When creating an online survey, it is important to take into account that respondents can also access the survey via mobile devices (e.g. smartphones and tablets). In 2021, almost 60% of 1KA's respondents completed the survey via smartphone.

Therefore, when designing the questions in the survey, mobile respondents should be taken into account, both in the actual writing of the questions and in ensuring that the questions are displayed correctly regardless of the type of device used by the respondent. If surveys are not adapted to mobile devices, there may be problems related to data quality, e.g. incorrect recording of responses, satisficing, respondents skipping questions, etc. If respondents become dissatisfied, mobile respondents are much more likely to leave and not complete the survey.

To avoid these problems, PEW has developed eight recommendations for creating a survey that is also mobile-friendly:

1. Software must be optimised for mobile devices

Mobile optimisation means that the software automatically detects the device used by the respondent, in particular the screen size, and adjusts the layout of the survey accordingly. Fonts, spacing and buttons should be large enough so that respondents do not need to enlarge the screen. There should also be no horizontal scrolling, but vertical scrolling is acceptable.

2. Shorter is better

This refers to both the number of questions and the scope of the question. The longer the survey, the more likely respondents are to disengage - this is true for all types of surveys, but even more so for mobile surveys. This is because of the small screen sizes of smartphones, which can be solved by shorter questions and answers, which will make it easier for respondents to read and complete the survey, and should improve the quality of the data itself.

3. Avoid more advanced features

When creating a survey, it is important to try to keep it simple and not complicate things by including more advanced functionality such as sliders and the like. Research shows that respondents on mobile devices have difficulty using these features correctly as they require a high level of skill. It can also take more time than simpler formats. Thus, radio buttons, checkboxes or text boxes are the most recommended if possible.

4. Avoid tables

Surveys presented in table form are designed to make efficient use of space by formatting questions and response categories in a table or matrix. However, the use of tables on all devices has certain drawbacks. For example, researchers have found that with tables, respondents often choose the same, usually middle, answer for each item in the table, leading to non differentiation, also known as straightlining. This data quality problem is even more pronounced on mobile devices, as tables often require horizontal scrolling in addition to vertical scrolling, which can mean that all questions and answers may not be visible at the same time. This may cause respondents to leave the survey.

To avoid problems associated with tables, many online survey tools, including 1KA, allow you to optimize the table question type for mobile devices. More >>

5. Multiple questions on the same screen

As tables are not ideal for mobile surveys, it is better to ask multiple questions on the same screen. In the past, it was popular for online surveys to have each question on its own page or screen. However, due to slower page loading on mobile phones, this can put unnecessary strain on respondents. This can be avoided by putting questions about the same topic on the same page or screen.

6. Making the most of the smartphone screen

Smartphones have limited screen space, which makes it even more valuable. Avoid embedding logos, advertising and graphics (which also take longer to load), as this can saturate the screen. Navigation buttons should be at the bottom of the screen to force respondents to scroll past all the questions and answer options; this way they do not miss anything.

7. Survey invitations should contain a unique URL

Respondents should not be required to enter an access code or password to complete the survey. This creates another barrier to completing the survey, which is especially true for respondents on smartphones, as entering the code means that they will have to switch between the survey invitation received in their email inbox and the mobile web browser. For a better user experience, it is preferable to use individualized URLs for each respondent, so that when the respondent clicks on the URL in the invitation, the survey will open directly and the respondent can start completing the survey right away.

8. Invite respondents to complete the survey by SMS

In case we have the explicit consent of the respondents to send text messages, in addition to invitations by email, we can also send an invitation to complete the survey by SMS. We include the URL to the survey in the message and the respondent just clicks on the URL in the message.

For further reading on this topic, see the report by the American Association for Public Opinion Research (AAPOR).

