Gorilla
Academy Home

Gorilla Academy

  • Welcome to Gorilla Academy
  • Attention
  • Overview of selective attention
  • Attention experiment summary
  • Building the task
  • Creating the experiment
  • Analysing the data (attention)
  • Advanced analysis using R Studio & JASP
  • Language
  • Overview of speech perception
  • Language experiment summary
  • Creating the task and experiment
  • Analysing the data (language)
  • Advanced analysis using R Studio
  • Learning
  • Overview of learning
  • Learning experiment summary
  • Creating the n-armed bandit task
  • Organising the data in Excel (learning)
  • Calculating learning rate and total score in R Studio
  • Data analysis in JASP
  • Social Influence
  • Overview on social influence
  • Social influence experiment summary
  • Creating the Rotten Tomatoes task
  • Are you more influenced by critics or fans?
  • Sustained Attention to Response Task (SART)
  • History of the SART with Prof Ian Robertson
  • SART experiment summary
  • Building the SART
  • Are there differences in sustained attention with age and dementia?
  • Advanced pre-processing and using ggstatsplot in R Studio
  • Consumer Psychology & Shop Builder
  • An Introduction to Consumer Psychology with Prof Gareth Harvey
  • Consumer Psychology experiment summary
  • Building a mood induction experiment with Shop Builder
  • Analysing a mood induction experiment with Shop Builder
  • Influencing Consumers
  • Influencing Consumers
Expert walkthroughs from theory to data analysis

Gorilla Academy: Where Gorilla tools meet the real world of practical online research

Each module contains:

  • A short lecture about a psychological or behavioural science topic

    • Grounding each practical experiment in a theoretical framework

  • A walkthrough of designing a replication or extension in Gorilla that builds on a previously published paper

    • The design is extracted from the paper. A step that is rarely explicitly taught!

    • Open Materials for you to use or adapt for your own studies

  • Open Data and walkthrough data analysis (including Excel, JASP, and R)

    • From raw Gorilla data to your results section

    • Advanced data organisation, pre-processing and data visualisation

Who this is for:

For Students

Are you trying to replicate or extend a published study, and don’t know where to start?

Have our lecturers take you through designing, creating and analysing a research project.

Seeing the development of a research project from start to finish will help you succeed with your dissertation or thesis.

For Researchers

Do you aspire to find the best way to translate your research question into a Gorilla protocol?

Or the most elegant way of turning Gorilla raw data into publication-ready results?

See how Gorilla components can be configured to support real research, and learn practical data analysis skills from our video walkthroughs, open data, and analysis scripts.

Avoid common pitfalls in study design and data analysis, so that your research reaches its full potential.

For Teachers

Are you in need of a research methods practical grounded in theory? Are you struggling to find an integrated source of materials, data and analysis scripts ready to go for your students?

Use our open materials, tools and data as you see fit to support your own lectures.

Students will benefit from interacting with real experimental data and seeing first-hand how much goes into generating robust results.

Overview of selective attention


The video underneath will guide you through the fundamental experiments and theories in the field of selective attention.

For more in-depth coverage of this topic I'd recommend a good cognitive psychology textbook. When researching this lecture, I used Cognitive Psychology: A student's handbook by Eysenck & Keane (2020)



Length(mins): 14:21

Experiment summary


Below is a useful summary of the experiment created and analysed for this Gorilla Academy topic. Here you will see which Gorilla features were utilised, what data was analysed and which statistical tests and data visualisations were performed using different softwares!


Selective Attention

Study summary:

This is a study of inattentional deafness, using the Dichotic Listening task. Participants are directed to listen to either men or women; men will be audible through the left headphone and women through the right, or vice versa. Someone walks next to the men's conversation saying “I’m a gorilla”. The participant is asked if they noticed anything strange.

Task features:

Data used:

50 participants

4 conditions (2*2 factorial design; between participants):

  • Listen to Men on Left
  • Listen to Men on Right
  • Listen to Women on Left
  • Listen to Women on Right

Analyses performed:

  • Microsoft Excel: data preprocessing using pivot tables
  • R: data preprocessing using dplyr to combine .csv files and extract relevant data, stringr to code text responses
  • JASP: Chi Square analysis

Creating Dalton & Fraenkel in Task Builder


In this video, I build a dichotic listening task in Gorilla Task Builder. In this task participants listen to an audio scene and respond using text buttons and text entry boxes.

