library(metafor)
## Zorunlu paket yükleniyor: Matrix
## Zorunlu paket yükleniyor: metadat
## Zorunlu paket yükleniyor: numDeriv
## 
## Loading the 'metafor' package (version 4.8-0). For an
## introduction to the package please type: help(metafor)
library(readxl)
veri <- read_excel("C:/Users/Salih/Desktop/mut.xlsx")
## New names:
## • `` -> `...1`
res3 <- rma(measure = "ABT",
            ai = ai,
            mi = mi,
            ni = ni,
            data = veri)
res3
## 
## Random-Effects Model (k = 28; tau^2 estimator: REML)
## 
## tau^2 (estimated amount of total heterogeneity): 0.2025 (SE = 0.0580)
## tau (square root of estimated tau^2 value):      0.4500
## I^2 (total heterogeneity / total variability):   96.61%
## H^2 (total variability / sampling variability):  29.50
## 
## Test for Heterogeneity:
## Q(df = 27) = 1090.0171, p-val < .0001
## 
## Model Results:
## 
## estimate      se     zval    pval   ci.lb   ci.ub      
##   1.6028  0.0873  18.3677  <.0001  1.4317  1.7738  *** 
## 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
predict(res3, transf = transf.ilogit)
## 
##    pred  ci.lb  ci.ub  pi.lb  pi.ub 
##  0.8324 0.8072 0.8549 0.6692 0.9242
c(
  Ortalama_Alfa   = mean(veri$ai, na.rm = TRUE),
  Standart_Hata   = sd(veri$ai, na.rm = TRUE) / sqrt(sum(!is.na(veri$ai))),
  Medyan_Alfa     = median(veri$ai, na.rm = TRUE)
)
## Ortalama_Alfa Standart_Hata   Medyan_Alfa 
##    0.77964286    0.01615469    0.76500000
forest(
  res3,
  slab = veri$Calisma,
  atransf = transf.ilogit,
  xlab = "Cronbach Alfa",
  mlab = "Rastgele Etki Modeli (Bonett)",
  addpred = TRUE,
  header = c("Yazar(lar) ve Yıl", "Cronbach Alfa [95% GA]")
)

funnel(res3,
       atransf = transf.ilogit,
       xlab = " Cronbach Alfa (Bonett, 2002)",
       ylab = "Standart Hata")

fsn(res3, type = "Rosenthal") 
## Warning: Setting type='General' when using fsn() on a model object.
## 
## Fail-safe N Calculation Using the General Approach
## 
## Average Effect Size:         1.6028 (with file drawer: 0.0211)
## Amount of Heterogeneity:     0.2025 (with file drawer: 0.2363)
## Observed Significance Level: <.0001 (with file drawer: 0.0500)
## Target Significance Level:   0.05
## 
## Fail-safe N: 2062
ranktest(res3) 
## Warning in cor.test.default(yi.star, vi, method = "kendall", exact = exact):
## Cannot compute exact p-value with ties
## 
## Rank Correlation Test for Funnel Plot Asymmetry
## 
## Kendall's tau = 0.0318, p = 0.8125
regtest(res3, model = "rma")   
## 
## Regression Test for Funnel Plot Asymmetry
## 
## Model:     mixed-effects meta-regression model
## Predictor: standard error
## 
## Test for Funnel Plot Asymmetry: z = -0.1892, p = 0.8499
## Limit Estimate (as sei -> 0):   b =  1.6516 (CI: 1.1164, 2.1868)
res3_yayim <- update(res3,
                    mods = ~ factor(yayim))

