Experiment Tree Nodes


Overview


Welcome to the Experiment Tree Nodes Tooling Reference Guide.

Here you can find out information on how and when to use a particular Experiment Tree Node. There are 15 Nodes available browse the list of Nodes in the menu on the left to find out more information about each one.

Experiments are designed by combining the Questionnaires and Tasks you create into an 'Experiment Tree'. With the addition of Control Nodes which add powerful functionality such as Branching, Randomisation and Counterbalancing and you can create complex experiments with ease.


Tooling Reference Guide Format:


For each Experiment Tree Node listed you will find a description of the Node and an image example. This will be followed by a configuration settings box (like the one below) which will list and explain all available configuration settings for the selected Node.

Most Nodes offer this common configuration setting:

Configuration Settings:

Group

This is the name given to the path that your participants will be directed down if they match the criteria specified for this group.

It is a good idea to call it something meaningful you will be able to recognise later because this name will be written to the Metrics Spreadsheet under the heading of the Node-key (e.g. Branch-xyz).


A Nodes Key can be found at the very bottom of every Node in the Experiment Tree. You can also find it when you open up a nodes configuration menu (by double clicking on the node); it's located at the top, just under the Node title.

Click the Demo icon to view a Demo Project which demonstrates how you can use the selected Experiment Tree Node.

Click the Example icon to view an example of a Sample Project which makes use of the selected Experiment Tree Node.

Questionnaire Node


Questionnaire Nodes are green with a icon in the top left corner.

The Questionnaire Node has a single connection point and contains a single questionnaire within your experiment.

Once the participant has completed the questionnaire they will advance to the next Node in the experiment tree.

Image Example: Questionnaire Node

Configuration Settings:

Randomise questionnaire elements?

Default = No; No Randomisation will occur
i.e. widgets will appear in the order you have put them in your questionnaire.

Options: No, Yes, Yes(except first)

Here you can choose to randomise the order of the widgets in the selected questionnaire.
The questionnaire can be fully randomised [Yes] or have all widgets but the first randomised [Yes(except first)]. This later option is particularly useful if the first element in the questionnaire is a rich-text widget which provides the instructions to the participant.

Task Node


Task Nodes are blue with a icon in the top left corner.

The Task Node has a single connection point and contains a single task within your experiment.

Double-click a Task Node to configure its node-specific manipulations. These manipulations must first be set in the task itself. Each Task Node which uses the same Task but with different manipulations will have any differences show up within all related Task Nodes of the Experiment Tree.

Once the participant has completed the Task they will advance to the next Node in the experiment tree.

Image Example: Task Node

Configuration Settings:

Manipulations

See image example of the configuration settings above.

  1. Spreadsheet
    The Spreadsheet manipulation is the most common manipulation you will see available in your task nodes. In the dropdown menu you can choose from any spreadsheets you have uploaded.

  2. Other Manipulation by Name Any other manipulations you have set in your Task will appear here under the name you have given them. If you have also given them a default value and description (via the manipulations-tab within the task-builder) then this will appear here as well.

Start Node


Start Nodes are dark-grey with a icon in the top left corner.

The Start Node has a single connection point and marks the entry-points into your experiment.

The initial screen shown to participants when they enter your experiment via your start node will depend on which recruitment option you have selected. After this initial start screen, the participant progresses directly on to the Questionnaire or Task Node connected immediately after your Start Node.

As soon as a participant enters a Start Node they will be recorded on your experiments 'Participants' page as 'Live' and counted towards your total Recruitment Progress. If you are on a Pay-per-Participant account, a Token will be reserved for the participant at this point.

Using Multiple Start Nodes:
Most experiments will only need one Start Node. However, it is possible to have multiple Start Nodes.

If your participants are already divided into groups, you can associate each Group to a separate Start Node. This way you can use multiple Start Nodes to ensure each group sees a different version of your experiment.

Using multiple start nodes is one way of dividing your participant pool. You may want to use this method if you wish to display different tasks or tasks with different manipulations to your different participant groups. Participants can either be assigned to different groups by offering them different links, or by manually assigning them to a group. Use the recruitment options, Email Shot, Email ID or Supervised if you wish to employ this method.

Image Example: Start Node

Configuration Settings:

Group

Default = blank

Name of group participants will be matched to. Leave blank to match to any group.


Include at Start

[Advanced Option]
If checked, all recruited participants who enter the Start Node will immediately be Included in your downloadable data at that point, regardless of how far they progress through your experiment. Use this option if you wish to view 'Live' participants data as they are progressing through your experiment.

