Repository files navigation

Thediscovr package contains resources for my 2026 textbookDiscovering Statistics Using and.There are tutorials written usinglearnr. Once a tutorial is runningit’s a bit like reading a book but with places where you can practicethe code that youhave just been taught. Thediscovr package is free and offered tosupport tutors and students using my textbook who want to learn.

NOTE Over summer 2025 the tutorials within this package were fullyupdated (see News).

To usediscovr you first need to install and andfamiliarise yourself with, and goodworkflow practice. You can do this usingthis interactivetutorial.Once you have installed and you caninstalldiscovr.

The current released version is available from CRAN:

install.packages("discovr")

The package is currently in constant development to get the most recentversion install it from github (but see the note above).

if(!require(remotes)){  install.packages('remotes')}remotes::install_github("profandyfield/discovr")

I recommend working throughthis playlist oftutorialson how to install, set up and work within and beforestarting the interactive tutorials.

• discovr_01: Introducing, RStudio andQuarto: What is R, tour of RStudio and Quarto, getting help,installing packages, coding style and loading packages.
• discovr_02: Code fundamentals: Functions and objects, packages andfunctions, style, data types.
• discovr_03: The tidyverse: tidy and messy data, tibbles, adding andselecting variables, filtering cases.
• discovr_04: Summarizing data: mean, median, variance, standarddeviation, interquartile range, normal and bootstrap confidenceintervals, tables of summary statistics. Includes an interactive appdemonstrating what a confidence interval is.
• discovr_05: Visualizing data. The ggplot2 package, boxplots,plotting means, violin plots, scatterplots, grouping by colour,grouping using facets, adjusting scales, adjusting positions.”
• discovr_06: The beast of bias. Restructuring data from messy to tidyformat (and back). Spotting outliers using histograms and boxplots.Calculatingz-scores (standardizing scores). Writing your ownfunction. Usingz-scores to detect outliers. Q-Q plots. Calculatingskewness, kurtosis and the number of valid cases. Grouping summarystatistics by multiple categorical/grouping variables.
• discovr_07: Associations. Plotting data with GGally. Pearson’sr,Spearman’s Rho, Kendall’s tau, robust correlations. Usingdisplay()to round output more flexibly.
• discovr_08: The general linear model (GLM). Visualizing the data,fitting GLMs with one and two predictors. Viewing model parameterswith broom, model parameters, standard errors, confidence intervals,fit statistics, significance.
• discovr_09: Categorical predictors with two categories (comparingtwo means). Comparing two independent means, comparing two relatedmeans, effect sizes, robust comparisons of means (independent andrelated), Bayes factors and estimation (independent and relatedmeans).
• discovr_10: Moderation and mediation. Centring variables (grand meancentring), specifying interaction terms, moderation analysis, simpleslopes analysis, Johnson-Neyman intervals, mediation with onepredictor, direct and indirect effects, mediation usinglavaan.
• discovr_11: Comparing several means. Essentially ‘One-wayindependent ANOVA’ but taught using a general linear model framework.Covers setting contrasts (dummy coding, contrast coding, and linearand quadratic trends), theF-statistic and Welch’s robustF,robust parameter estimation, heteroscedasticity-consistent tests ofparameters, robust tests of means based on trimmed data,post hoctests.
• discovr_12: Linear models involving continuous and categoricalpredictors. The first example looks at the case o moderation(non-paralell slopes models), whereas the second explores comparingmeans adjusted for other variables (a parallel slopes model or‘Analysis of Covariance (ANCOVA)’). The tutorial covers settingcontrasts, fitting the models, evaluating effects usingF-statisticsbased on Type III sums of squares and diagnostic plots, andinterpretting the model using heteroscedasticity-consistent tests ofparameters andpost hoc tests.
