Setup

#this is to check to see if package are installed and if not to install them or just load them if they are installed!
if(!require(pacman)) {install.packages(c("pacman", "remotes"))}

if(!require(papaja)) {remotes::install_github("crsh/papaja")}

pacman::p_load(here, tinylabels, apaTables, tidyverse, gtsummary, car, GGally, MASS, rcompanion, moments, utils, sjPlot, interactions, kableExtra, report, papaja, bookdown)

p_load_gh("crsh/papaja@devel")


#get relative path
path =  here("RESSOURCES") #this is really important ! you juste need to put the name of YOUR folder and here() will find the relative path for you !
setwd(path) #set working directory

# load data
base  <- read.csv("data.csv")


session = utils::sessionInfo() # get session informations

#opts_chunk$set(echo = F, message=F, warning=F) # set echo F for all
options(scipen = 666, warn=-1, contrasts=c("contr.sum","contr.poly"), knitr.kable.NA = '')  #remove scientific notation # remove warnings #set contrasts to sum ! 
set.seed(666) #set random seed

`%notin%` <- Negate(`%in%`)

# panderOptions('knitr.auto.asis', FALSE) #remove auto styling

\(~\) \(~\)

Introduction

Blabla citation in parentesis (James 1890) blbabla, or citation in text James (1890)

\(~\) \(~\)

Clean up

\(~\) \(~\)

base$Priming = as.factor(base$Priming); base$Product = as.factor(base$Product);

\(~\) \(~\)

Demographics

\(~\) \(~\)

base[c("Age", "Gender", "Profession")] %>%  tbl_summary(statistic = list(all_continuous() ~ "{mean} ({sd})", all_categorical() ~ "{n} / {N} ({p}%)"),)  %>% modify_caption("**Table 1. Demographics **") %>%bold_labels()
**Table 1. Demographics **
Characteristic N = 2961
Age 22 (4)
Gender
Autre/NA 2 / 296 (0.7%)
Femme 239 / 296 (81%)
Homme 55 / 296 (19%)
Profession
Actif.ve 3 / 296 (1.0%)
Etudiant.e 291 / 296 (98%)
Les deux 2 / 296 (0.7%)

1 Mean (SD); n / N (%)

\(~\) \(~\)

Inspections

\(~\) \(~\)

base %>% 
  keep(is.numeric) %>%
  gather() %>%
  ggplot(aes(value)) +
  facet_wrap(~key, scales = "free") + 
  geom_histogram()

#or GGally

\(~\) \(~\)

ggplot(base, aes(x=Priming, y=Age)) +
  geom_violin(trim=FALSE, fill='#A4A4A4', color="darkred")+
  geom_boxplot(width=0.1) + theme_minimal()

\(~\) \(~\)

Statistics

\(~\) \(~\)

Reccomendation

base_recco = filter(base, id %notin% c("13594")) #remove huge outliers, see appendix

final_recco = lm(Reccomend ~Priming*Product + Behav_enviro + Product:Behav_enviro,  data = base_recco)

apa_lm <- papaja::apa_print(anova(final_recco))
papaja::apa_table(apa_lm$table,  caption = "Anova table for Reccomendation.", label =NULL)
(#tab:Report_Regression)
Anova table for Reccomendation.
Effect \(F\) \(\mathit{df}_1\) \(\mathit{df}_2\) \(\mathit{MSE}\) \(p\) \(\hat{\eta}^2_G\)
Priming 0.76 1 289 2.44 .383 .003
Product 32.23 1 289 2.44 < .001 .100
Behav enviro 33.11 1 289 2.44 < .001 .103
Priming \(\times\) Product 0.06 1 289 2.44 .808 .000
Product \(\times\) Behav enviro 5.30 1 289 2.44 .022 .018 
sjPlot::plot_model(final_recco)

\(~\) \(~\)

interactions::interact_plot(final_recco, pred = Behav_enviro, modx = Product, interval = TRUE)


\(~\) \(~\)

Conclusions

\(~\) \(~\)

Example for reporting :

Product (\(F(1, 289) = 32.23\), \(\mathit{MSE} = 2.44\), \(p < .001\), \(\hat{\eta}^2_G = .100\)) and Behav_enviro (\(F(1, 289) = 33.11\), \(\mathit{MSE} = 2.44\), \(p < .001\), \(\hat{\eta}^2_G = .103\)) affected reccomendation. However, the effect of Behav_enviro differed by Product, \(F(1, 289) = 5.30\), \(\mathit{MSE} = 2.44\), \(p = .022\), \(\hat{\eta}^2_G = .018\).

\(~\) \(~\)

Appendix

\(~\) \(~\)

Visual check normality

plotNormalHistogram(base$Reccomend, main = "Reccomendation", sub = paste("skewness =", round(skewness(base$Reccomend, na.rm = TRUE),2)))