This means all participants recruited will count towards your recruitment progress and for Pay-per-Participant accounts a token is consumed immediately rather than at a Finish Node.

Checking this option will overwrite the functionality of the Reject Nodes, meaning regardless of being rejected they will be included in your downloadable data and for Pay-per-participant account a token will be consumed for each participant entering the Start Node.

Finish Node


Finish Nodes are dark-grey with a icon in the top left corner.

The Finish Node marks the end-point(s) of your experiment.

Once participants reach a Finish Node they can perform no further tasks or questionnaires. At this point the participant:

  • Will be sent to Gorilla's finish screen and are marked as 'Completed' on the Participants page.
  • Will be counted towards your recruitment progress.
  • Will be included in your downloadable data.
  • If you are on a Pay-per-Participant account their reserved Token will be consumed. (unless already consumed by a Start Node's advanced 'Include' option)

Using Multiple Finish Nodes:
Most experiments will only need one Finish Node. However, it is possible to have multiple Finish Nodes.

This can be useful if you want to have participants end at different points in your experiment; if for example they do not meet the requirements for your experiment or did not score high-enough to progress to the next stage of testing.

Image Example: Finish Node

Configuration Settings:

Onward URL

Enter a URL to redirect participants to this address once they reach the Finish Node. By Default, the participant is sent to the Gorilla finish screen.


Append PublicID

If set, append the participant's PublicID to the Onward URL above, along with the prefix-value assigned here. For example, if you enter the prefix-value here as id this will result in the participant being redirected to http://www.your-onward-url-above.com?id=PUBLICID

Use this option if your onward URL will need an identifier from the participant.


Show Completion Code

If checked, the participant will be shown their completion code. They can then use this code on third-party tools & software such as Amazon Mechanical Turk, to verify that they have completed the task.


Disable default behaviour

If this setting is checked, the default behaviour will be disabled.

If you are using Prolific, SONA or Qualtrics Panels you only need to enter your return URL in the above settings, because Gorilla will automatically send other required information back.

Default behaviours are the types of information that are passed back to your recruitment service, such as Survey Codes or Session IDs, or presented to your participant, such as Completion Codes.

The Third Party recruitment policy does not have default behaviours, so the above settings need to be configured manually.

If disabled, this information will not be sent automatically by Gorilla. Instead you must manually configure the settings above. Without this, you will be unable to validate which participants have completed the study or have been rejected, and so you may pay participants who have been rejected. If you are using a custom configuration, you may choose to disable this setting.

Click to view a Demo of using multiple Finish Nodes.

Reject Node


Reject Nodes are dark-grey with a icon in the top left corner.

The Reject Node marks an end-point of your experiment and functions similarly to the Finish Node.

Reject Nodes differ from the Finish Node in four ways: Participants who end at a Reject Node are by Default;

  • Marked as 'Rejected' on your Participants page.
  • Not counted towards you Recruitment Progress.
  • Not Included in your downloadable data.
  • For Pay-per-Participant accounts this means, the reserved token is Returned.

Note: If you have used the Start Node's advanced 'Include' option, this overwrites the functionality of a Reject Node and a token is consumed for every participant regardless of where they end your experiment.

Using Multiple Reject Nodes:
It is possible to have multiple Reject Nodes in your experiment.

This can be useful if you want to reject participants at different points in your experiment; if for example they do not meet the requirements for your experiment or did not score high-enough to progress to the next stage of testing.

Image Example: Reject Node

Configuration Settings:

Onward URL

Enter a URL to redirect participants to this address once they reach the Reject Node.

By Default, the participant is sent to the Gorilla finish screen. This is identical to the Finish Nodes default finish screen.


Append PublicID

If set, append the participant's PublicID to the Onward URL above, along with the prefix-value assigned here. For example, if you enter the prefix-value here as id this will result in the participant being redirected to http://www.your-onward-url-above.com?id=PUBLICID

Use this option if your onward URL will need an identifier from the participant.


Show Completion Code

If checked, the participant will be shown their completion code.

They can then use this code on third-party tools & software such as Amazon Mechanical Turk, to verify that they have completed the task.


Disable default behaviour

If this setting is checked, the default behaviour1 will be disabled.

If default behaviour is disabled, Gorilla will cease to send any information usually sent automatically to fascilitate integration. Instead you must manually configure the settings above.