• discovr_13: Factorial designs. Fitting models for two-way factorialdesigns (independent measures) usinglm(). This tutorial builds onprevious ones to show how models can be fit with two categoricalpredictors to look at the interaction between them. We look at fittingthe models, setting contrasts for the two categorical predictors,interaction plots, simple effects analysis, diagnostic plots androbust models.
• discovr_13_afex: Factorial designs. Fitting models for two-wayfactorial designs (independent measures) using theafex package.This tutorial takes an ANOVA approach to factorical designs. We lookat fitting the models, interaction plots, simple effects analysis,diagnostic plots, partial omega-squared and robust models.
• discovr_14: Multilevel models. This tutorial looks at fittingmultilevel models using theglmmTMB package (all code will also workwithlme4). It begins with an optional section on checking andcoding categorical variables before moving on to show you how to fitand interpret a multilevel model. We also look briefly at thepurrrpackage.
• discovr_15: Repeated measures designs. Fitting models for one- andtwo-way repeated measures designs using theafex package. Thistutorial builds on previous ones to show how models can be fit withone or two categorical predictors when these variables have beenmanipulated within the same entities. We look at fitting the models,setting contrasts for the categorical predictors, obtaining estimatedmarginal means, interaction plots, simple effects analysis, diagnosticplots and robust models.
• discovr_15_growth: Modelling change over time. Growth models usingmultilevel modelling and theglmmTMB package. (All code will alsowork withlme4.) First we explore growth over time by building up amodel to include a random intercept and slope for time. We then modelnon-linear change using both an exponential effect of time and apolynomials. We then extend the model to an example based on aclinical trial in which a fixed effect of an intervention moderateschange over time.
• discovr_15_mlm: Repeated measures designs. Fitting models for one-and two-way repeated measures designs using a multilevel modelframework usingglmmTMB. (All code will also work withlme4.) Theexamples matchdiscovr_15 but the modelling approach differs. Thistutorial builds on previous ones to show how models can be fit withone or two categorical predictors when these variables have beenmanipulated within the same entities. We look at fitting the models,setting contrasts for the categorical predictors and diagnostic plots.
• discovr_16: Mixed designs. Fitting models for mixed designs usingtheafex package. This tutorial builds on previous ones to show howmodels can be fit with one or two categorical predictors when at leastone of these variables has been manipulated within the same entitiesand at least one other has been manipulated using different entities.We look at fitting the models, setting contrasts for the categoricalpredictors, obtaining estimated marginal means, and interaction plots.
• discovr_17: Exploratory factor analysis (EFA). This tutorial looksat using exploratory factor analysis in the context of questionnairedesign. It covers factor analysis, parallel analysis and reliabilityanalysis using MacDonald’s Omega.”.
• discovr_18: Categorical variables. Entering categorical data,contingency tables, associations between categorical variables, thechi-square test, standardized residuals, Fisher’s exact test.
• discovr_19: Categorical outcomes (logistic regression). Thistutorial builds on previous ones to show how the general linear modelmodel extends to situations where you want to predict a binary outcome(logistic regression). We look at fitting the models and interpretingthe odds ratio.
• discovr_19_xmas: Christmas edition ofdiscovr_19 to match thelecture I givehttps://youtu.be/yniFrp8vQLQ?si=DaUVAmAL6sZQ2tkT.
• discovr_bayes: Bayesian taster tutorial. This tutorial offers ataster of Bayesian statistics by showing how to estimate models fromother tutorials within a Bayesian framework usingrstanarm. We alsolook at Bayes factors. The tutorial includes five examples of linearmodels: (1) predicting a continuous outcome from several continuouspredictors; (2) comparing two means; (3) comparing multiple means; (4)comparing means adjusted for a covariate (ANCOVA); and (5) predictinga continuous outcome from two continuous predictors (a factorialdesign).