This task was originally created by Dalton & Fraenkel (2012), you can read the full manuscript here.

You can also find the task on Gorilla Open Materials here.



Length(mins): 20:41

Creating Dalton & Fraenkel in Experiment Builder


In this video, I bring evertyhing together to create the full experiment in the Gorilla Experiment Builder. This includes going through the questionnaires, and using a randomiser to direct participants into one of four versions of the task.

This task was originally created by Dalton & Fraenkel (2012), you can read the full manuscript here.

You can also find the full experiment on Gorilla Open Materials here.



Length(mins): 14:31

Analysing Dalton & Fraenkel (2012) using Excel & JASP


In this video, I'll show you how to analyse the data. First, we'll use Microsoft Excel and pivot tables to pre-process the data. Then we'll run a Chi Square analysis in JASP.

You can download a copy of the data filename=data_attention_exp.zip here. 

Length(mins): 16:04

Advanced data organisation and analysis using R Studio & JASP


In this video, I'll show you a more advanced approach for analysing your data. This includes using R Studio to fully pre-process the data, creating a more comprehensive data spreadsheet. We'll run a Chi Square analysis in JASP and look at the effect of filtering participants that didn't pass the checks.

You can download a copy of the data filename=data_attention_exp.zip here. 

You can download my R script filename=Attention_organise_data.R here. 

Length(mins): 15:37

Overview of speech perception


This video will introduce the key problems, solutions, and models in speech perception. This includes an understanding of Motor Theory, TRACE, and Cohort models of speech perception.

For more in-depth coverage of this topic I'd recommend a good cognitive psychology textbook. When researching this lecture, I used Cognitive Psychology: A student's handbook by Eysenck & Keane (2020)



Length(mins): 20:16

Experiment summary


Below is a useful summary of the experiment created and analysed for this Gorilla Academy topic. Here you will see which Gorilla features were utilised, what data was analysed and which statistical tests and data visualisations were performed using different softwares!


Speech Perception

Study summary:

This study involved a Speech Perception task. Participants either see videos with a person speaking or hear audio only, and are instructed to write down the word they hear. Words are either Mouth Leading (where mouth shape provides information about the word before audio onset) or Voice Leading (where mouth shape does not provide information about the word before audio onset).

Task features:

Data used:

50 participants

4 conditions (2*2 factorial design; within participants):

  • Mouth Leading Video
  • Voice Leading Video
  • Mouth Leading Audio
  • Voice Leading Audio

Analyses performed:

  • Microsoft Excel: Data preprocessing using filtering and pivot tables
  • JASP: Repeated Measures ANOVA
  • R: Data preprocessing using dplyr, text editing using stringr, combining data frames using rbind, plotting using ggplot, mixed effects modelling using lme4

Building Karas et al (2019) in Gorilla


In this video, I build a speech perception task in Gorilla Task Builder. In this task participants will either listen to an audio file or watch a video of someone saying a word with or without noise. Participants are instructed to type in what they heard. This task uses video zones and text entry boxes.

This task was originally created by Karas et al (2019), you can read the full manuscript here.

You can also find the task on Gorilla Open Materials here.



Length(mins): 29:04

Analysing accuracy data using an ANOVA


In this video, I'll show you how to analyse the data. First, we'll use Microsoft Excel and pivot tables to pre-process the data. Then we'll run a repeated measures ANOVA in JASP.

You can download a copy of the data filename=data_language_exp.zip here. 

Length(mins): 13:24

Advanced analysis of accuracy using mixed effects models in R Studio


In this video, I'll show you a more advanced approach for analysing your data. This includes using R Studio to fully pre-process the data and run a generalised linear mixed effects model as in the original study.

Mixed effects models are a great tool to learn about. I don't have enough time to go into a lot of detail in this video so have a read of these useful links by Michael Clark and Coding Club.

You can download a copy of the data filename=data_language_exp.zip here.  You can download my R script filename=Language_preprocess_analysis.R here. 

Length(mins): 31:32

Learning: A computer science perspective


This video will introduce the three main learning types; Supervised, Unsupervised, and Reinforcement Learning. Here, I'll discuss the key differences between each type of learning and how each type of learning is engaged for different situations.