Additional Information:
1 Default behaviours are the types of information that are passed back to your recruitment service, such as Survey Codes or Session IDs, or presented to your participant, such as Completion Codes.

Default behaviour of integrated recruitment policys enables you to automatically validate which participants have completed your study and which have been rejected. If you choose to disable default behaviour without including full set up details you may loose vital information which will mean you may need to pay participants who have been rejected.

Typically, if you are using Prolific, SONA or Qualtrics Panels you only need to enter your return URL in the above settings, because Gorilla will automatically send all other required information back.

However, if you are using a custom configuration or the general Third Party Recruitment Policy, you may choose to disable this setting.

Note: The Third Party recruitment policy does not have default behaviours, so all of the above settings may need to be configured manually.


Participant Inclusion

[Advanced Option]

Default = Exclude

Options:

  1. Exclude: Default
  • For Pay-per-Participant accounts reserved token is Returned.
  • Participant not included in your experiments downloadable data.
  1. Include:
  • For Pay-per-Participant accounts the reserved token is Consumed.
  • Participant is included in your downloadable data.
  • Participant counts towards your Recruitment Progress.

Some experimental set-ups require collection of 'rejected' participants data. In this specific case, you may want to include participants who end your experiment at a Reject Node.


Rejection Status

Select an option from the dropdown to set the rejection type you wish participants to reach this Reject Node to be set to. For some recruitment policies, this will affect the link that your participants are returned on.

Default = 'Rejected'; Participants entering this Reject Node will be marked as 'Rejected' on your participants Status Page.

Options: 'Rejected' or 'Rejected - Over Quota'

'Rejected - Over Quota': Selecting this option will result in participants entering this Reject Node to be marked as 'Rejected - Over Quota' on your participants Status Page.
Note: This option is significant if you are using the Research Now Recruitment policy and/or Quota Nodes.

Redirect Node


The Redirect Node is grey with a icon in the top left corner.

Redirect Nodes allow you to move participants between Gorilla and an external site, questionnaire or task and return to Gorilla. This can be useful for longitudinal studies or when you wish to use more than one research platform. For example: you can use the External Node when you want to link to a Qualtrics Survey, millisecond Task or a pre-hosted jsPsych or PsychoPy task.

You can even chain surveys from multiple sources together with your Gorilla Tasks and Questionnaire components, making use of Gorilla's advanced Control Nodes to, for example, randomise participants between your external tasks.

Link the Redirect Node into your tree at the point when you wish the participant to do a task or questionnaire which is hosted outside of Gorilla.

When a Participant reaches a Redirect Node, they are sent via a URL to the external site. The participant then completes the external task, questionnaire, or reads the documentation, and then returns to Gorilla to continue your experiment via a link.

Redirect Nodes only allow a participant to return to a Gorilla task if paired with a return link in the external resource. Otherwise, the participant will remain on the external site they have been redirected to.

When the participant follows the link and returns to Gorilla, they may have to log in again, for example with a public ID. The participant would then progress to the next Node in your experiment tree dependant upon the type of 'Completion' option choosen in your Configuration Settings (see below).

Note: You will need to use a recruitment policy which supports a participant returning into your experiment. The Simple Link recruitment policy can therefore only be used if you are using the 'Completion Token' method for the 'completion' configuration option.

Image Example: Redirect Node

Configuration Settings:

URL

Paste here the link of the external site that you want your participants to be redirected to.


Append Public ID

This allows you to identify the participant to the external site much like when linking to a recruitment service. Name this according to the requirements of your external site.


Append Embedded Data

This allows you to carry information about responses given by participants on Gorilla to the external site. You may wish to do this if you intend to manipulate what external task/questionnaire etc. is presented to participants.


Redirection

Select either ‘Redirect immediately’ or ‘Show a message first’. Redirect immediately means that as your participant finishes the previous node, they will be automatically redirected. Show a message first allows you to inform your participants of this redirection. Participants will be redirected when they press a ‘redirect’ button. When selected, the following configuration settings appear:

  1. Title
  2. Message
  3. Button Text

Enter your content as you would in the Task or Questionnaire Builders.


Completion

This controls what happens to participants when they return to Gorilla after being redirected. There are three different completion options:

  1. Default = ‘Immediate’,
  2. ‘Delay’
  3. ‘Completion Token’.

1. ‘Immediate’ is the default option. When participants return to your experiment after being redirected they will immediately move to the next Node in your tree.

Use the Experiment URL to link the participant back to your experiment. You will need to use a recruitment policy which allows a participant to return under the same ID.