In Version 1.3onwards there is a tutorial pane. Having executed

library(discovr)

A list of tutorials appears in this pane. Scroll through them and clickon thebutton to run the tutorial:

Alternatively, to run a particular tutorial from the console execute:

library(discovr)learnr::run_tutorial("name_of_tutorial",package="discovr")

and replace “name of tutorial” with the name of the tutorial you want torun. For example, to run tutorial 2 execute:

learnr::run_tutorial("discovr_02",package="discovr")

The name of each tutorial is in bold in the list above. Once the commandto run the tutorial is executed it will spring to life in a web browser.

The tutorials are self-contained (you practice code in code boxes) soyou don’t need to use at the sametime. However, to get the most from them I would recommend that youcreate anproject and within that open (and save) a new RMarkdown file each timeto work through a tutorial. Within that Markdown file, replicate partsof the code from the tutorial (in code chunks) and use Markdown to writenotes about what you have done, and to reflect on things that you havestruggled with, or note useful tips to help you remember things.Basically, write a learning journal. This workflow has the advantage ofnot just teaching you the code that you need to do certain things, butalso provides practice in using itself.

See this video explaining my suggested workflow:

Inspired by therockthemes package andadapting code form that package I have come up with a bunch of colourthemes based around the studio albums of my favourite bandIronMaiden. Full disclosure, I’m not adesigner, so this largely involved uploading images of their sleeves tocolorpalettefromimage.com andseeing what happened. If you have a better palette design send me thehex codes for the colours! If you’re wondering why some albums aremissing, here’s the explanation: X Factor (would basically be 8 shadesof gray), Fear of the Dark (shit album), The Book of Souls (wouldbasically be 8 shades of black).

There are also colour blind accessible pallettes based onOkabe andIto andPaul Tol’s mutedpalette.

The following palettes exist.

• amolad_pal(): Colour palette (8 colour) based on Iron Maiden’sAMatter of Life andDeathalbum sleeve. Inggplot2 usescale_color_amolad() andscale_fill_amolad().
• bnw_pal(): Colour palette (8 colour) based on Iron Maiden’sBraveNew Worldalbum sleeve. Inggplot2 usescale_color_bnw() andscale_fill_bnw().
• dod_pal(): Colour palette (8 colour) based on Iron Maiden’sDanceof Deathalbum sleeve. Inggplot2 usescale_color_dod() andscale_fill_dod().
• frontier_pal(): Colour palette (8 colour) based on Iron Maiden’sThe FinalFrontieralbum sleeve. Inggplot2 usescale_color_frontier() andscale_fill_frontier().
• im_pal(): Colour palette (8 colour) based on Iron Maiden’seponymousalbum sleeve. Inggplot2 usescale_color_im() andscale_fill_im().
• killers_pal(): Colour palette (8 colour) based on Iron Maiden’sKillers albumsleeve. Inggplot2 usescale_color_killers() andscale_fill_killers().
• nob_pal(): Colour palette (8 colour) based on Iron Maiden’sTheNumber of theBeastalbum sleeve. Inggplot2 usescale_color_nob() andscale_fill_nob().
• okabe_ito_pal: Colourblind-friendly palette (8 colour) fromOkabeand Ito. Inggplot2 usescale_color_oi() andscale_fill_oi().
• pom_pal(): Colour palette (8 colour) based on Iron Maiden’sPieceof Mind albumsleeve. Inggplot2 usescale_color_pom() andscale_fill_pom().
• power_pal(): Colour palette (8 colour) based on Iron Maiden’sPowerslavealbum sleeve. Inggplot2 usescale_color_power() andscale_fill_power().
• prayer_pal(): Colour palette (8 colour) based on Iron Maiden’sNoPrayer for theDyingalbum sleeve. Usescale_color_prayer() andscale_fill_prayer().
• senjutsu_pal(): Colour palette (10 colour) based on the innergatefold image of Iron Maiden’sSenjutsualbum albumsleeve. Inggplot2 usescale_color_senjutsu() andscale_fill_senjutsu().
• sit_pal(): Colour palette (8 colour) based on Iron Maiden’sSomewhere inTimealbum sleeve. Inggplot2 usescale_color_sit() andscale_fill_sit().
• ssoass_pal(): Colour palette (8 colour) based on Iron Maiden’sSeventh Son of a SeventhSonalbum sleeve. Inggplot2 usescale_color_ssoass() andscale_fill_ssoass().
• virtual_pal(): Colour palette (8 colour) based on Iron Maiden’sVirtual IXalbum sleeve. Inggplot2 usescale_color_virtual() andscale_fill_virtual().

To view the palette execute

scales::show_col(name_of_palette()(8))

Replacingname_of_palette() with the name, for example

scales::show_col(pom_pal()(8))

To apply, for example, the Powerslave palette to the colours of aggplot2 plot addscale_color_power() as a layer:

library(ggplot2)# Get albums in the classic era from the discovr::eddiefy data.# I'm not including fear of the dark because it's not in any way classic.# No prayer for the dying was pushing its luck too if I'm honest.classic_era<- subset(discovr::eddiefy,year<1992) ggplot(classic_era, aes(x=energy,y=valence,color=album_name))+  geom_point(size=2)+discovr::scale_color_power()+  theme_minimal()

Similarly to apply the Powerslave palette to the fill of objects in aggplot addscale_fill_power() as a layer:

ggplot(classic_era, aes(x=album_name,y=valence,fill=album_name))+  geom_violin()+discovr::scale_fill_power()+   theme(axis.text.x= element_text(angle=45))+  theme_minimal()

See the book or data descriptions for more details. This is a list ofavailable datasets within the package. Raw CSV files are available fromthe book’s website.