summary(res3_yayim)
## 
## Mixed-Effects Model (k = 28; tau^2 estimator: REML)
## 
##   logLik  deviance       AIC       BIC      AICc   
## -17.2266   34.4531   40.4531   44.2274   41.5440   
## 
## tau^2 (estimated amount of residual heterogeneity):     0.2100 (SE = 0.0612)
## tau (square root of estimated tau^2 value):             0.4583
## I^2 (residual heterogeneity / unaccounted variability): 96.69%
## H^2 (unaccounted variability / sampling variability):   30.20
## R^2 (amount of heterogeneity accounted for):            0.00%
## 
## Test for Residual Heterogeneity:
## QE(df = 26) = 1087.2590, p-val < .0001
## 
## Test of Moderators (coefficient 2):
## QM(df = 1) = 0.0859, p-val = 0.7694
## 
## Model Results:
## 
##                   estimate      se     zval    pval    ci.lb   ci.ub      
## intrcpt             1.6132  0.0957  16.8535  <.0001   1.4256  1.8009  *** 
## factor(yayim)tez   -0.0751  0.2562  -0.2932  0.7694  -0.5773  0.4271      
## 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
predict(
  res3_yayim,
  newmods = model.matrix(~ factor(yayim), 
                          data = data.frame(yayim = c("Makale", "Tez")))[, -1],
  transf = transf.ilogit
)
## 
##     pred  ci.lb  ci.ub  pi.lb  pi.ub 
## 1 0.8339 0.8062 0.8583 0.6672 0.9263 
## 2 0.8232 0.7450 0.8812 0.6286 0.9276
veri_kat <- veri[-25, ]
res3_kat <- update(
  res3,
  mods = ~ factor(katilim),
  data = veri_kat
)

summary(res3_kat)
## 
## Mixed-Effects Model (k = 27; tau^2 estimator: REML)
## 
##   logLik  deviance       AIC       BIC      AICc   
## -16.5551   33.1102   41.1102   45.8224   43.2155   
## 
## tau^2 (estimated amount of residual heterogeneity):     0.2215 (SE = 0.0670)
## tau (square root of estimated tau^2 value):             0.4706
## I^2 (residual heterogeneity / unaccounted variability): 96.87%
## H^2 (unaccounted variability / sampling variability):   31.95
## R^2 (amount of heterogeneity accounted for):            0.00%
## 
## Test for Residual Heterogeneity:
## QE(df = 24) = 1066.0003, p-val < .0001
## 
## Test of Moderators (coefficients 2:3):
## QM(df = 2) = 0.4154, p-val = 0.8124
## 
## Model Results:
## 
##                            estimate      se     zval    pval    ci.lb   ci.ub 
## intrcpt                      1.7412  0.2809   6.1975  <.0001   1.1905  2.2918 
## factor(katilim)universite   -0.1032  0.3134  -0.3293  0.7419  -0.7174  0.5110 
## factor(katilim)yetiskin     -0.1863  0.3135  -0.5944  0.5522  -0.8007  0.4280 
##                                
## intrcpt                    *** 
## factor(katilim)universite      
## factor(katilim)yetiskin        
## 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
predict(
  res3_kat,
  newmods = model.matrix(~ factor(katilim),
                          data = data.frame(
                            katilim = c("yetiskin", "karma", "universite")
                          ))[, -1],
  transf = transf.ilogit
)
## 
##     pred  ci.lb  ci.ub  pi.lb  pi.ub 
## 1 0.8256 0.7828 0.8614 0.6441 0.9253 
## 2 0.8508 0.7668 0.9082 0.6608 0.9435 
## 3 0.8373 0.7967 0.8710 0.6629 0.9308