2. ‘Delay’ prevents your participant from returning to Gorilla from the redirected site for the time period that you specify, after which they will be moved to the next Node.

If you select this option, additional configuration settings will appear:

  • Days
  • Hours
  • Minutes

Enter numbers into each field as required. Decimal numbers cannot be entered.

3. ‘Completion Token’ allows you to monitor whether the participant has completed the task/questionnaire that you have redirected them to.

Instead of linking participants back to your experiment using the experiment URL, you must use the Return URL AND include the completion token at the end of your external task/questionnaire.

When participants return through this link, they will register as having completed the task and automatically continue your experiment in the next tree node. If they do not return with the completion code they will be redirected again to your external site until they finish the task and are returned to Gorilla with the completion code.

If you select this option, another configuration setting will appear, named ‘Completion Token Name’. This is optional and allows you to change ‘completion_token’ in the URL to a name of your choice.

Checkpoint Node


Checkpoint Nodes are dark-grey with a icon in the top left corner. There is no screen associated with the Checkpoint Node; it is invisible to the participant.

The Checkpoint Node is useful for when you want to keep track of how far your participants have gotten through your experiment. When a participant passes through a Checkpoint Node this will be recorded in the metrics. The information is also available on the Participants page for your experiment: The name of the last checkpoint passed through by a participant will be listed under the 'checkpoint' column.

The classic use case for the Checkpoint Node is immediately after a consent questionnaire. When used in this way you can determine how many participants left your experiment because they did not consent.

Image Example: Checkpoint Node

Configuration Settings:

Name

Enter a name for the Checkpoint.

It is most useful if the name is meaningful. e.g. 'consent given', if used as a checkpoint after your consent questionnaire.

Delay Node


Delay Nodes are orange with a icon in the top left corner.

The Delay Node allows you to 'suspend' or 'delay' a participant from continuing your experiment.

It prevents a participant from taking part in the rest of your experiment (i.e. nodes which come after the delay node) until the specified time has elapsed.
The delay will be timed from when the participant first reaches the Delay Node.

This node is useful if you wish to have controlled time delays between different parts of your experiment. For example, between a teaching and testing phase, or, in order to measure performance at regular intervals.

Image Example: Delay Node
Image of delay node icon

Configuration Settings:

Title

This is the title of the dialog box which will be displayed to participants when they reach this node.


Message

This is the message which will be displayed in the dialog box which the participants will read when they reach this node.


Days

Here you specify the number of Days you wish to delay your participant for. This will delay them from taking part in the rest of the experiment until 00:00 (midnight) after the specified number of days has elapsed.

A 'day' is measured as starting at 00:00 (midnight) the next day. Thus if a participant enters your experiment and hits the delay Node at 23:50, and the delay is set to 1 day. It will be possible for that participant to start their next experiment phase within 10 minutes i.e. at 00:00 (midnight). If you wish for a participant to have a minimum number of hours delay use the 'Hours' Setting in conjunction with this.

If you wish your delay to be less than a single Day then leave this blank or type 0 here. Instead enter your time under the 'Hours' or 'Minutes' settings.


Hours

Here you specify the number of hours you wish to delay your participant for. This will delay them from taking part in the rest of the experiment until the specified time has elapsed.

If you wish to set a delay of 24 hours or less then you must leave the 'Days' field empty and enter the number of 'hours' in this field you wish for the duration of the delay.

If you are using a delay in 'days'. This is the number of hours of delay that will be added after 00:00 (midnight) is reached. Use this combination if you wish your participant to start at a specific time of day. i.e. a delay of 1 Day + 8 hours, will mean your participant will be prevented from taking part in your experiment until 08:00 the next day.

If you wish your delay to be less than a single hour then leave this blank or type 0 here. Instead enter your time under the 'Minutes' setting.


Minutes

Here you specify the number of minutes you wish to delay your participant for. This will delay them from taking part in the rest of the experiment until the specified time has elapsed.

If you have entered a delay time under the 'Hours' setting as well, then the two times will be added together.


Send Reminder

Note: This setting will only have an effect if you are using a recruitment option which uses the participants email address or else you use the 'Reminder Form' Setting.
If checked, a reminder email will be sent to the participant once your specified delay time has elapsed and they are able to continue.


Reminder Form

Note: The participant's email address will be visible on your participant dashboard, so you must ensure you have ethical clearance to collect this data.
If checked, display a reminder form that allows a participant to enter their email address. If they do so, they can then be sent a reminder if the 'Send Reminder' option is also enabled.

