Why becoming an R user?

  • Mainstream in academia for data science, increasingly used in business. Job market advantage!

  • Free & open-source: wherever you go, R will be with you at no costs (unlike MPLUS, MATLAB, SPSS, etc.)

  • Real programming language: difficult at the beginning, but: 1) gives you lots of flexibility; 2) has transfer on other programming languages (e.g., Python).

  • Vast community support thanks to a large and active community (also, chatGPT, Lucrez-IA, etc., know it pretty well!).

  • Huge ecosystem, >20,000 packages on CRAN, more from other sources (e.g., GitHub), to do amazing stuff with statistical data analysis, machine learning, data visualization, developing webapps [shiny], writing reports and even entire books [bookdown, rmarkdown]), can integrate with quarto, github.

  • Facilitates reproducible research by sharing code and workflows.

What you may expect to learn in this course:

  • Executing fundamental operations and using basic functions;

  • Working with essential data types and structures;

  • Gaining some proficiency in managing and manipulating data with vectors and dataframes;

  • Understanding some fundamental concepts of programming.

Over the next couple of years, following this PhD program, you may have the opportunity to learn to use R to perform at least some fundamentals about:

  • Core statistical inference methods;

  • (Generalized) linear (mixed-effects) modeling;

  • Data visualization using ggplot2;

  • Power analysis via data simulation ;

  • Structural Equation Modeling (SEM);

  • Conducting Meta-Analysis.

you may even create greeting cards

you may even create greeting cards

or like fancy infographics

or like fancy infographics

perform classical data analysis

you may create interactive webapps with Shiny


see Shiny gallery


here’s a couple of recent real examples from Psicostat members:

you may create interactive webapps with Shiny

or entire websites and books

examples of other resources that can be created within the R ecosystem, integrating other tools such as GitHub and Quarto:

install R and Rstudio



first of all, for getting started, follow the instructions in Chapter 1 of Introduction2R to ensure that both

  • R (the basic software itself to run the programming language) and
  • RStudio (as the IDE of election)

are installed

R Console (just base R)

R Studio (full IDE)

Some R packages that you will or may need in the future (1/3)

Package Used for what Example(s) of functions
base
(base R)		 Basic functions sum, mean, sqrt, abs, c, data.frame, summary, scale, plot, +, -
stats
(base R)		 Basic statistical calculations and functions sd, cor, cor.test, t.test, lm, glm, AIC, rnorm, rbinom
graphics
(base R)		 Basic statistical calculations and functions plot, boxplot, hist, barplot

(You may actually use these “base” packages very often without even realizing that they are packages)

Some R packages that you will or may need in the future (2/3)

Package Used for what Example(s) of functions
effectsize Compute different effect sizes cohens_d, hedges_g, cohens_f, d_to_r
lme4 Fitting (generalized) (non-)linear mixed-effects models lmer, glmer, ranef
performance Useful tools for models check_collinearity, r2_nagelkerke, icc
effects Display effects for various statistical models allEffects
emmeans Estimate marginal means for various models emmeans

Some R packages that you will or may need in the future (3/3)

Package Used for what Example(s) of functions
ggplot2 Create beautiful plots using The Grammar of Graphics ggplot, geom_point, geom_line, …
lavaan Structural Equation Models (SEM) sem, cfa
semTools Useful tools for SEMs compRelSEM, measEq.syntax
metafor Perform meta-analysis rma, rma.mv, forest, funnel, regtest
brms Fitting practically any Bayesian model via MCMC with STAN brm, set_prior
blavaan Fitting Bayesian SEMs bcfa, bsem

Let’s Test the Environment!

Let’s run a few commands in RStudio to familiarize with its console and see if the installation works properly

rnorm(10) # draw 10 random values from a Standard Normal distribution
 [1] -2.9017454  1.6591803 -0.3806936  0.2842570 -2.1319494 -1.1003835
 [7]  1.5540280 -1.5755473 -0.5630551  0.4359361
?rnorm # open the help tab for the "rnorm" function
round( rnorm(10, mean=100, sd=15) ) # draw 10 values from IQ distribution, round them
 [1]  78 101  87  86 114  87  91  83 116 111


install.packages("psych") # install a package from CRAN
library(psych) # load the newly installed package

fisherz(rho=0.5) # use it to transform a correlation into a Fisher's z
[1] 0.5493061