An online survey created using the 1KA tool can also be answered by respondents via a mobile phone. Due to the smaller screen, certain adjustments are required to optimize the survey. For more advanced and new types of questions, or more advanced layouts of basic types of questions (such as image hotspot, slider, combined table, etc.), it may occur on mobile devices to have some offsets or opacity. In this case, we recommend using the basic types of questions and / basic layout questions.

FEATURES IN MOBILE PHONE ADJUSTMENTS

Category - single answer and category - multiple answers (basic vertical layout)

The basic questions, category - single answer, and category - multiple answers, are generally not problematic and are shown in the survey on a mobile device in such a way that they enable the survey to be completed normally. This applies to the following layouts: vertical, vertical - left text, drop-down menu, select from the list, image type.

For both question types and different layouts, the entire answer box is coloured when you click/tap on the selected answer (the entire box is also clickable).

Category - single answer and category - multiple answers (other layouts)

For one-answer and multiple-answer questions, layouts are also available that may appear unmanageable on mobile devices, so the layout selected on mobile devices is ignored and the question is displayed in a normal layout (vertical). This applies to the following layouts: horizontal beside the question, horizontal under the question, drag-drop.

Table

Table type question and all subtypes and table type question setups are automatically displayed on mobile devices as a set of basic question types. For example, if we have a question table - single answer with 5 different sub-questions, the question will appear on mobile devices in the form of a set of 5 consecutive questions of category type - single answer.

Example of a survey with a custom view of a table type question (mobile preview) >>

You can also display tables in the form of automatic expanding in the tab 'EDIT' - 'Settings' - 'Mobile settings'.

Example of a customized view of a table type question - auto-expand (mobile preview) >>

CUSTOMIZATION OF THE DOUBLE GRID TABLE QUESTION TYPE

For the question type 'Double grid table', the answers on the mobile phone appear as a series of questions of category type - single answer. For the sake of transparency, there is a horizontal continuous line between the individual answers on the cellphone and a dashed line between the individual tables.

Example of a survey with question type double grid table (mobile preview) >> (for comparison, see how layout types look on the web version of the survey)

DISPLAY OF THE TABLE WHERE MOBILE CUSTOMIZATIONS ARE OFF

If you turn off mobile customizations (see the last paragraph), you can also view the spreadsheet on mobile devices in the same format as the web, or use any of the step-by-step questions.

When editing advanced question types and different table layouts, in the 'EDIT' - 'Questionnaire' tab, to the right of the 'Advanced' tab, there is the 'Step by step mode' option. If you select 'NO', the question will be automatically adjusted as a consecutive series of questions of category type - single answer (or multiple answers) when answering by mobile phone.

There are 6 different types of Step by step completing that you can see at the link below.

Example of a survey with different forms of step by step answering mode >>

Example of a survey with different forms of step by step answering mode (mobile preview) >>​

Input questions (text and number)

On a mobile phone, questions with an input field (question types 'text' and 'number'), buttons for questions of type roulette and selection list are automatically adjusted to 100% screen width.

Drag and drop question

You can choose the drag-and-drop layout for different types of questions: category single answer, multiple answers, table. For single or multiple-choice questions, the drag-and-drop layout automatically adjusts to the default vertical layout when completing the survey via mobile. Table type question - Drag-and-drop layout appears on your mobile device as a classic (mobile-friendly) question.

Questions with pictures

By default, if you have images in the survey, they are displayed by default when responding to the survey via a mobile device, and they are adjusted to occupy 80% of the screen width.

Question with inserted video

If you inserted a video in the questionnaire, the width of the video will adjust automatically on 100% of the width of the mobile device. Respondents will be able to see the full video, no matter the width of the device they’re using for filling out the questionnaire.

Example of the survey with embedded video (mobile preview) >>

ADDITIONAL ADJUSTMENTS

Hiding radio and checkbox buttons

With the type of questions such as 'Category – single answer' and 'Category – multiple answers' there is an extra option 'Hide radio button' or 'Hide checkbox button' in the 'Advanced' tab. Functionality is recommended for mobile devices, especially when you're using pictures for answers. In this case, you can hide these buttons (radio or checkbox buttons), respondents which will be filling out the questionnaire on their mobile device will be able to select pictures as an answer with the tap on it. It will be easier for them to choose the answer, as it would be selecting the 'checkbox' or 'radio' button, where you have to be more precise.