Click to view a Demo of the Delay Node.

Click to view a Classic Training Example of a Sample Project which makes use of the Delay Node.

Quota Node


Quota Nodes are orange with a icon in the top left corner.

The Quota Node allows you to specify a separate recruitment quota for a part of your experiment. Like the Start, Finish and Reject Nodes the Quota Node is partly configured on your Experiments Recruitment Page. Quota Nodes act like a ticket gate; only allowing participants to continue on with your experiment if there are enough spaces left in your quota.

Quota Nodes have two exit branches; 'ACCEPT' and 'REJECT'. By default participants will continue down the 'ACCEPT' branch of your experiment until your Quota limit is reached. Once your Quota is full, participants will be sent down the 'REJECT' branch.

How to set up a Quota Node:

  1. Add a Quota Node into your Experiment Tree flow
  2. Using the dropdown either select a Quota name from the list or select 'Create New...' to add a new quota.
  3. On your experiments Recruitment Page, set a recruitment target for each of your Quotas.

The Quota Node is typically used in conjunction with Branch or Randomiser Nodes when you require an exact number of complete participants per 'branch'. Usually the ACCEPT branch of the Quota Node will eventually end at a Finish Node. Whereas the REJECT branch of the Quota Node will often end at a Reject Node. View the image example below for a typical experiment set up using the Quota Node.

Note: Quota Nodes are used to control recruitment numbers for your experiment and therefore have the same effect on recruitment state as the overall Recruitment Target. For example, if all the Quotas for your experiment are full participants will be prevented from entering your experiment. Helpfully stopping you from overrecruiting participants! To allow more participants in your experiment simply increase your Quota size or else manually reject 'Live' Participants who have not completed to free up space in your Quota's.

Image Example: Quota Node
Click Here To view an image example of the Quota Node Configuration Setting
Configuration options available on the Quota Node:

Configuration options for Quota Nodes available on your Experiments Recruitment Page:
In the example above you can see that only 2 (StimuliSet1 and StimuliSet2) of the 4 possible Quotas are in use. Both Quotas are set to recruit 50 participants. The two other Quotas (YesQuota and NoQuota) have been disabled.
Click Here To view an set-up example for a Quota Node experiment.
The image below shows a typical experiment set up using the Quota Node:

The image below shows the Experiment Recruitment Page for the above experiment. Notice how the Quota's themselves have separate recruitment status and progress bars.
Click Here To view the Metric Columns generated by a Quota Node
The Quota Node adds 1 new data column to your metrics spreadsheet for each Quota Node passed through by the participant:
Column Name Description
quota-nodekey Will display the status (ACCEPT or REJECT) of the participant passing through the Quota Node.

ACCEPT means the participant was counted towards your quota and continued down the ACCEPT branch of your Quota Node.

REJECT means the quota was full when the participant reached your Quota Node and continued down the REJECT branch of your experiment.

Download an Example

Configuration Settings: Node

Quota

Select a name from a drop down list of possible quotas.

To create a new Quota select, 'Create New...' Option. Then type a name for your new Quota.


Configuration Settings: Recruitment Page

Recruitment Target

On the Rescruitment Page of your Experiment. Press the 'Change Recruitment Target' button to open the Recruitment Target Menu. Here you will be presented with two options for each quota:

1) Quotas

Type a number to set a Recruitment Target for each named Quota.

2) Enable/Disable

Click the button to toggle the status of a quota.

By Default all quotas start as Enabled.
This means participants who reach your Quota Node will continue down the ACCEPT branch until your Quota is full.

Click the 'Enable' button at any time to re-Enable a disabled Quota.
Recruitment for this quota will then continue from where you left off.

Click 'Disable' button to Disable a Quota.
Participants who reach a Quota Node which has been disabled will continue down the REJECT branch of your Quota Node until they complete your experiment - regardless of how many participants were previously recruited in this quota.

Note: Disabling a Quota will not remove the Quota Node from your experiment.
If you wish to stop using the Quota to control recruitment to a section of your experiment you should remove the Quota Node from your experiment on your experiments Design Page and Commit the new version of your Experiment.

Click to view a Tutorial of the Quota Node.

Repeat Node


Repeat Nodes are orange with a icon in the top left corner. Unlike other Nodes the Repeat Node is a dual Node system linked with arrows.

The Repeat Node allows you to specify a section of your experiment you wish your participants to repeat.