• acdc: Data about whether Bon Scott or Brian Johnson is the bestsinger of AC/DC. For details execute?acdc.
• album_sales: fictitious data about predicting album sales fromadvertising, airplay and the band’s image. For details execute?album_sales.
• alien_scents: fictitious data about training sniffer dogs todetect alien space lizards when they try to mask their identity withdifferent scents. For details execute?alien_scents.
• animal_bride: fictitious data about life satisfaction when marriedto a dog or a goat. For details execute?animal_bride.
• angry_pigs: fictitious data about whether playing the video gameangry pigs makes people more aggressive towards pigs. For detailsexecute?angry_pigs.
• angry_real: fictitious data about whether playing the video gameangry pigs makes people more aggressive in everyday life. For detailsexecute?angry_real.
• animal_dance: Fictitious data about training cats and dogs todance.?animal_dance
• beckham_1929: Data from a study by Beckham (1929). For detailsexecute?beckham_1929.
• big_hairy_spider: fictitious data about whether anxiety is greaterafter exposure to real spiders or pictures of spiders. For detailsexecute?big_hairy_spider.
• biggest_liar: fictitious data about creativity and telling lies.For details execute?biggest_liar.
• bronstein_2019: Data about whether delusion proneness predictsbelief in fake news because of less analytic thinking. For detailsexecute?bronstein_2019.
• bronstein_miss_2019: The data in [bronstein_2019] but withmissing values inserted using MCAR amputation. For details execute?bronstein_miss_2019.
• catterplot: fictitious data for plotting a catterplot. For detailsexecute?catterplot.
• cat_dance: fictitious data about training cats to dance. Fordetails execute?cat_dance.
• cat_reg: fictitious data about training cats to dance. For detailsexecute?cat_reg.
• cetinkaya_2006: data from a study by Cetinkaya and Domjan (2006)about quails with sexual fetishes. Seriously. For details execute?cetinkaya_2006.
• chamorro_premuzic: Data about what students want (personalitywise) from their lecturers. For details execute?chamorro_premuzic.
• child_aggression: fictitious data (based on real research) aboutpredicting aggression in children. For details execute?child_aggression.
• coldwell_2006: Data predicting childhood adjustment from variousparenting variables. For details execute?coldwell_2006.
• cosmetic: Fictitious multilevel data predicting quality of lifefrom cosmetic surgery. For details execute?cosmetic.
• daniels_2012: Data about the effects of sexualised sports imageson self-image. For details execute?daniels_2012.
• dark_lord: fictitious data about the subliminal messages in songs.For details execute?dark_lord.
• davey_2003: Data about the effects mood and stop rules on checkingbehaviour. For details execute?davey_2003.
• dog_training: Data about the training dogs to vocalise when theysniff alien life forms. For details execute?dog_training.
• download: fictitious data about the download music festival andbeing smelly. For details execute?download.
• df_beta: fictitious data used to illustrate the DF Beta statistic.For details execute?df_beta.
• eel: Fictitious data about a randomized control trial to testwhether eel therapy is an effective treatment of constipation. Fordetails execute?eel.
• elephooty: Fictitious data about elephants playing football(soccer). For details execute?elephooty.
• escape: Fictitious data about whether I’m a better songwriter thanmy old bandmate Malcolm?escape.
• essay_marks: fictitious data about essay marking. For detailsexecute?essay_marks.
• exam_anxiety: fictitious data about exam performance, anxiety andrevision. For details execute?exam_anxiety.
• exercise: Fictitious data from a randomised control trial of theeffect of exercise on emotional well-being. For details execute?exercise.
• field_2006: Data that tests a hypothesis that threat informationaffects children’s avoidance of novel animals. For details execute?field_2006.
• gallup_2003: Data that tests a hypothesis about why penises have abell end. For details execute?gallup_2003.
• gelman_2009: Data used to critically evaluate the explanations(and claim) that there are more beautiful women than men in the world.For details execute?gelman_2009.
• glastonbury: More fictitious data about music festivals and beingsmelly. For details execute?glastonbury.
• goggles: fictitious data about whether alcohol affects perceptionof physical attractiveness. For details execute?goggles.
• goggles_lighting: fictitious data about the moderating effect oflighting on the ratings of attractivenesses of faces after differentdoses of alcohol. For details execute?goggles_lighting.
• grades: fictitious data about statistics grades. For detailsexecute?grades.