*********sürekli değişkenler için moderatör analizi**********

res_kadin <- rma(measure = "ABT", 
                 ai = ai, 
                 mi = mi, 
                 ni = ni, 
                 mods = ~ kadin, 
                 data = veri, 
                 method = "REML")
## Warning: 1 study with NAs omitted from model fitting.
summary(res_kadin)
## 
## Mixed-Effects Model (k = 27; tau^2 estimator: REML)
## 
##   logLik  deviance       AIC       BIC      AICc   
## -16.6631   33.3263   39.3263   42.9829   40.4692   
## 
## tau^2 (estimated amount of residual heterogeneity):     0.2117 (SE = 0.0629)
## tau (square root of estimated tau^2 value):             0.4601
## I^2 (residual heterogeneity / unaccounted variability): 96.70%
## H^2 (unaccounted variability / sampling variability):   30.29
## R^2 (amount of heterogeneity accounted for):            0.00%
## 
## Test for Residual Heterogeneity:
## QE(df = 25) = 1026.0401, p-val < .0001
## 
## Test of Moderators (coefficient 2):
## QM(df = 1) = 0.0405, p-val = 0.8406
## 
## Model Results:
## 
##          estimate      se     zval    pval    ci.lb   ci.ub      
## intrcpt    1.6835  0.3365   5.0027  <.0001   1.0239  2.3431  *** 
## kadin     -0.0010  0.0049  -0.2012  0.8406  -0.0106  0.0087      
## 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
res_kadnull <- rma(measure = "ABT",
                ai = ai,
                mi = mi,
                ni = ni,
                data = veri,
                method = "REML")

R2_kad <- (res_kadnull$QE - res_kadin$QE) / res_kadnull$QE
R2_kad
## [1] 0.05869353
res_yil <- rma(measure = "ABT", 
                 ai = ai, 
                 mi = mi, 
                 ni = ni, 
                 mods = ~ yil, 
                 data = veri, 
                 method = "REML")


summary(res_yil)
## 
## Mixed-Effects Model (k = 28; tau^2 estimator: REML)
## 
##   logLik  deviance       AIC       BIC      AICc   
## -17.0091   34.0183   40.0183   43.7926   41.1092   
## 
## tau^2 (estimated amount of residual heterogeneity):     0.2056 (SE = 0.0600)
## tau (square root of estimated tau^2 value):             0.4535
## I^2 (residual heterogeneity / unaccounted variability): 96.61%
## H^2 (unaccounted variability / sampling variability):   29.52
## R^2 (amount of heterogeneity accounted for):            0.00%
## 
## Test for Residual Heterogeneity:
## QE(df = 26) = 1061.0205, p-val < .0001
## 
## Test of Moderators (coefficient 2):
## QM(df = 1) = 0.5500, p-val = 0.4583
## 
## Model Results:
## 
##          estimate       se     zval    pval     ci.lb     ci.ub    
## intrcpt   35.2140  45.3231   0.7770  0.4372  -53.6177  124.0457    
## yil       -0.0166   0.0224  -0.7416  0.4583   -0.0606    0.0273    
## 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
res_yilnull <- rma(measure = "ABT",
                ai = ai,
                mi = mi,
                ni = ni,
                data = veri,
                method = "REML")


R2_yil <- (res_yilnull$QE - res_yil$QE) / res_yilnull$QE
R2_yil
## [1] 0.02660196
res_mi <- rma(measure = "ABT", 
                 ai = ai, 
                 mi = mi, 
                 ni = ni, 
                 mods = ~ mi, 
                 data = veri, 
                 method = "REML")


summary(res_mi)
## 
## Mixed-Effects Model (k = 28; tau^2 estimator: REML)
## 
##   logLik  deviance       AIC       BIC      AICc   
## -17.2128   34.4256   40.4256   44.1999   41.5165   
## 
## tau^2 (estimated amount of residual heterogeneity):     0.2095 (SE = 0.0611)
## tau (square root of estimated tau^2 value):             0.4577
## I^2 (residual heterogeneity / unaccounted variability): 96.72%
## H^2 (unaccounted variability / sampling variability):   30.52
## R^2 (amount of heterogeneity accounted for):            0.00%
## 
## Test for Residual Heterogeneity:
## QE(df = 26) = 1081.6913, p-val < .0001
## 
## Test of Moderators (coefficient 2):
## QM(df = 1) = 0.1475, p-val = 0.7009
## 
## Model Results:
## 
##          estimate      se    zval    pval    ci.lb   ci.ub    
## intrcpt    0.8804  1.8828  0.4676  0.6401  -2.8099  4.5706    
## mi         0.1823  0.4745  0.3841  0.7009  -0.7477  1.1122    
## 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
res_minull <- rma(measure = "ABT",
                ai = ai,
                mi = mi,
                ni = ni,
                data = veri,
                method = "REML")