Participants who enter the bottom Repeat Node, will follow the arrow to the top Repeat Node. Which when linked in with the rest of your experiment tree enables the participant to repeat that section of your experiment.
Note: Make sure your Repeat Nodes are in the correct order - use the arrow as a guide - trace the path a participant will take through your nodes.

This node is useful if you wish to have participants repeat the same task or sequence of tasks/questionnaires. For example, to repeat a sequence of training task nodes in a training study.

Image Example: Repeat Node

Configuration Settings:

Repeats

Type a number for the maximum number of times a participant will see this sequence of nodes.

e.g. If you type 2, then the participant will see this sequence of nodes exactly 2 times in total. Once initially and then once when they repeat them.

Click to view a Demo of the Repeat Node.

Switch Node


Switch Nodes are orange with a icon in the top left corner.

The Switch Node, used in conjunction with the Switch Zone, or Switch Widget, allows participants to switch between two tasks, two questionnaires, or a task and a questionnaire during your experiment.
Note: The Switch Node will only work if you also use a Switch Zone within the tasks, or a Switch Widget within the questionnaires, tethered to your Switch Node.

Unlike other Nodes the Switch Node doesn't connect directly into the overall flow of the tree; instead, you 'tether' the Switch Node between two Tasks or Questionnaires. You can do this by dragging from the lighter areas, on each side of the Switch Zone, to the tasks or questionnaires you want to tether. One task/questionnaire (the primary task) will be linked directly into your tree and will be the first task the participant sees. The other task/questionnaire (the secondary task) will not be linked to your tree directly and will only be seen by the participant if they switch to it - via a Switch Zone button.

The Switch Node is also special in that it always requires both tasks/questionnaires connected to the Switch Node to contain either a Switch Zone, or a Switch Widget.

The Switch Zone grants the participant a button, within the task, which allows them to switch between the two tethered task or questionnaire nodes. You can find out more about the Switch Zone here.

The Switch Widget grants the participant a button, within a questionnaire, which allows them to switch between the two tethered task or questionnaire nodes. You can find out more about the Switch Widget here.

The Switch Node is useful in experiments where you wish to test a participant's task preference: e.g. see how long participants spend upon a maths task compared to a word-based task. It can also be useful when you wish to test comprehension of a document, e.g. see how many times participants refer to rich text in a questionnaire in order to answer questions about it.

Image Example: Switch Node
Click Here To view an image example of the Switch Node Configuration Setting
The Switch Node example below is set up to give the participant a maximum of 10 switches between the two tasks, with a completion criteria set to an overall 'Time Limit' of 30s (30000 ms):
Image of Switch Node Config
Click Here To view the Metric Columns generated by a Switch Node
The Switch Node adds 5 new data columns to your metrics spreadsheet:
Column Name Description
switch-nodekey-time-primary This is the total time (in ms) the participant spends on the primary task.
switch-nodekey-percentage-primary This is the time the participant spent on the primary task displayed as a percentage.
switch-nodekey-time-secondary This is the total time (in ms) the participant spends on the secondary task.
switch-nodekey-percentage-secondary This is the time the participant spent on the secondary task displayed as a percentage
switch-nodekey-switches This is a count of the total number of switches a participant made between the primary and secondary tasks.
Column Name Row Response Description
Response SWITCH This is the reaction time (in ms) at which the participant pressed the Switch Button.
Response SWITCH ELAPSED This is the total time (in ms) that the participant has spent on the task before pressing the Switch button.

Download an Example

Configuration Settings:

Max Switches

Type a number to specify the maximum number of times that the participant can switch between tasks.

Default = blank; infinite. Leave blank for unlimited switches.


Completion Criteria

Select from the dropdown the criteria for completing the switching task:

  1. Complete Both Tasks [Default]: regardless of number of switches the participant will be made to complete both tasks fully.
  2. Complete Primary Task: the participant will be made to complete the Primary Task.
  3. Complete Primary Task: the participant will be made to complete the Secondary Task.
  4. Time Limit: Once a set time limit is reached, regardless of participants progress in either task, the participant will be advanced to the next node, after the primary task, in the experiment tree.
    Note: you must set a time (in ms) in the 'total time limit' setting for this setting to work correctly. Note: Do not use this setting when switching between two questionnaires. This is because questionnaire metrics are recorded upon completion of a questionnaire. The Time Limit criteria means that neither questionnaire may be finished, meaning no metrics are recorded.

