Question Types

This page describes which question types you can use in SoSci Survey and what to consider when using them.

Most question types offer numerous options for further settings. Thus, many question types also allow a display that differs considerably from what is shown here.

You will find an overview of all question types and templates directly in SoSci Survey when you create a new question. Below you will find some additional notes on the question types – and tips on how to design questions that go beyond these defaults.

SoSci Survey is continuously being developed further. If you have further ideas for question types, please describe them briefly in the online support.

A selection offers different options to choose from and allows selecting only one option.

Whether to use a visible selection or a dropdown mainly depends on the length of the list and on taste. You can change the question type at any time.

The multiple choice question allows selecting several response options.

Important: While a simple selection results in one variable (which stores the selected option), a multiple choice question results in one variable for each item, which stores whether the item was checked or not.

The horizontal selection is another display form of the selection. In its appearance, it is similar to scale questions, but it offers other functions (e.g., open input fields for the options) and, unlike scale questions, does not support multiple items.

In the selection sequence, selection questions are displayed one after the other (they are created as partial questions within the question), which the respondent answers simply by clicking on the option (displayed in a box). There is no need to click “Next”.

Besides the selected option, this question also stores the respondent's response time.

Detailed manual: Selection Sequence

The extended selection can manage more than 99 options with custom codes and/or read options from the project's internal database (Using Data Fragments in PHP Code). This way, options can be read from external sources (e.g., patient data) or changed dynamically during the survey (e.g., adding new options).

Detailed manual: Extended Selection

On a scale (extremes only), only the extreme values can be labeled. In return, these scales offer somewhat more design freedom than scales with labeled interim values. If you do not want to label the interim values, use this scale. Advantages:

• The horizontal spacing of the input fields can be considerably narrower
• A visual break can be inserted into the scale
• The poles can be labeled within the respective rows

Technical note: The input fields are displayed (in contrast to “normal” scales and other question types) by means of CSS positioning. This makes the page display a little faster.

In the following scale, the verbal anchorage is placed within the row, and the 0 is visually separated.

In contrast to general scales, fully labeled scales also allow labeling the interim values. Optionally visible or invisible – in the latter case, the label is only displayed when the mouse is moved over one of the input fields.

If the display of your scale does not meet your expectations, feel free to experiment with the settings. Among other things, a visual anchorage, the orientation of the scale, or the labels can be changed.

For more information, see Scale.

In most respects, a slider is equivalent to a scale – but instead of checking a box, the respondent moves a slider handle on a scale.

Tip: If you would like to use a visual analogue scale (VAS), use the corresponding question template. Among other things, the slider handle is replaced by a cross, which is only displayed when the respondent clicks on the scale.

Further information on its use: Slider

Like scales, polarity profiles can have different numbers of levels.

In addition, SoSci Survey offers the option to vary the polarity of the items randomly.

The two-sided slider works like the normal slider – but you can place a label to the left and right of the slider, as in the polarity profile.

Tip: For the normal slider, you can set that the labels are displayed “at every item”, i.e., also to the left and right of the slider. In this case, an item text (e.g., a statement) can be displayed in addition to the labels. If the extremes of the slider are labeled the same every time (e.g., “does not apply”/“applies”), a normal slider often makes more sense than a two-sided slider.

As the name suggests, the text input allows the respondent to type text. The text input field can be single-line or multi-line. A label can be placed to the left of the input field, or the input field can span the full width.

Note: The input fields can be single-line or multi-line. The difference is important because different HTML form elements are used. If you use a multi-line input field but set its height as small as that of a single-line field, some browsers display a mini scrollbar on the right. Therefore: Do not specify a height for single-line input fields!

Tip: To get an input box spanning the full width (i.e., without a text in front of it), simply omit the text and specify only a height.

Tip: If the respondent shall use a certain format for the input (e.g., hh:mm for hours and minutes), you can specify the format by means of regular expressions in the corresponding item of the question. A corresponding hint is best placed in the explanation below the question text or in the label of the input field. A hint that appears dynamically or a “pre-label” within the input field is less promising¹⁾.

This question presents several input fields to the respondent (one below the other) – and whenever the respondent fills out one of the fields, a new one appears below. In the question, you specify the maximum number of input fields offered.