• hangover: fictitious data about the efficacy of different drinksas cures for a hangover. For details execute?hangover
• hiccups: fictitious data on digital rectal stimulation andhiccups. For details execute?hiccups.
• hill_2007: Data from Hill et al. (2007) testing the effect ofdifferent forms of psychoeducation on exercise behaviour. For detailsexecute?hill_2007.
• honesty_lab: fictitious data about perceptions of honesty. Fordetails execute?honesty_lab.
• ice_bucket: Data about the ice bucket challenge. For detailsexecute?ice_bucket.
• invisibility_base: Fictitious data about how much mischief peoplewould get up to if they had an invisibility cloak using a pre-poststudy design.?invisibility_base
• invisibility_cloak: fictitious data about how much mischief peoplewould get up to if they had an invisibility cloak using an independentdesign. For details execute?invisibility_cloak.
• invisibility_rm: fictitious data about how much mischief peoplewould get up to if they had an invisibility cloak but using a repeatedmeasures design. For details execute?invisibility_rm.
• jiminy_cricket: fictitious data about whether wishing on a starmakes you successful. For details execute?jiminy_cricket.
• johns_2012: Data about whether the colour red is a mating signalto men. For details execute?johns_2012.
• lambert_2012: Data about whether pornography use is related torelationhsip commitment and infidelity. For details execute?lambert_2012.
• massar_2012: Data about whether gossiping has an evolutionaryfunction. For details execute?massar_2012.
• mcnulty_2008: Simulated data to match the results of a study aboutwhether attractivenes sis linked to the support given within arelationship. For details execute?mcnulty_2008.
• men_dogs: fictitious data about whether men exhibit dog-likebehaviours (compared to dogs). For details execute?men_dogs.
• metal: Fictitious data about whether listening to metal musicmakes you angry?metal.
• metal_health: fictitious data about whether listening to heavymetal negatively affects mental health. For details execute?metal_health.
• metallica: Data about thrash metal band, Metallica. For detailsexecute?metallica.
• miller_2007: Data from Miller et al. (2007) testing thehidden-estrus theory. For details execute?miller_2007.
• mixed_attitude: Fictitious data about whether different type ofimagery in advertising affect ratings of different types of drinksbased on the gender identity of the participant. For details execute?mixed_attitude.
• murder: fictitious data about the number of murder each month atthree street locations (Ruskin Avenue, Acacia Avenue and Rue Morgue).For details execute?murder.
• muris_2008: Data about whether you can train children to interpretambiguous situations in a particular way. For details execute?muris_2008.
• nichols_2004: Data from the development of the Internet AddictionScale, IAS (Nichols & Nicki, 2004). For details execute?nichols_2004.
• notebook: fictitious data about whether watching the film thenotebook is emotionally arousing. For details execute?notebook.
• ocd: Fictitious data about interventions for obsessive compulsivedisorder. For details execute?ocd.
• ong_2011: Data about social media profile pictures and personalitytraits. For details execute?ong_2011.
• ong_tidy: Data about social media profile pictures and personalitytraits. For details execute?ong_tidy.
• penalty_shootout: Fictitious data about predictors of penalty kicksuccess in soccer (or whatever sport you enjoy). For details execute?penalty.
• profile_pic: Fictitious data related to whether the number offriend requests from random people on social media is affected bywhether your profile picture depicts you as single or part of aromantic couple. For details execute?profile_pic.
• pubs: Data illustrating the difference between an outlier and aninfluencial case. For details execute?pubs.
• puppies: Fictitious data related to whether puppy therapy works.For details execute?puppies.
• puppy_rct: Fictitious data related to whether puppy therapy workswhen you adjust for a person’s baseline happiness. For details execute?puppy_rct.
• puppy_love: Fictitious data related to whether puppy therapy workswhen you adjust for a person’s love of puppies. For details execute?puppy_love.
• r_exam: Fictitious data relating to an R exam at two universities.For details execute?r_exam.
• reality_tv: Fictitious data relating to whether being on a realityTV show exacerbates personality disorder traits. For details execute?reality_tv.
• raq: Fictitious data relating to a fictional questionnaire about Ranxiety that is not an actual questionnaire. For details execute?raq.
• roaming_cats: fictitious data about how far cats roam from theirhomes. For details execute?roaming_cats.
