, model = "within")
summary(fixedeffect)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = log(crime) ~ avgclr, data = crime.panal, model = "within")
## 
## Balanced Panel: n = 53, T = 2, N = 106
## 
## Residuals:
##        Min.     1st Qu.      Median     3rd Qu.        Max. 
## -5.1118e-01 -1.1521e-01  2.2204e-16  1.1521e-01  5.1118e-01 
## 
## Coefficients:
##          Estimate Std. Error t-value  Pr(>|t|)    
## avgclr -0.0219020  0.0038933 -5.6256 7.463e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    5.1266
## Residual Sum of Squares: 3.187
## R-Squared:      0.37834
## Adj. R-Squared: -0.25528
## F-statistic: 31.6469 on 1 and 52 DF, p-value: 7.4634e-07

NORMAL LM model

normal <- (lm(log(crime) ~ avgclr, data = crime.panal ) )
summary(normal)

## 
## Call:
## lm(formula = log(crime) ~ avgclr, data = crime.panal)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -1.35716 -0.24547 -0.00101  0.31497  0.93049 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  4.122663   0.156403  26.359  < 2e-16 ***
## avgclr      -0.035112   0.003663  -9.586 5.68e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.4514 on 104 degrees of freedom
## Multiple R-squared:  0.4691, Adjusted R-squared:  0.464 
## F-statistic: 91.88 on 1 and 104 DF,  p-value: 5.678e-16

plm ve LM bir arada yada karşılaştırma

stargazer(fixedeffect, normal, type = "text" )

## 
## ==================================================================
##                                  Dependent variable:              
##                     ----------------------------------------------
##                                       log(crime)                  
##                             panel                    OLS          
##                             linear                                
##                              (1)                     (2)          
## ------------------------------------------------------------------
## avgclr                    -0.022***               -0.035***       
##                            (0.004)                 (0.004)        
##                                                                   
## Constant                                          4.123***        
##                                                    (0.156)        
##                                                                   
## ------------------------------------------------------------------
## Observations                 106                     106          
## R2                          0.378                   0.469         
## Adjusted R2                 -0.255                  0.464         
## Residual Std. Error                           0.451 (df = 104)    
## F Statistic         31.647*** (df = 1; 52) 91.882*** (df = 1; 104)
## ==================================================================
## Note:                                  *p<0.1; **p<0.05; ***p<0.01