Example survey with hidden radio and checkbox buttons (mobile preview) >>

Adjusting the size of the radio and checkbox buttons

With question types 'Category - single answer' and 'Category - multiple answers' additional options of changing the size of radio and checkbox buttons are available to 2KA and 3KA private package users. These options can be accessed in the tab 'EDIT' - 'Design' - 'Theme modification', under 'Custom icons', 'Icons size'. Minimum size of buttons is 18 x 18 px and maximum size is 55 x 55 px. We especially recommend the use of this feature when designing surveys for mobile devices.

Example of a survey with a default size of buttons

Example of a survey with a modified size of buttons (40 x 40px)

Preview for mobile phone and tablet

You can use the preview to see how the survey will appear on different types of devices (computers, mobile phones or tablets). In addition to a basic PC preview, you have an option of a survey preview for mobile phones and tablets. More >>

THE OPTION FOR SWITCHING OFF THE ADJUSTMENTS FOR MOBILE DEVICES

All of the mentioned adjustments are automatically turned on; this way we can allow the optimal display of the questionnaire to those who use their mobile devices. In tab 'EDIT' – 'Settings' – 'Mobile settings' you can change these adjustments, which we dissuade. When you turn off the mobile adjustments the crossed-out mobile phone icon will appear in the menu line on the upper right corner (beside the icon for turning on comments in the questionnaire).

In case when 'Mobile friendly' display is turned on, you can separately turn off the display of images on mobile phones (if your questionnaire contains pictures that are not suitable for mobile devices) and adjust tables on the mobile phones.

In the 'EDIT' – 'Questionnaire'' tab, in the settings of the 'Single answer table' question type, you can select the option to customise the table, which means that you can set each sub-question in the table to be displayed separately to the respondent. The respondent will automatically move to the next sub-question by selecting an answer for the current sub-question. Alternatively, the respondent can scroll through the sub-questions themselves (depending on the customisation choice).  

To enable this option, select the desired customisation of the table on the computer and on the mobile device in the settings of the 'Single answer table' question type, on the right-hand side, in the 'Advanced' tab, under 'Properties'. In addition to the default setting, you can choose between the following customisation modes: Grid, Scroll, Unfolding and Horizontal Unfolding.

This option is recommended especially in the case of respondents completing the survey via mobile devices. If you keep the default settings, the question will be automatically adapted to the 'Scroll' format when respondents answer it via mobile devices.

The link below shows examples of all types of customisation of a table question type:

Example of a survey with different options of step by step survey completing >>

The functionality is available for users of the 2KA and 3KA packages for individuals, and for users of the business packages for groups.

Where you have questions with a large number of possible answers, you can use the drop-down question type to create sets of connected questions (drill-down), where every following question is conditionally displayed according to the answer given to the previous question.

For example: if we are asking about a car model, the respondent first has to choose the country of the manufacturer, followed by the brand of the car, and based on that answer, they choose the model. This can be displayed in the form of consecutive drop-downs in one row.

How to set sequential filtering of questions? (Example: Car Model)

• Insert questions: e.g. (Q1) Country of car manufacturer, (Q2) car brand and (Q3) model of car.
• Questions have to be of the 'Category – Single answer' type, then we select the 'Drop-down' layout in the 'Settings' menu positioned on the right.  
• Insert all possible answers or categories.
• Go to the editing of question Q2, where you set the condition for every category (car brand) based on Q1. Do the same for question Q3, where the condition is set based on the category in question Q2.
• A condition for the display of a category is set by clicking on the IF* icon on the right side of a category. In the 'pop-up' window, select Q1 or Q2 as a base for selection and choose the appropriate value. Repeat for all categories.
• Then all three questions should be put in a block, where we choose a horizontal question layout. We can also choose the standard (default) layout.
• If the questions are not structured in a block, the drill-down filtering will work exactly the same way, but the questions will be placed independently - vertical, with their own label and with the normal spacing of questions. All three options (questions structured in a block - horizontal and vertical, and questions not structured in a block) can be viewed at the link below.