For more in-depth coverage of this topic I'd recommend a good cognitive psychology textbook. When researching this lecture, I used Cognitive Psychology: A student's handbook by Eysenck & Keane (2020)



Length(mins): 18:14



High Score

If participants wanted they could leave their initials and be placed on the Gorilla High Score Screen. Congratulations to the top 5 players!

Name Score
GW 5171
sb 5082
DG 4728
Vasili 4624
KA 4613

Experiment summary


Below is a useful summary of the experiment created and analysed for this Gorilla Academy topic. Here you will see which Gorilla features were utilised, what data was analysed and which statistical tests and data visualisations were performed using different softwares!

Update: We have now created a version of this task using Task Builder 2, without requiring any scripting. Check out the Utility Task in our Task Builder 2 Samples project!


Learning

Study summary:

This study involves an n-armed bandit task. Participants choose between 2 slot machines (green or blue) to win the most points. Best option (utility) is the combination of 1) which machine pays out more often (probability) and 2) how much each machine pays out (magnitude; 1-99 written in middle of machine). Utility = Magnitude*Probability

Task features:

Data used:

50 participants

4 conditions (2*2 factorial design; between participants):

  • Stable first; blue pays out
  • Stable first; green pays out
  • Volatile first; blue pays out
  • Volatile first; green pays out

Analyses performed:

  • Microsoft Excel: calculating correct answer using IF, data preprocessing using pivot tables, questionnaire scoring using SUM
  • R: combining data frames using rbind, selecting and filtering data using dplyr, hierarchical Bayesian modelling using hBayesDM (to infer learning rate for each participant)
  • JASP: Paired t-test, mixed ANOVA, correlation analysis

Building Behrens et al (2007) in Gorilla


In this video, I build an n-armed bandit task in Gorilla Task Builder. Participants will choose between a blue and green slot machine in order to win the most points. Each slot machine has a different probability of paying out, as well as a different number of points available. The probilities and points are driven by a spreadsheet. At the end of the video, I used a little bit of scripting to calcalate who's a winner and what the total score is.

Update: We have now created a version of this task using Task Builder 2, without requiring any scripting. Check out the Utility Task in our Task Builder 2 Samples project!

The task was originally created by Behrens et al (2007).

You can find the Gorilla Academy version of the task, built using Task Builder 1 and scripting, on Gorilla Open Materials.

You can download an example of the Excel task spreadsheet including formulas filename=learning_task_spreadsheet.xlsx here. 

Length(mins): 27:48



High Score

If participants wanted they could leave their initials and be placed on the Gorilla High Score Screen. Congratulations to the top 5 players!

Name Score
GW 5171
sb 5082
DG 4728
Vasili 4624
KA 4613

Calculating learning in Excel


In this video, we'll use Microsoft Excel and pivot tables to pre-process the data. This includes using Excel formulas to calculate new variables.

You can download a copy of the data filename=data_learning_exp.zip here. 

Length(mins): 16:14



High Score

If participants wanted they could leave their initials and be placed on the Gorilla High Score Screen. Congratulations to the top 5 players!

Name Score
GW 5171
sb 5082
DG 4728
Vasili 4624
KA 4613

Calculating learning rate and total score in R Studio


In this video, we'll use the hBayesDM package in R Studio to calculate the learning rate for each participant. We'll also calculate scores for winning trials along with the total scores for stable and volatile periods.

If you're interested in learning more about computational modelling of beahaviour there are some great online resources including this page by Drs den Ouden and O'Reilly.

I would strongly recommend going to this page for a course run by Miriam Klein-Flügge, Jacqueline Scholl, Laurence Hunt, and Nils Kolling from Oxford University where they discuss modelling a variation of this exact task.

You can download a copy of the data filename=data_learning_exp.zip here.  You can download my R script filename=Learning_organise.R here. 

Length(mins): 10:51



High Score

If participants wanted they could leave their initials and be placed on the Gorilla High Score Screen. Congratulations to the top 5 players!