• rollercoaster: Fictitious data about how roller-coaster inducedfear affects attractiveness ratings. For details execute?rollercoaster.
• santas_log: Fictitious data related to whether the type andquantity of treat consumed on Christmas night affects whether elvessuccessfully deliver presents. For details execute?santas_log.
• self_help: fictitious data about whether self-help books improverelationship satisfaction. For details execute?self_help.
• self_help_dsur: fictitious data about whether self-help booksimprove relationship satisfaction compared to statistics books. Fordetails execute?self_help_dsur.
• sharman_2015: Data from Sharman & Dingle (2015) about whetherlistening to metal music increases anger?sharman_2015.
• shopping: fictitious data about shopping For details execute?shopping_exercise.
• sniffer_dogs: fictitious data about training sniffer dogs todetect alien space lizards. For details execute?sniffer_dogs.
• social_anxiety: fictitious (I think) data about whether socialanxiety symptoms are specific to social anxiety. For details execute?social_anxiety.
• social_media: fictitious data about the effects of social media ongrammar. For details execute?social_media.
• soya: fictitious data about the effects of eating soya on spermcount. For details execute?soya.
• speed_date: Fictitious data related to the extent to whichinterest in dating someone is affected by their looks, personality orthe dating strategy they adopt. For details execute?speed_date.
• stalker: fictitious data about therapy for stalking. For detailsexecute?stalker.
• students: I can’t even remember what this data file contains. Fordetails execute?student.
• superhero: fictitious data about whether wearing differentsuperhero costumes leads to more severe physical injuries. For detailsexecute?superhero.
• supermodel: fictitious data about supermodel salaries. For detailsexecute?supermodel.
• switch: Fictitious data relating to whether injuries from playingvideo console games can be mitigated by a warm up.?switch
• tablets: fictitious data about predicting the desirability ofcomputing tablets. For details execute?tablets.
• tea_15: fictitious data based on real data about cognitivefunctioning and drinking tea. For details execute?tea_makes_you_brainy_15.
• tea_716: fictitious data based on real data about cognitivefunctioning and drinking tea. For details execute?tea_makes_you_brainy_716.
• teaching: fictitious data about the success of different methodsof teaching. For details execute?teaching.
• teach_method: more fictitious data about the success of differentmethods of teaching. For details execute?teach_method.
• text_messages: fictitious data about whether use of messaging appsruins your grammar. For details execute?text_messages.
• tosser: Fictitious data relating to a fictional questionnaireabout The Teaching of Statistics for Scientific Experiments, which isfictional. For details execute?tosser.
• tuk_2011: Data about whether needing to urinate helps decisionmaking. For details execute?tuk_2011.
• tumour: fictitious data about mobile phone use and brain tumours.For details execute?tumour.
• tutor_marks: fictitious data comparing 4 tutors marks of the sameessays. For details execute?tutor_marks.
• van_bourg_2020: Data from van Bourg et al (2020) relating towhether dogs would release their distressed owners from a box. Fordetails execute?van_bourg_2020.
• video_games: fictitious data about the relationship between videogame use, callous unemotional traits and aggression. For detailsexecute?video_games.
• williams: Data relating to the development of a questionnaire tomeasure organizational ability. For details execute?williams
• xbox: Fictitious data relating injuries to the type of videoconsole game played and the console it was played on. For detailsexecute?xbox.
• zhang_sample: Data about whether performing a maths test under adifferent name assists performance. For details execute?zhang_2013_subsample.
• zibarras_2008: Data from Zibarras, Port, and Woods (2008) relatingto the relationship between personality and creativity. For detailsexecute?zibarras_2008.
• zombie_growth: fictitious data that mimics a randomised controltrial over time testing an intervention to transform zombies back totheir pre-zombified state. For details execute?zombie_growth.
• zombie_rehab: fictitious data that mimics a randomised controltrial testing an intervention to transform zombies back to theirpre-zombified state in different clinics. For details execute?zombie_rehab.

Solutions for end of chapter tasks are available atwww.discovr.rocks.

Solutions for the Labcoat Leni tasks are available atwww.discovr.rocks.

Although I recommend working through the interactive solutions, eachbook Chapter has online code and a downloadable R Markdown fileavailable fromwww.discovr.rocks.