2. Type in a sample of data into the Text box. For example, for the date pattern you might type in 20201231.
3. If the pattern works correctly it will show that it has matched the data, the result that will display in Snap and the Values of each of the component parts.

If the test fails either the text does not fit the pattern or the pattern needs editing.

Result and Modify Case only alter the way the data is displayed in the analysis.

The post Testing a pattern appeared first on SnapSurveys.

The post Character selection dialog appeared first on SnapSurveys.

The post Repeat pattern components appeared first on SnapSurveys.

The post Match Pattern dialog appeared first on SnapSurveys.

The post Component properties dialog appeared first on SnapSurveys.

Viewing the patterns

1. Open the Patterns dialog from the View | Patterns menu on the Snap XMP Desktop menu.
2. To open the Pattern Properties dialog you can
   • Select a pattern in the list and click the View or Modify button. The patterns supplied with Snap XMP are read-only.
   • Click the New button to create a new pattern.

The post Patterns properties dialog appeared first on SnapSurveys.

Snap XMP provides a number of ways to randomise the order of your questions:

• You can define a group of consecutive questions that will be randomised. This is called a randomised block of variables.
• You can randomise several blocks of variables and this may contain further combinations of fixed blocks of variables or randomised blocks of variables
• Advanced options let you build complex randomisation with nested blocks.

In Snap XMP, you can also randomize the order of question choices in a closed question or the order of rows in a grid question.

Adding the randomised blocks

The randomised question blocks are created and edited in the Randomisation dialog, available in the Questionnaire Design window.

1. Open the questionnaire in the Questionnaire Design window.
2. Click the Randomise icon on the Snap XMP Desktop toolbar. This opens the Randomisation dialog. In this dialog, you can create blocks of variables that randomise the question order in the questionnaire.

3. Click the Add button to start defining your randomisation blocks. This opens the Block Details dialog.
4. There are 2 block types available:
   • A randomised block of variables: This lets you randomise a block of consecutive questions or variables.
   • Randomise several blocks of variables: This lets you build randomisation blocks containing more than one block of variables.

A randomised block of variables

This option lets you define one block containing consecutive questions or variables that will be shown in a random order.

1. Select the Block Type: A randomised block of variables. This enables the First variable and Last variable drop-down lists.
2. In the First variable drop-down, select the variable that starts the block.
3. In the Last variable drop-down, you will see all the variables that come after the First variable. Select the variable that finishes the block.

4. Click OK. The block appears in the list of Questionnaire blocks.

Randomise several blocks of variables

This option lets you define multiple blocks containing consecutive questions or variables that are shown in a random order.

1. Select the Block Type: Randomise several blocks of variables. This defines the main block that will contain other child or nested blocks.
2. Click OK. The block appears in the list of Questionnaire blocks. Now you can add other blocks of variables.

3. To add a block, select the Randomise several blocks item in the list of Questionnaire blocks then click the Add button. This opens the Block Details dialog.

4. There are 2 block types available. (You may see 4 block types when the advanced option is on.)
   • A fixed block of variables: This lets you define a block of variables that will be shown in the order set in the questionnaire.
   • A randomised block of variables: This lets you randomise a block of consecutive questions or variables.
5. Select the required block Type. The First variable and Last variable drop-down lists become enabled.
6. In the First variable drop-down, select the variable that starts the block.
7. In the Last variable drop-down, you will see all the variables that come after the First variable. Select the variable that finishes the block.
8. Click OK. The block appears in the list of Questionnaire blocks.

6. Repeat this process until you have entered all the required blocks.

Fixed Variables

When creating a group of randomised variables, you may wish to fix the position of some of the variables, such as headings or instructions that need to be shown at the top.

Variables within the randomised block can be set as fixed using the Fixed Variable setting.

1. In the Randomisation dialog, select the Block Type: A randomised block of variables. This enables the First variable and Last variable drop-down lists.
2. In the First variable drop-down, select the variable that starts the block.
3. In the Last variable drop-down, you will see all the variables that come after the First variable. Select the variable that finishes the block. This enables the Select button.