Name Score
GW 5171
sb 5082
DG 4728
Vasili 4624
KA 4613

Comparing behaviour in stable and volatile periods using JASP


In this video, we'll use paired T-tests, repeated measures ANOVAs, and correlations to explore this dataset. We'll also look at assumptions and how results change when you use the correct assumptions. All the analyses are conducted in JASP.

You can download a copy of the data filename=data_learning_exp.zip here. 



Length(mins): 16:48



High Score

If participants wanted they could leave their initials and be placed on the Gorilla High Score Screen. Congratulations to the top 5 players!

Name Score
GW 5171
sb 5082
DG 4728
Vasili 4624
KA 4613

An overview of Social Influence


This video will introduce some of the different types of social influence, including: conformity, obedience, social learning, nudging, framing, and contagion.

In this lecture, I also made reference to Stirling University's Nudge database. Click on this link to see lots of real world examples of nudging

For more in-depth coverage of this topic I'd recommend a good social psychology textbook. When researching this lecture, I used Social Psychology by Hogg & Vaughan (2017)



Length(mins): 21:35

Experiment summary


Below is a useful summary of the experiment created and analysed for this Gorilla Academy topic. Here you will see which Gorilla features were utilised, what data was analysed and which statistical tests and data visualisations were performed using different softwares!


Social Influence

Study summary:

This study uses a Social Influence task. Participants rate a movie on a scale from 0 to 100. They then see how the same movie was rated by critics and fans. After that, participants rate the movie a second time.

Task features:

Data used:

49 participants

Analyses performed:

  • Microsoft Excel: data preprocessing using filtering and pivot tables
  • JASP: non-parametric t-test analysis, correlation analysis, mixed-effects modelling
  • Psychometrica: comparing correlations

Building a social influence task in Gorilla


In this video, I build a novel social influence task. Participants rated a series of movies using sliders, after which they learnt what the critics and audience thought and made a second rating. This video shows you how to use branching and embedded data, as well as a bit of scripting to change an attribute (move slider tip to participant's previous rating).

Although there isn't a task out there like this, a lot of inspiration was drawn from De Martino et al (2017). You can read the full manuscript here.

You can also find the task on Gorilla Open Materials here.



Length(mins): 27:28

Analysing a social influence task in Gorilla


In this video, we filter the data in excel as an alternative approach to preprocessing the data. We analyse the data using T-tests, correlations, and linear mixed effects modelling. All the analyses are conducted in JASP.

You can download a copy of the data filename=data_social_exp.zip here. 



Length(mins): 27:07

History of the SART with Prof Ian Robertson


I was incredibly honoured to have one of my mentors and creator of the SART, Professor Ian Robertson, tell me all about Sustained Attention, the creation of the SART, and how to use this amazing tool effectively.

Professor Ian Robertson is Co-Director of the Global Brain Health Institute and Professor Emeritus at Trinity College Institute of Neuroscience

For more in-depth coverage of this topic I'd recommend a good cognitive psychology textbook. I've used Cognitive Psychology: A student's handbook by Eysenck & Keane (2020) in the past.

Prof Ian Robertson has also written a number of award winning books on this topic and others, including Mind Sculpture, The Winner Effect, The Stress Test and his newest book How Confidence Works is coming out in June 2021. You can find out more about Prof Ian Robertson and his works here.



Length(mins): 18:13

Experiment summary


Below is a useful summary of the experiment created and analysed for this Gorilla Academy topic. Here you will see which Gorilla features were utilised, what data was analysed and which statistical tests and data visualisations were performed using different softwares!


SART

Study summary:

This study involves the Sustained Attention to Response task (SART). Participants see numbers 1-9 on screen. They are instructed to press space for each number except number 3. There are two variations of the task: Fixed SART, where the numbers appear in order 1-9, and Random SART, where the numbers are in pseudo-random order.

Task features:

Data used:

60 participants

3 groups (Young, Old, People with Dementia)

Analyses performed:

  • Microsoft Excel: data preprocessing using pivot tables, questionnaire scoring using SUM
  • JASP: correlation analysis, one-way ANOVA, mixed ANOVA
  • R: data preprocessing using dplyr to remove multiple keypresses, plotting and statistics using ggstatsplot

Building the Fixed and Random Sustained Attention to Response Tasks (SARTs) in Gorilla


In this video, I build the fixed and random Sustained Attention to Response Tasks as described in Robertson et al (1997).