Total Time Limit

Note: This setting will only be applied if using the 'Time Limit' completion criteria.

Type a numerical value for the total (maximum) amount of time (in ms) a participant may spend across the two tasks.

Click to view a Demo of the Switch Node with Questionnaires (complete both).

Click to view a Example Experiment: Switch between Stroop Task and Instruction Questionnaire (complete primary).

Click to view a Example Experiment: Switch paradigm - switch between two different Tasks (time limit).

Click to view a Example Experiment: Switch between - a Task's Instructions and a Task (complete secondary).

Branch Node


The Branch Node is orange with a icon in the top left corner.

Branch Nodes allow you to direct participants down different paths (or branches!) of your experiment tree based on their responses to questionnaires or their performance in tasks.

Branch Nodes are evaluated as soon as the participant reaches them. The participant then advances to the next Node in the tree, on the branch that they are assigned to by the Branch Node.

In order to use the Branch Node effectively, you will need to make use of the embedded data function in the questionnaires and tasks that come before your Branch Node. You can learn more about Embedded Data here.

If you wish to direct your participants down different paths of an experiment randomly - but not based on performance - then it is better to use a Randomiser Node instead.

Image Example: Branch Node

Configuration Settings:

Group

Enter the name of the branch that the participant will be directed down if they match its criteria.


Property

This is case sensitive.

Enter the name of the property from a proceeding node that you want to test. This is the name of your Embedded Data that will be evaluated.

In the case of Questionnaires: this is the name of the 'Key' of the associated embedded data.

In the case of Tasks: this is the name given to a (setting) in the 'Correct Answer: Embedded data settings' when using Active Response Zones.


Rule

Select a rule from the dropdown menu to apply to the embedded data.


Value

This is case sensitive.

This is the value required to satisfy the rule. The value of the embedded data will be compared to this Value in accordance with the selected rule.

In the case of Questionnaires: this 'Value' will be one of the corresponding 'answers' offered in your questionnaire configuration associated with the key above.

In the case of Tasks: there are three possibilities for the 'Value' depending upon which embedded data setting you are using to branch from.

When branching from 'Correct Answer: Embedded data settings':

  1. 'most recent answer': The 'Value' will be the expected name or value stored in the most recent answer.
  2. 'correct/incorrect/total answer count': The 'Value' will be a numerical value relative to your choosen 'answer count'.
  3. 'percentage correct answers': The 'Value' will be a numerical value between 0 and 100.

Default

When checked this branch is the default branch for progression. Participants who do not match any rules in the branch node will be assigned to the default branch.

Click to view a Demo of the Branch Node.

Randomiser Node


Randomiser Nodes are orange with a in icon the top left corner.

The Randomiser Node allows you to distribute participants at random between 2 or more different paths through your experiment tree.

Randomiser Nodes are evaluated as soon as the participant reaches them. The participant then advances to the next node in the tree, on the branch that they are assigned to by the Randomiser Node.

There are two parts to setting the Randomiser Node.

  1. The Group Allocations allows you to give a name to each group and assign a ratio to the group. Additional groups can be added using the 'plus' button.
  2. The Randomisation Mode allows you to decide how the randomisation will occur.

For example, consider a situation where you wanted to have two groups, one for an easy branch and the other for a hard branch, with a ratio of 2:1 respectively. The first group can be given the name 'easy' and a ratio of '2'. The second group can be given the name 'hard' and a ratio of '1'.

Note: Randomiser Nodes have no knowledge of subsequent attrition.

Image Example: Randomiser Node
Attrition and the Randomiser Node

The Randomiser Node has no knowledge of subsequent attrition.

To take an extreme example, imagine you have two groups, one with a positive mood induction (fluffy animals) and one with a negative mood induction (aversive images). You'd expect a lot more attrition in the aversive stimuli group. Consequently, you might want to run 1/3 of the experiment at 4:4, work out the attrition rate and then change it to 4:8 or 4:12. This would allow you to compensate for the asymetric attrition without removing random assigment. If you had set the randomiser to 50:50 and had quickly collected all 50 in the positive group, your choice would then be to assign people to the negative group, or to over recruit in the positive group. Neither would be ideal.

Configuration Settings:

Group Allocations

  1. Group

Enter the name of the branch that the participant will be directed down if they are assigned to this group.


  1. Ratio
What ratio do I need? (Click here)

The best way to think about the balanced randomiser is like a deck of cards. The colour of the card (red or black) determines which branch you go down.