4. Click the Select button. This shows a list of all the variables in the randomised block.

5. Select the variables that you want to stay in the fixed position.
6. Click OK to save the changes.
7. The Details section shows the block information, including the fixed variables.

Removing blocks

1. In the Randomisation dialog, select the block that you wish to remove. If you select a block with child blocks, then this action removes all the child blocks as well as the block selected.
2. Click the Remove button.
3. You are asked to confirm the deletion. Click OK to remove the blocks, otherwise click Cancel.

Editing blocks

1. In the Randomisation dialog, select the block that you wish to edit.
2. Click the Edit button.
3. Edit the Block type, First variable and Last variable.
4. Click OK to save the changes, otherwise click Cancel.

Errors

When there are errors in the questionnaire blocks, you will see a red cross next to the blocks that have an error. If you save the blocks with an error, then the Randomise icon is shown with a red cross .

For example, the randomisation blocks cannot contain overlapping groups of variables. If this happens, a red cross is displayed next to the questions that overlap. When you select the block with an error, a description of the error is shown in the Details section.

Conflicts with routing and text substitution

If your survey uses routing or text substitution, you need to consider the effects of randomising the question order, as routing and text substitution use the answers to previous questions.

For example, if you create a randomisation block containing three questions: Q1, Q2 and Q3 where there is also routing set on the questionnaire that Q2 and Q3 are shown only when the first answer is selected in Q1 (Q1=1).

Here are all the possible combinations:

• Q1, Q2, Q3: randomisation OK (because it is the original sequence)
• Q1, Q3, Q2: randomisation OK
• Q2, Q1, Q3: randomisation cancelled because Q2 needs to be after Q1
• Q2, Q3, Q1: randomisation cancelled because Q2 and Q3 need to be after Q1
• Q3, Q1, Q2: randomisation cancelled because Q3 needs to be after Q1
• Q3, Q2, Q1: randomisation cancelled because Q3 and Q2 need to be after Q1

In this example, you can see that routing is prioritised and the randomisation is cancelled when Q1 is not first in the randomisation sequence. Removing Q1 from the randomisation block, would allow both randomisation of Q2 and Q3, and routing to work.

Examples

In these examples, the question numbers have been included to show the randomisation. Usually, you would not show the question number so the respondent is unaware that the questions are randomised.

Creating a randomised block of variables

This is a group of individual, consecutive questions that are shown in a random order when the respondent is completing the questionnaire.

1. Open the questionnaire in the Questionnaire Design window. This example has 5 questions about an instructor’s performance.

2. Click the Randomise icon on the Snap XMP Desktop toolbar. This opens the Randomisation dialog.
3. Click the Add button to start defining the randomised block of questions. This opens the Block Details dialog.
4. Select the Block Type: A randomised block of variables. This enables the First variable and Last variable drop-down lists.
5. In the First variable drop-down, select the variable that starts the block.
6. In the Last variable drop-down, you will see all the variables that come after the First variable. Select the variable that finishes the block.

7. Click OK. The block appears in the list of Questionnaire blocks.

When the questionnaire is completed by a respondent they will see the questions in a random order. From the question numbers you can see that the questions are in a different order than the questionnaire design.

Randomising several blocks of variables

You can randomise blocks of questions. The two options are

1. Randomising the blocks, but keeping the questions within those blocks in the original order
2. Randomising the blocks, and also randomising the questions within those blocks

Randomising several blocks that contain questions in the original order

1. Open the questionnaire in the Questionnaire Design window.
2. Click the Randomise icon on the Snap XMP Desktop toolbar. This opens the Randomisation dialog.
3. Click the Add button to start defining your randomisation blocks. This opens the Block Details dialog.
4. Select the Block Type: Randomise several blocks of variables. This defines the block that will contain the other blocks.

5. Click OK. The block appears in the list of Questionnaire blocks. Now you can add other blocks of variables.
6. To add a block, select the Randomise several blocks item in the list of Questionnaire blocks then click the Add button. This opens the Block Details dialog.