This question is useful if you expect several independent mentions from the respondent. The individual mentions are stored in different variables – in contrast to a (large) open input field, this saves a lot of work during the analysis.

Note: This question uses JavaScript.

The cloze text is a special form of the text input. Here, the input fields are placed directly within the text.

The balloon test is another special form of the text input. As a basis, however, you need an image in which speech bubbles are already drawn. With SoSci Survey, you can then display texts and input fields in this image.

HTML does not provide a form element for rankings. Since version 2.0, SoSci Survey offers a solution that uses JavaScript for this purpose. A series of cards is displayed, which can be arranged on ranks with the mouse.

Note: A small proportion of internet users has JavaScript disabled. These respondents see a different display with dropdowns.

Note: The ranking question has to be answered with the mouse; filling it out with the keyboard is not possible when JavaScript is enabled. For people with visual impairments, this can be a significant barrier. In the screen reader mode, SoSci Survey uses the alternative display with dropdown input fields.

Especially in psychology, the time between the presentation of a stimulus and the respondent's answer (response time, latency) is often used as an indicator of the accessibility of an attitude.

Detailed manual: Assignment Task

Response time generally depends heavily on further factors (e.g., complexity of the stimulus, general response time of the person, current concentration, etc.). In the online questionnaire, further distortions are added, e.g., loading times or the individual skill in clicking with the mouse. In a normal question, e.g., a scale, measuring the response time is therefore largely pointless, since only a marginal relevant share of variance remains.

SoSci Survey therefore offers a special assignment with response time measurement. The stimuli (terms, images, or graphics) are presented individually, and the rating is done by pressing a key. The terms are thus assigned to two or more categories (e.g., right/wrong, pleasant/unpleasant, rating 1 to 5).

Note: A small proportion of internet users has JavaScript disabled. These respondents see a different display; a response time cannot be determined then. The same applies to users with visual impairments who answer the questionnaire in the screen reader mode.

Some templates simply serve to quickly create frequently used questions (e.g., a 5-point Likert scale), other templates are special modifications of standard question types (e.g., the puzzle design for ranking questions). If a template is selected when creating a new question, a question with the underlying standard question type and the matching settings is created.

Note: The questions for implicit attitude measurement are only available if the add-on module “Implicit Methods” has been activated for the user account. If a survey project is shared with several user accounts (Shared Access to Projects), the module must be activated for the project owner.

In the IAT, words or (more frequently) images are assigned to several categories. An implicit attitude is inferred from the response time for the assignment and (indirectly) from the number of incorrect assignments.

SoSci Survey automatically analyzes the test and provides the test result (d-score) as well as a number of indicators in the dataset. The data for each individual assignment decision is available as JSON in the dataset.

Besides the standard IAT (Implicit Association Test), the Single-Category IAT (SC-IAT) and the Brief IAT (BIAT) are available as variants.

Affect Misattribution Procedure (AMP)

In the Affect Misattribution Procedure (AMP), test stimuli are presented to the respondent as primes, followed by neutral symbols (e.g., Chinese characters). The respondent shall rate the symbols as positive or negative and not be distracted by the previously presented images. The implicit attitude towards the test stimuli is inferred from the rating of the neutral symbols.

SoSci Survey directly provides the test score (proportion of positive ratings per category) in the dataset.

Internal variables are not displayed as a question. The only task of internal variables is to store values in the dataset, e.g., a random number that was drawn.

Internal variables are used together with the PHP function put().

To protect against programs that fill out online forms automatically (bots), CAPTCHAs have become established on the internet.

SoSci Survey facilitates including CAPTCHAs with the question type of the same name, CAPTCHA.

Sometimes an email address shall be stored, but without breaking the anonymity of the survey. This special question type serves this purpose. How it works is described in detail in the chapter Collect Contact Data Separately.

If you would like to send an email to an email address specified by the respondent, a question of the type E-Mail to personal Contact will help you. This question shows the respondent an input field for an email address and sends an email to this address when “Next” is clicked.

Manual for this question type: E-Mail to a Personal Contact

In combined questions (also: matrix questions), two questions with the same items can be displayed side by side. How this works is explained in the section Combined Questions

… are explained in the section Positioning Input Fields Freely

... Filter Questions

... A question in which I can program different texts