  • 2:2 means you have 4 cards, 2 are black and 2 are red. They get handed out in a random order (BBRR). Then picked up, shuffled and dealt out again (BRBR, RRBB, RBBR).
  • So 2:2 is different from 1:1. With 1:1 there are only 2 cards, 1 red and 1 black, so there are only two possible orders (BR and RB). Some reviewers don't like this because they think it's assignment. It's not, but it's not worth the argument!
  • 10:10 means you have 20 cards, 10 red and 10 black. You could (by chance) have a run of 10 red followed by 10 black, then 10 black and 10 red. With a high ratio, you may have a balanced set of participants at only 20 and 40 Ns.

Consequently, We wouldn't recommend putting the exact number in the randomiser (i.e. 30:30), but rather instead go for something like 2:2 or 4:4.

Ratio of participants that will be assigned to this branch of the Randomiser.

The exact mathematics of how the ratio used depends on the randomisation mode (see below), but the proportion of participants assigned to any one branch will always be equal to that branch's ratio divided by the sum of all ratios.


The following settings apply to the node as a whole:

Randomisation Mode

  1. Balanced means that participants are allocated in the above ratios in sets of the total ratios.

[Default]
So for two branches with ratios 10 and 10, for every 20 participants, 10 will get the first branch and 10 will get the second. This is random without replacement.

  1. Random means completely random, with each branch having the probability of its ratio divided by the total ratio.

Therefore, two branches with ratios 1 and 1 is identical to two branches with ratios 50 and 50. This is random with replacement.

Click to view a Demo of the Randomiser Node.

Order Node


Order Nodes are orange with a icon in the top left corner.

The Order Node allows you to randomise the order of Task Nodes and Questionaire Nodes within your experiment tree.

Order Nodes are special in that they are not part of the overall flow of the tree; instead, you 'tether' Task and Questionnaire Nodes to the Order Node, allowing it to control their order. You can do this by dragging from the lighter area of the Order Node to the task or questionnaire you want to tether.

Image Example: Order Node

Configuration Settings:

Randomisation Mode

Select a Randomisation Mode from the dropdown menu:

  • Latin Square uses a standard Latin Square design to ensure that all possible permutations of task order are as equally represented as possible.
  • Balanced considers all possible permutations of task order that will be used.

In both cases, as far as is possible, Gorilla will attempt to assign the same number of participants to each permutation.

Click to view a Demo of the Order Node.

Counterbalance Node


Counterbalance Nodes are Orange with a icon in the top let corner.

The Counterbalance Node can be used in conjuction with the source type spreadsheet-manipulations to assign different stimuli sets to each participant.

For exampl, if you have 4 stimuli sets, you could set up a Randomiser node with 4 paths, where each path goes to an instance of an instance of your task set up with a different spreadsheet, such that each task instance will show a different stimuli set. With 4 stimuli sets this is doable, but if using any more this set-up becomes tedious. Instead, a counterbalance node can be used to achieve the same end, but only needing one instance of the task.

This is particularly powerful method if you want to run large numbers of participants (1000) covering a large number of stimuli sets.
If however you wish to make use of the quota node for each stimuli set, then you will need to use the Randomiser Node to individually set a quota for each branch option.

How to set up the Counterbalance Node

  1. First, be sure to have set up your Task correctly using the spreadsheet-manipulation source for your stimuli set.
  2. Secondly, in the spreadsheet of your Task, be sure to set up all your required spreadsheet-manipulation stimuli set options as separate columns.
  3. In the Task Node following your Counterbalance Node enter '$${your_Counterbalance_Node_Config_name}' into the Manipulation setting on your Task Node.
  4. Finally, in your Counterbalance Node, list your stimuli set spreadsheet-manipulation column names under the Value option in the Counterbalance Node configuration menu. Each option should be separated with commas. Note: Remember that names are case-sensitive. This will assign each participant to each spreadsheet manipulation in turn.
Image Example: Counterbalance Node
Click Here To view an image example of the Counterbalance Node Configuration Setting
1) Image of Configuration Options of the Counterbalance Node:
2) Image of Configuration Options of the corresponding Task Node set up using spreadsheet-manipulations:

Configuration Settings:

Name

Set the name for your Counterbalance Node. This name will be used as the embedded data in the spreadsheet manipulation of your task.


Values

List the names of your spreadsheet columns used in your spreadsheet manipulation, separated by commas.

Click to view a Demo of the Counterbalance Node.