7. Select the block type: A fixed block of variables. The First variable and Last variable drop-down lists become enabled.
8. In the First variable drop-down, select the variable that starts the block.
9. In the Last variable drop-down, you will see all the variables that come after the First variable. Select the variable that finishes the block.
10. Click OK. The block appears in the list of Questionnaire blocks.

11. Repeat this process until you have entered all the required blocks.

Randomising several blocks that contain questions in a random order

1. Open the questionnaire in the Questionnaire Design window.
2. Click the Randomise icon on the Snap XMP Desktop toolbar. This opens the Randomisation dialog.
3. Click the Add button to start defining your randomisation blocks. This opens the Block Details dialog.
4. Select the Block Type: Randomise several blocks of variables. This defines the block that will contain the other blocks.

5. Click OK. The block appears in the list of Questionnaire blocks. Now you can add other blocks of variables.
6. To add a block, select the Randomise several blocks item in the list of Questionnaire blocks then click the Add button. This opens the Block Details dialog.
7. Select the block type: A randomised block of variables. The First variable and Last variable drop-down lists become enabled.
8. In the First variable drop-down, select the variable that starts the block.
9. In the Last variable drop-down, you will see all the variables that come after the First variable. Select the variable that finishes the block.
10. Click OK. The block appears in the list of Questionnaire blocks.

11. Repeat this process until you have entered all the required blocks.

The post Randomising the order of your questions appeared first on SnapSurveys.

The validation available in Snap XMP includes:

• making a question mandatory, where the participant must answer it
• setting a response type
• setting the maximum length for a free format text answer
• adding a pattern, such as a date or email address, that the answer needs to match
• setting a minimum number of answers for multiple response questions
• setting a maximum number of answers for multiple response questions
• setting an initial value for a question
• making a question read only

Setting up validation in Snap XMP Desktop

1. Each variable has a set of individual properties. The properties depend on the type of question. You can examine and change the properties when you are in Design mode   in the Questionnaire window.
2. Select one or more questions and click Variable Properties   on the Questionnaire toolbar. This opens the Variable Properties dialog.

3. Edit the property values by entering or selecting a value for the required property. If you have selected several questions of the same type, you can change any properties they share.

More information about the variable properties can be found at Variable properties.

There is also a video available that includes a section on response validation in Snap XMP Desktop at Snap XMP – Survey Logic in Snap XMP Desktop from Snap Surveys on Vimeo.

Setting up validation in Snap XMP Online

In Snap XMP Online open the survey in the Build tab to edit the questionnaire. The Validation & Masking menu is in the Build side menu.

The validation properties shown depend on the question that is selected.

More information about the validation options in Snap XMP Online can be found at Validation.

Some properties such as the response type or pattern are set in the question toolbar.

More information about the question toolbar can be found at Inserting questions.

There is also a video available that includes a section on response validation in Snap XMP Online at Snap XMP – Survey Logic in Snap XMP Online from Snap Surveys on Vimeo.

Making a question mandatory

When you want to make sure that a respondent answers a question, set the Must answer to Yes in Snap XMP Desktop or select the Mandatory box in Snap XMP Online. The respondent needs to answer the question before they can progress through the questionnaire. The mandatory option is useful when questions further on in the questionnaire rely on an answer to an earlier question.

Setting the response type

When you create a question, it has the default response type for the pre-defined question style. You can change the response type according to the question style and the requirements of the questionnaire.

The response types for Single Choice, Multi-choice and Grid questions are

• Single for a Single choice, Multi choice or Grid question when the respondent only selects one answer. For example, Yes/No or Age group questions
• Multiple for a Single Choice, Multi choice or Grid question when the respondent can select one or more answers

The response types for Open Ended, Open Series questions are

• Quantity which allows only numbers.
• Literal which allows any free format text, such as names and comments.
• Date which allows a range of date formats.
• Time which allows a range of time formats.

More information about setting the response type can be found at Setting the response type.

