How do I use the Tasks tab in Website Testing?

In Website Testing, the respondents try to complete a set of tasks that you give to them. In this help, we’ll discuss how to use the Tasks tab to:

• Create tasks for your respondents to fulfill
• Set up calculation of the Lostness metric
• Adjust the order and number of tasks that the respondents see
• Create post-task questionnaires
• Import and export your tasks in a CSV file
• Set up Crowd Feedback

Create tasks for your respondents to fulfill 

• The tasks that you create should represent specific use cases of your website or app
• To write a task, simply open a task and fill in the task text
• Set the Start URL – the URL of the initial screen for this task
• Set the End URL – the task is considered successfuly fulfilled if the respondent ends the task on this exact URL
• Set the length of the optimal respondent’s path for the lostness metric – This setting is optional. Look at the section below this one for more instruction on how the lostness metric works and how you can set it up.
• To insert more tasks into the Website Testing study, click Add task
• By default, respondents can’t skip tasks. You can enable them to skip tasks by checking Allow respondents to skip tasks

Write your tasks

Each task needs some text to be given to the respondent as an assignment, a starting URL and an ending URL.

Set up calculation of the Lostness metric 

The Lostness metric measures the effectiveness of navigation on your website. It tells you how lost respondents get when they use your website (when they solve your tasks). It’s based on the difference between the number of pages visited by the user and the minimum number of pages (the optimal path) that need to be visited to solve the task.

You can set up the length of the optimal path for each task here. This setting is fully optional, however setting it properly will lead to lostness being calculated more accurately. If you don’t set the length of the optimal path, the shortest path found in task recordings will be used to calculate lostness automatically. In practice, this means that the path of all respondents might not be compared to the actual optimal path, but rather the path of the respondent who traversed the least pages (who may or may not have also been lost).

Enter the minimum number of pages from the start URL to the success URL (start and success pages included). Here are some examples:

• If the start and end URL are the same (the task can be solved on one page) the optimal path is 1
• If a link can take you directly from the start URL to end URL the optimal path is 2
• If the optimal way to solve a shopping task is to buy a product on the homepage (start URL – page 1), go to the shopping cart (page 2) and click Checkout which leads to the order confirmation (end URL – page 3) the optimal path is 3

Since you will become unable to save changes to the optimal path once you launched your study, we recommend setting the length of the optimal path for your tasks even if you’re currently unsure if you’re going to make use of Lostness or not.

Adjust the order and number of tasks that the respondents see 

• Unless the order of tasks is randomized, the respondents solve tasks in the order they’re listed in
• To move a task up or down on the list, simply drag-and-drop it to the desired destination. Alternatively, use the Move up/down buttons under the ellipsis button
• To show the tasks to respondents in different order, select Randomize task order
• To keep the first task static (e.g. to use it as an example), check Don’t randomize the first task
• If the order of tasks is randomized, by changing Number of tasks to show to a respondent , you can show every respondent only a certain number of tasks from the total number of tasks available

Edit task options

Here, you can customize how respondents interact with tasks – whether respondents can skip tasks and whether the order of tasks should be randomized.

Create post-task questionnaires 

• Aside from the surveys in the Questionnaire tab, you can also ask questions immediately after each task
• This is recommended for gathering immediate feedback to prevent important information about solving individual tasks being mixed up of forgotten
• Question creation works the same as in the Questionnaire tab

Accompany your tasks with a survey

Ask your respondents additional questions after they complete a task.

Import and export your tasks in a CSV file 

Above the task editor, there is an option to import and export tasks stored in a .CSV file. Use this option to easily reuse your tasks between Website Testing studies, or if you want, to bring in tasks that have been created somewhere other than in UXtweak.

Use UTF-8 encoded CSV files to ensure that the import works correctly. The simplest way to save your imported file with UTF-8 encoding is to edit it through Google Sheets (click the link for more instructions). To get a UTF-8 encoded file when saving your file in Excel, go to File -> Save As and select CSV UTF-8 before you click Save.

Only the text, the start URL and the end URL get imported and exported within the tasks. The post-task questionnaire has its own Import and Export buttons.

Set up Crowd Feedback 

Use the power of the crowd to efficiently generate ideas regarding your product. Crowd Feedback is an apparatus through which the first few respondents submit their Likes, Dislikes and suggested Improvements. Later repondents then vote for which ideas they agree with, making popular ideas raise to the top. When respondents vote, they can also leave an additional comment to explain their decision.

• By default, Crowd Feedback is turned off. To activate it, select whether you want it Run after study, Run after each task or Run after selected task. If you choose after selected tasks, you can select individual tasks to run Crowd Feedback after in the dropdown list, or by checking Crowd Feedback enabled after this task in a task’s options.
• Decide whether respondents can submit and vote on ideas, or only vote. In vote-only mode, respondents only vote on ideas that you pre-defined, without the ability to submit their own ideas. There needs to be at least one pre-defined idea if you wish to run Crowd Feedback in vote-only mode.
• Crowd Feedback can be used to generate ideas for three categories: Likes, Dislikes and Improvements. By default, they will all appear if Crowd Feedback is enabled. However, you can disable either feedback category by unchecking the checkbox next to it.
• Click Pre-define next to a feedback category to create Likes, Dislikes or Improvements that respondents will be able to vote for. Simply write the text of the idea that the respondents will see and click Add. If you run Crowd Feedback after tasks, you also need to select which tasks the pred-defined idea belongs to. To remove a pre-defined idea, click the trash bin icon next to it.

In advanced options, you can further customize how Crowd Feedback is collected. If you’re unsure how to use this option, the default settings have been selected to be suitable for the majority of projects.

• If you’ve given respondents the ability to submit their own ideas, you can Limit number of ideas all respondents can submit in total. Once the number of total ideas is reached, further respondents will still be able to vote on ideas, but not able to submit new ones any longer. Its value can be set for Likes, Dislikes and Improvements separately. The default value is 10.
• Similarly, you can also Limit number of ideas a respondent can submit (Increasing this number allows creative/talkative respondents to submit more ideas, but your ideas will also come from less people). Adjustable for Likes, Dislikes and Improvements separately. The default value is 2.
• Check Ideas required so respondents have to submit the determined number of ideas before they can continue on with the study.
• You can Limit number of ideas a respondent can vote for. You can let them vote for just one idea, without restrictions or for percentages (25%, 50% or 75%) of total ideas, rounded up. The default value is 25% of total ideas.
• Respondents can vote for ideas with Simple check marks or a 5-point or 7-point Likert scale. With the Likert scale, respondents determine how much they agree with a given idea – from Slightly agree to Strongly agree. Use the Likert scale to get nuance about how much respondents identify with ideas.
• Check Voting required so respondents have to vote for the determined number of ideas before they can continue on with the study.