Gorilla - Wondering if they can take their research online

When you think of an online study, what type of study comes to mind? A ques­tion­naire? A simple reac­tion time task with images and a button click?

It’s easy to imagine how someone else can take a simple study online. But when you think about all the com­plex­i­ties of your own research, it’s harder to see how it could work.

Here at Gorilla we’re in the busi­ness of lis­ten­ing to what researchers want to create, and then devel­op­ing the tools to make that research pos­si­ble. I can’t promise that your research will work online, but I’d ask you to keep an open mind while I show you why there’s a very good chance it will.

Before I talk more about the tools, I’d to share with you some evi­dence that

  1. lots of research can go online and
  2. design­ing online exper­i­ments is often straight­for­ward (maybe even enjoyable)!

 

Success in your research area

The pub­li­ca­tions that cite Gorilla, will give you a peer-reviewed and con­crete sense of what research can be taken online successfully.

We’ve got over 200 pub­li­ca­tions and pre-prints listed, with a lot of diver­si­ty. And that’s not even all of them.

Have a look through the list and see if there’s any­thing in your research area.  If there is, take a look at their methods. Hope­ful­ly it will give you con­fi­dence that you can be suc­cess­ful too.

The second thing to do is to look at our spot­light inter­views.  These are short inter­views with a variety of researchers using Gorilla.  Maybe there’s someone that does research a bit like yours?

Audi­to­ry: Kyle Jasmin carried out an audi­to­ry cat­e­go­riza­tion experiment, inves­ti­gat­ing how we use pitch and dura­tion to detect word emphasis.

Speech: Violet Brown was inter­est­ed in why there are indi­vid­ual dif­fer­ences in sus­cep­ti­bil­i­ty to the McGurk effect.

Linguis­tics: Matthew Hunt pre­dict­ed that par­tic­i­pants would have a greater bias towards select­ing that they had heard an ‘-in’ variant of words ending in the mor­pheme ‘-ing’ when lis­ten­ing to swear words.

Motor: Jonathan Tsay used Gorilla to design a battery of cog­ni­tive tasks, inves­ti­gat­ing how the cere­bel­lum con­tributes to cognition.

Memory: Jade Pick­er­ing inves­ti­gat­ed whether retrieval prac­tice on a subset of inte­grat­ed infor­ma­tion would still be an effec­tive learn­ing strat­e­gy for all of the inte­grat­ed infor­ma­tion. Par­tic­i­pants were asked to imagine three ele­ments (e.g. spidercarrotcircus) inter­act­ing as vividly as pos­si­ble, then did retrieval prac­tice on a subset of pairs (spidercarrot).

Indi­vid­ual Dif­fer­ences: Nick Harp has been using Gorilla for a lon­gi­tu­di­nal study, in which he is track­ing emo­tion­al bias through­out the COVID-19 pan­dem­ic. His results show that self-report­ed cog­ni­tive reap­praisal ten­den­cies are asso­ci­at­ed with a smaller shift toward neg­a­tiv­i­ty in our measure of emo­tion­al bias.

Devel­op­men­tal: Anqi Hu used Gorilla to examine the role of social inter­ac­tion in word learn­ing in school-aged children.

Edu­ca­tion: Jessica Mas­son­nié studies how class­room noise and dis­trac­tions can impact on chil­dren’s learning.

Health: Juliet Usher Smith con­duct­ed a con­trolled online study with 1000 par­tic­i­pants, looking at the effects of per­son­alised lifestyle advice on cancer risk.

Cog­ni­tion: Sahira van de Wouw used Gorilla to build an online version of the fiancée task (par­tic­i­pants rate 90 faces, then encounter 8 faces in sequence and accept or reject each one as their date). Sahira aimed to repli­cate over­sam­pling effects found in the lab version of the study.

Abnor­mal Psy­chol­o­gy: Jen Pink analysed the rela­tion­ships between dif­fer­ent types of aggres­sion and traits in psy­chopa­thy for 3‑factor and 7‑factor psy­chopa­thy models in order to compare the models.