Initial value

A question can have an initial value. This value initially shows or selects the initial answer in the question, when a respondent is completing the questionnaire. You can set the initial value to select a code, or show a literal constant or expression using values from the preceding questions.

More information about setting the initial value can be found at Setting an initial value.

Read only

Setting a question as Read only prevents the respondent from answering or overwriting that question during an interview.

Use the read only property where:

• the question contains seeded values from a participant database that you do not want the participant to change
• the question contains an initial value that is based on answers to previous questions in the survey that you do not want the participant to change

More information about read only questions can be found at Creating a read only question.

Setting the minimum and maximum number of answers

You can set the minimum and/or maximum number of answers required for multiple response questions. When the minimum number of answers is set to a value greater than 0, the question automatically becomes mandatory. The maximum number of answers required cannot be greater than the code count.

Rating Check

Often you would like to ask the respondents of your survey to rank a number of options in a given order. You can use a grid question to achieve this by setting the Rating Check property. This property ensures respondents answer the question with only one ranking in each aspect or row.

More information about rating check can be found at Using Rating Check to ensure ranking

Setting an exclusive response

When you create a multiple choice question containing a list of items, it is unlikely that the list will contain all available options. You may provide one or more addition codes such as Other, Don’t know or None of the above. When these are selected, none of the other responses are applicable and should not be selected.

In the Exclusive property, enter the list of the codes that are not available in combination with the other responses. When any of these codes are selected, any other responses to the question are cleared.

Maximum length or number of characters

You can set the maximum length or number of characters entered in a free format text question, such as an Open Ended question. The answer cannot exceed the maximum length.

Valid ranges

You can add an expression in the Valid field of a variable so that the respondent can only enter data which matches the expression. For complex expressions it is better to use patterns.

The expression must be a list or range of possible values.

These can be

• limited range e.g., 1 TO 5 or 1~5
• unlimited range using relational operators such as > (greater than) e.g., >10 would include all values greater than 10
• ranges defined by variables in the questionnaire e.g., <Q9 would include all values less than the response to Q9
• a list e.g., 1,2,3

More information about setting valid ranges can be found at Setting valid ranges on a question

Setting the response format using Patterns

Patterns define the format of the data that a respondent should enter when completing the questionnaire. Open ended questions use them to check that the entered data is in the correct format.

In data entry, patterns help to

• Ensure that respondents or data entry personnel enter the correct format for information such as postcodes, zip codes or National Insurance numbers. Pattern matching checks whether the answer is a possible match in the pattern.
• Convert the data into a standard form, such as adding currency symbols in quantity data.

Snap XMP contains a library of pre-defined patterns. You can also create your own using the Patterns dialog in Snap XMP Desktop.

More information about patterns can be found at Introduction to patterns.

Scanning validation

Patterns can also help validate responses from scanned questionnaires.

There are two types of scanning validation:

• Spelling validation checks a response against a stored dictionary of words and tries to match responses to words in the dictionary.
• Pattern validation attempts to match a response against the expected pattern of characters. This is often used to validate responses such as dates, telephone number and zip or post codes.

More information about patterns can be found at Using scanning validation patterns

The post Setting up response validation appeared first on SnapSurveys.

• asking how many hours a week respondents spend doing different things
• asking how much money respondents spent on different activities during a visit
• asking respondents to share a fixed amount or percentage between a set of options

The amount allocated to each option gives the relative importance of each. This gives good numeric data for analysis, but is quite difficult for respondents.

This worksheet shows you how to calculate a running total as the respondent completes the answers. It explains how to create a set of Open Series First and Open Series Next questions for quantity responses and put a running total into the final response. It also shows you how to set a valid range on the total, so that respondents cannot move to the next page if the total has overrun.

Please note: The Open Series First and Open Series Next question styles are new in Snap XMP Desktop. If you are using a survey that has been upgraded from Snap 11 Professional then the equivalent question style is Open Grid First and Open Grid Next.

Step 1: Create your set of questions

The first step is to create the group of questions that need to be totalled. These must be quantity questions.