Example of a survey with sequential filtering of questions >>

The functionality is available for users of the 3KA package for individuals, and for users of the business packages for groups.

For all question types where you have multiple categories or answers, there is an option to hide the individual categories or answers. You can hide a category or answer so that new respondents no longer see it in the survey, without losing the data you have already collected. If you delete a category or answer, the data already collected is deleted from the database, so we recommend the hide function for such cases.

This option is particularly recommended for event registrations, individual workshop registrations, product orders and other cases where, for some reason, one or more of the answers is no longer relevant (e.g. a product is out of stock or a workshop is full).

By clicking on a specific question, you can access additional settings on the right side for each category, such as advanced editing, conditions, delete category and as described above – 'hide category' (icon in the shape of crossed-out circle). When you hide a category, the icon will turn red.

Example of a survey with a hidden answer feature >>

For all question types containing radio buttons or checkboxes ('Categories - Single answer', 'Categories - Multiple answers', 'Tables - Single answer' and 'Table - Multiple answers'), there is an additional option to change the size of the radio buttons and checkboxes. The option is available in the ‘EDIT' - 'Design' tab when editing a theme (Theme customisation). In the 'Radio buttons and checkbox' segment (PC theme customisation) or 'Mobile phone custom icons' segment (Mobile theme customisation), you can choose any size of radio buttons or checkboxes (from 18 to 55 px for PC and from 20 to 55 px for mobile).

The selected setting is applicable for the whole survey (all questions with radio and checkbox buttons), but not to a device used for accessing the survey. Different icon sizes can be chosen for surveys shown on personal computers (and tablets) and surveys shown on mobile phones. This functionality enables us to, for example, only enlarge icons on mobile phones, where this is recommended because of smaller screen sizes, and not on personal computers and tablets, where icon size remains the same.

Example of a survey with the default size of radio and checkbox buttons (minimum size 18 px) >>

Example of a survey with a customized size of radio and checkbox buttons (maximum size 55 px) >>

The functionality is available for users of the 2KA and 3KA packages for individuals, and for users of the business packages for groups.

For 'Categories - Single Answer' and 'Categories - Multiple Answers' questions, you can enable advanced image upload as an answer, so that you can upload an image to be displayed as an answer with a single click.

To turn this option on, click on the toolbar of the question for which you want to set this. The question settings will open on the right-hand side.

In the 'Basic' tab under 'Answers', check the 'Image as answer' option and icons to add an image will be shown on the left-hand side before the answer text.

Clicking on the icon will open the 'Image Properties' pop-up window, where you can enter the URL of the image you want or click on the 'Choose file' button to upload the image from your computer.

When uploading an image from your computer, a new window will open. Select the desired image from the directory on your computer and then click 'Open'. In the 'URL' input field, the URL of your image that has been uploaded to the 1KA server will be displayed. You can also set the alignment and size of the uploaded image as you wish. Add it to the answer by clicking on the 'OK' button.

When editing a questionnaire in the 'EDIT' - 'Questionnaire' tab, 'Advanced settings' are available above the questionnaire.

Find and replace words

You can easily replace words that appear repeatedly in the questionnaire. In the first field, enter the word or phrase you want to correct, and in the second field, enter the word or phrase you want to replace the first one with.

To search the questionnaire only, use the CTRL + F command.

Add page break between all of the questions

If you want respondents to see each question on its own page, you can add a page break between all questions in your questionnaire with one click.

Check the correctness of the conditions

To help you correct any errors, there is an option to 'Check the correctness of the conditions'. This feature is particularly useful if you use conditions and loops in your survey. By clicking on the bug icon, the application will warn you in a pop-up window if the survey contains errors and also where the error is located.