Com­par­i­tive Psy­chol­o­gy: Kirsty Graham and Cat Hobaiter used Gorilla for a large scale study inves­ti­gat­ing whether humans can under­stand the ges­tures pri­mates use to com­mu­ni­cate. Thanks to some excel­lent PR with cov­er­age in The Atlantic and on the BBCs Life Sci­en­tif­ic, they managed to recruit 15,000 participants!

 

Prac­ti­cal Online Research Methods

We also sponsor BeOn­line: a con­fer­ence all about online research, with a par­tic­u­lar focus on online research methods.

The speak­ers present their spe­cif­ic research methods and results, but you’ll also find a lot of gen­er­al­is­able infor­ma­tion on how to ensure great data quality online and how to pilot your study suc­cess­ful­ly. You can find great talks on all kinds of sub­jects:

Audi­to­ry: Violet Brown was inter­est­ed in how face masks affect speech per­cep­tion. She record­ed speech with 5 dif­fer­ent face mask con­di­tions (no mask, sur­gi­cal mask, two cloth masks, and a trans­par­ent mask) and then played the clips to participants.

Edu­ca­tion: Diana Lau­ril­lard explains how she worked with Gorilla to design a game that would help dyscal­cu­lia and early learn­ers to under­stand the inter­nal struc­ture of numbers.

Devel­op­men­tal: Kather­ine Ellis is used to con­duct­ing face-to-face research with special pop­u­la­tions, but she’s been learn­ing how to move online since the start of the pan­dem­ic. Her research team has embarked on apply­ing their pre­vi­ous face-to-face pro­to­col, which includes eye-track­ing, cog­ni­tive and behav­iour­al assess­ments, to online methods, with neu­rotyp­i­cal chil­dren, and chil­dren with genetic syn­dromes and intel­lec­tu­al disability.

Social Psy­chol­o­gy: Dan Richard­son at UCL uses Gorilla on a tablet to conduct exper­i­ments in person, in some very unusual places. Learn about the time he tested cre­ativ­i­ty from an under­wa­ter night club in Venice, or in a theatre in London, or during a live YouTube stream…

Behav­iour­al Eco­nom­ics: Pasquale Rummo carries out research on how to promote healthy food pur­chas­es. He’s created an online grocery store website to inves­ti­gate how finan­cial incen­tives and nudges can be effec­tive in online retail spaces.

Eye track­ing: Gaia Sceriff is an early pioneer of online eye track­ing. She dis­cuss­es her journey toward webcam-based eye-track­ing over the last three years, with pros, cons, and excite­ment ahead.

Mouse track­ing: Alex Anwyl-Irvine is one of the devel­op­ers who created MouseView.js, an alter­na­tive to eye track­ing. Mou­se­view creates an occlud­ing layer over a website, so that par­tic­i­pants need to view their mouse to the area of the screen they want to view. Alex dis­cuss­es the tool and the val­i­da­tion research he’s been working on.

Motor: Ryan More­head and his lab have been moving motor control research online. They’ve created a gam­i­fied version of an inter­cep­tive timing task (inter­cept fruit with a fruit bat) and inves­ti­gat­ed how they can stan­dard­ise and account for dif­fer­ences in participant’s equipment.

Health: Dim­itris Koutoukidis tested effec­tive­ness of an indi­vid­ual-level inter­ven­tion and an envi­ron­men­tal-level inter­ven­tion on the sat­u­rat­ed fat content of the shop­ping basket during an online shop­ping experiment.

 

The Gorilla Ecosystem

Now let’s talk about the tools! Gorilla con­tains a variety of tools to support a wide range of research. The glue of Gorilla is our experiment tree, which is used to design the overall flow of the experiment.

Setting up online experiments easily

The experiment tree makes ran­domi­sa­tion, branch­ing, coun­ter­bal­anc­ing, regroup­ing simple.  These are often hor­ri­bly com­pli­cat­ed in code. Phew!