1. In Snap XMP Desktop, open the survey in the Questionnaire window.
2. Click New Question to create a new question.
3. Select Open Series First as the question style
4. Select Response in the next dropdown then select Quantity as the response type.

5. Enter the question text, for example “How did you spend your day at the activity centre? Please allocate hours to the activities up to a maximum of 8 hours.”
6. Enter the first activity in the grid label.
7. Press Tab to create the next question in the list. By default, the next question added is an Open Series Next question with a Quantity response. Enter another activity in the grid label. Repeat until you have entered all the activities.
8. Add an extra grid question to display the running total.

9. Click Save to save the questionnaire.

Step 2: Display the total in the calculated field and set the valid range

1. Select the open series question that will display the total.
2. Click Variable Properties to open the Variable Properties dialog.
3. In this example, the total hours have a valid range of 0 to 8. Scroll to the Valid property and enter 0 to 8.

4. Scroll to the Initial Value property and enter the total: Q1a+Q1b+Q1c+Q1d+Q1e. When the open series entries change the value of the total field is updated when the respondent moves to the next entry. The questions used in the total calculation must always be asked earlier in the questionnaire.

5. By default, the Read Only property is set to No. This allows the respondent to edit the total field. In this example, the total should not be edited so set the Read Only property to Yes. This renames the Initial Value property to the Calculation property.

6. Click OK.
7. Click Save to save the questionnaire.
8. When the questionnaire is complete. Publish the survey in Snap XMP Desktop or Snap XMP Online.

If there is a topic you would like a worksheet on, email to snapideas@snapsurveys.com

The post Creating running totals appeared first on SnapSurveys.

You can find these questions in the Reference Survey Pak (Personal demographics) supplied with Snap XMP.

Step 1: Create a derived variable

To set up an interlocking quota, you first need to set up a derived variable in your survey.

1. Open the survey in Snap XMP Desktop.
2. Click Variables on the Snap XMP Desktop toolbar. This opens the Variables window.
3. Click New variable to create a new variable.
4. Name the variable AgeGenderGroup, and label it Age and gender groups.
5. Set the Type to Derived, the Response to Single.

6. In the code list add the code labels and values:

7. Click Save to save the variable.

Step 2: Set up the quota for the variable

1. Select View | Quotas and Screen Outs. This opens the Quota and Screen Outs window.
2. Click New Quota or Screen Out then select New Quota. This opens the Quota dialog.
3. In Name, enter AgeGenderQuota.
4. In Variable, select the variable AgeGenderGroup. When you select a variable, the Details table lists all the answer codes. The target for each category is available as a number or percentage. When you enter a target number, this automatically updates the percentage and overall target. Once interviewing has started Actual and % show the latest response counts.

5. When the quota is exceeded, the participant is directed to a web page or shown a custom message. In multilingual surveys, select the language then enter the message for the selected language. Further details are available at: Notifying the participant when a quota or screen out is exceeded
6. Click OK to save the quota. The Quota and Screen Outs window shows the new quota.

Step 3: Enable the quota

You can enable quotas individually or at the survey level, which is the default. You can change this in the Quota Control dialog.

In the Quota dialog, select Enabled to activate the selected quota. Clear Enabled to disable the selected quota.

If there is a topic you would like a worksheet on, email to snapideas@snapsurveys.com

The post Interlocking quotas appeared first on SnapSurveys.

Routing allows you to select which is the next question in the questionnaire based on previous answers. There are three types of routing:

• Conditionally Ask Question. The respondent sees the question when certain conditions are met. For example, if the respondent answers a previous questions giving a particular answer.
• Goto On Answer. The respondent moves through the questionnaire depending which answer they select. It’s normally used to skip sections of a questionnaire according to the response to a single-response question.
• Goto After Question. The respondent moves through the questionnaire after a question has been answered. It’s normally used to jump irrelevant sections of a questionnaire.

