Lakens

#Script based on Carter & McCullough (2014) doi: 10.3389/fpsyg.2014.00823

#Load Libraries
library(meta)
library(metafor)

#Insert effect sizes and sample sizes
es.d<-c(0.38,0.41,-0.14,0.63,0.22)
n1<-c(75,48,22,18,60)
n2<-c(75,52,21,20,55)

Lakens

library(MBESS)
Lims <- conf.limits.ncf(F.value = 7, conf.level = 0.90, df.1 <- 4, df.2 <- 50)
Lower.lim <- Lims$Lower.Limit/(Lims$Lower.Limit + df.1 + df.2 + 1)
Upper.lim <- Lims$Upper.Limit/(Lims$Upper.Limit + df.1 + df.2 + 1)
Lower.lim
Upper.lim

Lakens

#####
p1 <-numeric(200) #set up empty container for all p-values
bf1 <-numeric(200) #set up empty container for all bf
p2 <-numeric(200) #set up empty container for all p-values
bf2 <-numeric(200) #set up empty container for all bf
p3 <-numeric(200) #set up empty container for all p-values
bf3 <-numeric(200) #set up empty container for all bf
t <- 1.96 #set start point t-value

for (i in 1:201)

Lakens

#####SIMULATE & PLOT Z-SCORES#########
 
nSims <- 100000 #number of simulated experiments
p <-numeric(nSims) #set up empty container for all simulated p-values
z <-numeric(nSims) #set up empty container for z-scores
 
for(i in 1:nSims){ #for each simulated experiment
x<-rnorm(n = 23, mean = 100, sd = 20) #produce simulated participants
y<-rnorm(n = 23, mean = 110, sd = 20) #produce simulated participants
t<-t.test(x,y) #perform the t-test

Lakens

############DATA############
#All significant Z values: 1.9609,1.9611,1.9727,1.9859,2.0063,2.017,2.0188,2.0231,2.0316,2.0377,2.04,2.0444,2.0494,2.0579,2.062,2.0624,2.0662,2.0737,2.0802,2.0803,2.0821,2.0863,2.0895,2.095,2.0964,2.0983,2.1,2.1,2.1113,2.1219,2.124,2.1299,2.1314,2.1318,2.1444,2.1448,2.1481,2.1491,2.1551,2.1584,2.1599,2.1615,2.1622,2.163,2.1656,2.1711,2.1807,2.1836,2.1865,2.1879,2.1909,2.2068,2.2151,2.2205,2.2212,2.2237,2.2316,2.2369,2.2457,2.2572,2.2668,2.2716,2.2777,2.2892,2.2993,2.3001,2.3034,2.31,2.3171,2.3183,2.3232,2.3282,2.337,2.343,2.3455,2.3459,2.3473,2.3517,2.354,2.3543,2.36,2.3623,2.3697,2.3707,2.3767,2.3781,2.3833,2.3871,2.396,2.4221,2.43,2.4408,2.4463,2.4584,2.4714,2.4751,2.485,2.488,2.4889,2.5121,2.5141,2.5306,2.5387,2.5531,2.5698,2.5711,2.5726,2.5793,2.5803,2.5988,2.6322,2.6335,2.6396,2.6397,2.6468,2.6475,2.6664,2.6687,2.6788,2.702,2.7032,2.7212,2.7245,2.7289,2.729,2.7324,2.736,2.7371,2.7402,2.744,2.7537,2.7578,2.761,2.7775,2.7846,2.7929,2.7931,2.8054,2.8131,2.8474,2.8605,2.8657,2.87

Lakens

nSims <- 11000 #number of simulated experiments
psig <-numeric(nSims) #set up empty container for all simulated p-values
pnonsig <-numeric(nSims) #set up empty container for all simulated p-values
z <-numeric(nSims) #set up empty container for z-scores

for(i in 1:nSims){ #for each simulated experiment
  x<-rnorm(n = 38, mean = 100, sd = 20) #produce  simulated participants
  y<-rnorm(n = 38, mean = 108, sd = 20) #produce  simulated participants
  t<-t.test(x,y) #perform the t-test
  if(t$p.value<=0.05&t$p.value>=0.01) {

Lakens

#                               phack

# Requires
# {psych}

# See Also http://rynesherman.com/blog/phack-an-r-function-for-examining-the-effects-of-p-hacking/

phack <- function(initialN=50, hackrate=10, grp1M=0, grp2M=0, grp1SD=1, grp2SD=1, maxN=100, alpha=.05, alternative="greater", graph=TRUE, sims=100000) {
  require(psych)
  outmat <- matrix(NA, nrow=sims, ncol=6)

Lakens

#####EFFECT WITH PUBLICATION BIAS#########

nSims <- 10000 #number of simulated experiments
z1 <-numeric(nSims) #set up empty container for z-scores

for(i in 1:nSims){ #for each simulated experiment
  x<-rnorm(n = 23, mean = 100, sd = 20) #produce  simulated participants
  y<-rnorm(n = 23, mean = 114, sd = 20) #produce  simulated participants
  t<-t.test(x,y) #perform the t-test
  if(t$p.value<0.05) {

Lakens

#Adapted from on EV Nordheim, MK Clayton & BS Yandell https://www.stat.wisc.edu/~yandell/st571/R/append7.pdf

n.draw = 20 #number of tests you draw
mu = 0.3 #true difference
SD = 1

#determine sample sizes for three tests
library(pwr)
pwr.t.test(d=0.3, power=0.9, sig.level=0.05, type = "one.sample", alternative="two.sided")
pwr.t.test(d=0.3, power=0.9, sig.level=0.01, type = "one.sample", alternative="two.sided") 

Lakens

#You will need to load the R package "hypergeo" to use the vstat function

library(hypergeo)

#Below, I'm vectorizing the function so that I can plot curves. 
#The rest is unchanged from the vstat function by Stober-Davis & Dana.
#If you want to use R unbiased, remove the # before the Rsq adjustment calculation below
vstat <- Vectorize(function(n,p,Rsq)
{
  #Rsq = Re(1-((n-2)/(n-p))*(1-Rsq)*hypergeo(1,1,(n-p+2)*.5,1-Rsq))