R2_mi <- (res_minull$QE - res_mi$QE) / res_minull$QE
R2_mi
## [1] 0.007638216
res_ni <- rma(measure = "ABT", 
                 ai = ai, 
                 mi = mi, 
                 ni = ni, 
                 mods = ~ ni, 
                 data = veri, 
                 method = "REML")


summary(res_ni)
## 
## Mixed-Effects Model (k = 28; tau^2 estimator: REML)
## 
##   logLik  deviance       AIC       BIC      AICc   
## -16.7256   33.4513   39.4513   43.2256   40.5422   
## 
## tau^2 (estimated amount of residual heterogeneity):     0.2006 (SE = 0.0586)
## tau (square root of estimated tau^2 value):             0.4479
## I^2 (residual heterogeneity / unaccounted variability): 96.32%
## H^2 (unaccounted variability / sampling variability):   27.15
## R^2 (amount of heterogeneity accounted for):            0.91%
## 
## Test for Residual Heterogeneity:
## QE(df = 26) = 865.2573, p-val < .0001
## 
## Test of Moderators (coefficient 2):
## QM(df = 1) = 1.1555, p-val = 0.2824
## 
## Model Results:
## 
##          estimate      se     zval    pval    ci.lb   ci.ub      
## intrcpt    1.4800  0.1435  10.3171  <.0001   1.1989  1.7612  *** 
## ni         0.0003  0.0003   1.0749  0.2824  -0.0003  0.0009      
## 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
res_ninull <- rma(measure = "ABT",
                ai = ai,
                mi = mi,
                ni = ni,
                data = veri,
                method = "REML")

R2_ni <- (res_ninull$QE - res_ni$QE) / res_ninull$QE

R2_ni
## [1] 0.2061984
res_full <- rma(measure = "ABT",   # Güvenirlik katsayıları için uygun dönüşüm
                ai = ai,           # Gözlenen güvenirlik katsayıları (Alpha gibi)
                mi = mi,           # Madde sayısı
                ni = ni,           # Örneklem büyüklüğü
                mods = ~ kadin + yil + mi + ni, # Moderatörler
                data = veri)
## Warning: 1 study with NAs omitted from model fitting.
# Sonuçları inceleyin
summary(res_full)
## 
## Mixed-Effects Model (k = 27; tau^2 estimator: REML)
## 
##   logLik  deviance       AIC       BIC      AICc   
## -15.2952   30.5904   42.5904   49.1367   48.1904   
## 
## tau^2 (estimated amount of residual heterogeneity):     0.2233 (SE = 0.0706)
## tau (square root of estimated tau^2 value):             0.4726
## I^2 (residual heterogeneity / unaccounted variability): 96.59%
## H^2 (unaccounted variability / sampling variability):   29.29
## R^2 (amount of heterogeneity accounted for):            0.00%
## 
## Test for Residual Heterogeneity:
## QE(df = 22) = 795.1838, p-val < .0001
## 
## Test of Moderators (coefficients 2:5):
## QM(df = 4) = 1.6820, p-val = 0.7940
## 
## Model Results:
## 
##          estimate       se     zval    pval     ci.lb     ci.ub    
## intrcpt   23.4348  50.0112   0.4686  0.6394  -74.5853  121.4549    
## kadin     -0.0034   0.0058  -0.5852  0.5584   -0.0148    0.0080    
## yil       -0.0113   0.0247  -0.4577  0.6472   -0.0597    0.0371    
## mi         0.2749   0.5107   0.5383  0.5903   -0.7260    1.2758    
## ni         0.0004   0.0003   1.1009  0.2709   -0.0003    0.0010    
## 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1