Participants will see a series of digits (1-9) in a fixed or random order. Participants were required to respond whenever a number came on screen, unless it was a 3. Being able to snap out of a routine rhythmic patterns is a hallmark of sustained attention. Building this task required using different content types, encoding keyboard responses, using screen time limits, and different spreadsheet randomistion rules. There's even a bit of scripting to randomly change the size of the numbers on each trial.

You can also find the task on Gorilla Open Materials here.



Length(mins): 29:13

Analysing errors in SART performance across Young, Old, and MCI/Dementia


In this video, I use pivot tables in Excel to preprocess the data. We also use filters in Excel to diagnose some crazy values in our data. Later, we'll use correlations, a one way ANOVA, and a mixed ANOVA to address our hypotheses.

All the analyses are conducted in JASP.

You can download a copy of the data filename=data_SART_exp.zip here. 

Length(mins): 25:57

Advanced pre-processing and analysis of the SART in R Studio


In this video, I'll show you how to preprocess your data in R Studio using some basic dplyr functions. In R Studio I was able to easily deal with the double tap problem mentioned here.

I'll also show you some simple and powerful analyses using ggstatsplot. It's a wonderful tool that covers pretty much every type of analysis, combining beautiful plots with detailed stats.

You can download a copy of the data filename=data_SART_exp.zip here.  You can download my R script filename=SART_organise_analyse.R here. 

Length(mins): 12:44

An Introduction to Consumer Psychology with Prof Gareth Harvey


Have you ever left a shop thinking "I don't know why I just bought that"? It's likely because of Consumer Psychology.

It was a joy to sit down and chat with Professor Gareth Harvey and discuss what consumer psychology can and can't do, along with how you can get involved in this area of work. Professor Gareth Harvey is Associate Professor of Consumer Psychology at the Haute école de gestion de Genève (HEG-Genève). You can see his talk at BeOnline 2021 here.

If you want to learn more about Consumer Psychology, Gareth (and many others) recommend The Choice Factory by Richard Shotton which outlines 25 behavioural biases that influence what we buy.



Length(mins): 11:47

Experiment summary


Below is a useful summary of the experiment created and analysed for this Gorilla Academy topic. Here you will see which Gorilla features were utilised, what data was analysed and which statistical tests and data visualisations were performed using different softwares!


Consumer Psychology

Study summary:

This study involved a mood induction task followed by a virtual shopping simulation. Participants were randomised to watch either happy, neutral or sad news videos. They were then given an affect questionnaire (PANAS-SF) as a manipulation check. Participants were then taken to a simulated online shop, where they were instructed to buy a bottle of wine and a dessert for dinner with friends. Participants also filled out demographic questions and the Barratt Impulsiveness Scale (BIS-11).

Task features:

Data used:

100 participants

3 conditions (Happy, Neutral, Sad; between participants)

Analyses performed:

  • Microsoft Excel: questionnaire scoring using SUM
  • JASP: data preprocessing using filters, one-way ANOVA, mixed ANOVA, correlation analysis, Bayes factor

Does mood influence purchasing? Building a mood induction experiment with Shop Builder


In this video, I build a Mood Induction task before participants go into an Online Shop and make purchasing decisions.

Participants will be randomly assigned to watch a series of either happy, neutral, or sad news videos (around 4mins total) after which they fill out an affect questionnaire (PANAS-SF). Participants are then tasked with buying a bottle of wine and a dessert for a dinner party with friends in an online shopping simulation. Will they spend more or less money on supplies if they're happy or sad? Participants also filled out demographic questions and the Barret Impulsivity Scale (BIS-11).

You can find the task on Gorilla Open Materials here.

Learn more about Shop Builder here or see all the features in our Support Docs here



Length(mins): 28:10

Does mood influence purchasing? Analysing a mood induction experiment with Shop Builder


In this video, we use excel and JASP to preprocess the data using simple formulas and filtering. Later, we'll use ANOVAs and correlations to check our mood induction worked and address our hypotheses about mood and purchasing.

All the analyses are conducted in JASP.

You can download a copy of the data filename=data_consumer_exp.zip here. 

Length(mins): 28:30

Influencing Consumers


You can find all the information about our Influencing Consumers course in our guide.