Within the tree, each node can be created in any one of our tools:

  • Ques­tion­naire Builder (for surveys like Qualtrics, Survey Monkey or Google Forms)
  • Task Builder (for reac­tion time tasks like ePrime or Presentation)
  • Code Editor (for build­ing tasks from scratch like MATLAB, or jsPsych)
  • Game Builder (for making 2D games, like you might in Unity or Unreal Engine)
  • Shop Builder (for con­sumer deci­sion making, like Shopify but for research)
  • Mul­ti­play­er Builder (for mul­ti­play­er exper­i­ments, like OTree)

You can link as many dif­fer­ent tasks from dif­fer­ent tools as you like: the pos­si­bil­i­ties are endless.

Within each of those tools, there is a huge range of fea­tures.  Gorilla is a bit like making a model: we give you the build­ing blocks, you come up with end­less­ly cre­ative ways to fit them together.

The flag­ship Gorilla tool is our Task Builder, which con­tains over 60 dif­fer­ent zones (build­ing blocks), includ­ing text and image zones, key­board and response button zones, audio and video zones, eye and mouse track­ing zones… Out of these you can create almost anything!

But, unlike with tra­di­tion­al build­ing blocks, we also give you the ability to create your own pieces. In plainer terms, in most of our tools you can add snip­pets of JavaScript to upgrade the func­tion­al­i­ty we provide.

Most studies can be made within our tools, but if you want to extend the func­tion­al­i­ty that we provide, you can use script­ing or the Code Editor and extend Gorilla to your needs.

 

How to get started: Repli­cate then innovate!

The easiest way to get started is to select a study that you have done in the lab and repli­cate it online. This will help in two ways. Firstly, you have the pro­to­col already clearly defined.  So you aren’t going back to the drawing board and design­ing a new study. You’ve done the lit­er­a­ture review, you’ve got the method, you’ve got the stimuli, so it’s simply a ques­tion of trans­lat­ing it for online use.

Sec­ond­ly, once you’ve col­lect­ed the data you’ll be able to compare the data to the data you col­lect­ed in the lab. This will help you con­vince your own inter­nal Review­er 2 that the data quality online is good enough.

Most researchers are pleas­ant­ly sur­prised that, while online data may have higher vari­ance than lab data, the effects are still there and the larger sample more than compensates.

 

So what do I do now?

Take Gorilla for a test drive!  Cre­at­ing an account is free and easy, so you can take some time to see what Gorilla is like:

We have a few dif­fer­ent guides to help you get started:

Mas­sive­ly impressed with Gorilla. Logged on for the first time yes­ter­day lunchtime. Had a demo expt ready for stu­dents to adapt by 6pm the same day.
- Dr. Jenni Rodd — Senior Lec­tur­er, Uni­ver­si­ty College London

Dr. Jenni Rodd is only one of the many people who love to use Gorilla.

 

What if I get stuck?

We offer a lot of support options.  Once you’ve signed up, book an onboard­ing webinar.  In 1 hour, we’ll show you round the ins and outs of Gorilla so that you can take your research online con­fi­dent­ly. They’re live every other week, but if you can’t wait to get started don’t worry: There’s also an on-demand version.

Beyond that, we have lots of dif­fer­ent ways we can help you:

  • We have a fast and friend­ly support desk who respond to most ques­tions within 24 hours.
  • You can also join our Face­book group for support from our com­mu­ni­ty of researchers.
  • Or book a slot in our Office Hours to talk to a Support officer 1−2−1.

Go ahead, give it a try. You’ll be sur­prised how easy and fun setting up your exper­i­ments is. And if you are still not sure wether it is going to be too tech­ni­cal­ly dif­fi­cult for you, then we have just the right piece of reading for you.

Jo Ever­shed

Jo is the CEO and co-founder of Caul­dron and Gorilla. Her mission is to provide behav­iour­al sci­en­tists with the tools needed to improve the scale and impact of the evi­dence-based inter­ven­tions that benefit society.