The application will also warn you of errors if you want to activate a survey that contains errors in the terms and conditions, without you checking that the terms and conditions are correct. More on this in FAQ >>

Using conditions is only available for users of the 2KA and 3KA packages for individuals, and for users of the business packages for groups.

Enumerate

The numbering of questions is in principle generated automatically in order (Q1, Q2, Q3, etc.), but moving or deleting questions can break the order. By clicking on the enumerate option, you can restore the correct order. This does not apply to variables which you have previously changed the names of. More on this in FAQ >>

Display collapsed view

Additional icons for the collapsed and expanded view can be found on the right-hand side.

In the collapsed view, only questions without answers are shown, which enhances transparency. Transparency is especially useful when questionnaires are complex and long. It is recommended to use the collapsed view for as few as 30 questions. You can switch between the collapsed and expanded view at any time (from one icon to the other).

If you have structured your questions into blocks or used loops, you will also see the options 'Open all blocks and conditions' and 'Close all blocks and conditions' under 'Advanced settings', which allow you to open and close your question blocks with a single click.

Structuring questions into blocks and some of the functionalities described are only available for users of the 2KA and 3KA packages for individuals, and for users of the business packages for groups.

In the ‘Edit’ – ‘Questionnaire’ tab, additional editing options are available in the question taskbar.

In the taskbar, you can:

• Open the question editor by clicking on the pencil icon;
• Add conditions by clicking on the IF icon (Conditions are only available to users of the 2KA and 3KA packages for individuals and to users of the business packages for groups.);
• Copy the question;
• Open the question preview;
• Delete the question.

For question types Category – Single answer, Category – Multiple answer and Table, the option of choosing the measurement scale type is also available:

• ordinal;
• nominal.

You can read more about the differences between ordinal and nominal scales here >>.

Some of the functionalities described are only available for users of the 2KA and 3KA packages for individuals, and for users of the business packages for groups.

For each variable, you can limit the time available for entering a response. This can be done in the 'Advanced' tab in the question settings (in the 'EDIT' – 'Questionnaire' tab), where you can select 'Time limit for answering (Timer)' and set the display time for that question. Only one timer can be set for each page. When the time is up, the survey will be redirected to the next page automatically, regardless of whether there are other questions that do not have a timer on the previous page. We therefore recommend that if a timer is set, you add a page break after each question.

Pending the next application update, it is currently not possible to utilize the timer feature in cases where the timed page is followed by a page with a question that's not visible to all respondents (eg. due to group randomization or other IF conditions). Depending on the individual survey design and methodological needs, consecutive timed pages with separate conditions may instead be interspersed with pages visible to all respondents (in cases where question order is not imperative), or all timed questions within conditions can be grouped on one page (in cases where each respondent is only shown one question out of a pool of multiple questions).

If you wish to measure the answering time for each question, you must also create a separate page for each question, since time is only measured for each page. You can enable paradata in the 'DATA' – 'Browse' tab ('+ Show columns' - 'Respondent paradata') to record, in addition to the survey data, the date and time at which the respondent arrived at a particular page of the questionnaire.

In the table with the respondent units, we then have the date and exact time of completion according to the individual pages of the survey:

• Date:
  • Begin: the exact date and time (hours, minutes and seconds) when the respondent opened the survey page (click on the URL of the link to open the survey);
  • End: The last time the respondent was active on the survey, excluding the outro page;
• Page 1: date_1: Exact time when the respondent clicked on the "Next page" button on page 1;
• Page 2: date_2: Exact time when the respondent clicked on "Next page" button on page 2;
• Page n: date_n (n = number of pages): Exact time when the respondent clicked on the "Last page" button on page n;
• The time of the outro page (end of the survey) is not visible.

The average page times are listed in the 'TEST' - 'Duration' - 'Actual time'.