This worksheet describes how to set up Conditionally Ask routing so that only the appropriate elements in a grid question are displayed based on responses from previous questions and also displaying question based on responses in a grid.

Display a grid based on the response to a question

In this example, a questionnaire starts with a multi-choice question asking which means of transport the respondent uses to go to work. Several sets of grid questions display depending upon the answer to the first question. For example, the grid asking about travelling to work by car only shows if the respondent answers Car (code 1). Routing only shows the grid when it is relevant.

1. Select the first row of the grid, (Q3.a) I choose to travel by car because….
2. Click Routing Rules to open the Routing Rules dialog for that question.

3. Click Add to create a routing rule.

4. Set the type to be Conditionally Ask Question and click OK. This opens the Routing Rules dialog.
5. In the If field, enter Q1=1. This will display the first grid row if the respondent has selected code 1 (Car) in response to question 1.

6. Click OK. A small arrow is added to the start of the question to show that there is routing applied to it. You can double-click this arrow to open the Routing Rules details.

7. This Routing Rule need to be applied to all the rows in the grid. This can be done using in the Variable Properties. Select the grid question containing routing and click Variable Properties to display the Variable Properties window. The routing condition is shown in the Not Asked property.
8. Copy the Not Asked property then select all the grid rows and any associated titles, sub titles or instructions. Click Variable Properties and paste the Not Asked property.

Display a question based on the responses in a grid

A further question (Q4) asks respondents why they do not find public transport convenient. It only makes sense to ask this question if the respondent has indicated that they do find public transport inconvenient.

This means that you need to set routing on this section so it is only asked if they have selected the Strongly agree or Agree responses to the question …public transport is not convenient (Q3.b).

1. Select question 4, and click Routing Rules  to open the Routing Rules dialog.
2. Click Add to create a new routing rule with the type set to Conditionally Ask Question and click OK.
3. In the If field, enter Q3.b=(1,2). This will display question 4 if the respondent selects Strongly agree (code 1) or Agree (code 2) in response to question 3.b.

4. Click Variable Properties to display the Variable Properties window and copy the text in the Not Asked field (unless( q3.b=(1,2))).
5. Select all the grid rows and any associated titles, sub titles or instructions. Click Variable Properties and paste the Not Asked property. This page should now only appear if the respondent finds public transport inconvenient.

Display specific rows in a grid based on the response to a question

This example uses a multi-choice question for respondents to select the facilities they have used.

The survey then offers a rating scale for the facilities. Routing is applied that shows the row for each facility that the respondent has used.

1. Select the first line of the first grid section, (Q2.a) Rowing machine
2. Click Routing Rules  to open the Routing Rules dialog for that question.
3. Click Add to create a new routing rule with the type set to Conditionally Ask Question and click OK.
4. In the If field, enter Q1=1. This displays the first question of the grid if the respondent has selected code 1 (Rowing machine) in response to question 1.

5. Click OK. A small arrow is added to the start of the question to show that there is routing applied to it. You can double-click this arrow to open the Routing Rules dialog.
6. Repeat this for each question in the grid, using the required code.
7. Click Save to save the questionnaire.
8. Test the routing works by publishing the questionnaire and launching the preview.
9. Select the facilities in the multi-choice question.

10. Click Next and confirm that the correct elements of the grid appear.

Further information on routing is available at Adding routing and Routing rule expressions.

If there is a topic you would like a tutorial on, email to snapideas@snapsurveys.com

The post Using routing in grid questions appeared first on SnapSurveys.

• show a default page with a customised message
• send them to a page of your choice

These can be set for the

• overall target quota in the Quota Control dialog
• variable based quotas and screen outs in the Quota dialog and the Screen out dialog

Setting a message

1. Select View | Quotas and Screen Outs. This opens the Quota and Screen Outs window. (Prior to build 12.10 this was on the Tailor menu.)
2. Open the quota or screen out as follows:
   • If you wish to set the message of a variable based quota, double-click the quota in the list to open it in the Quota dialog.
   • If you wish to set the message of a screen out, double-click the screen out in the list to open it in the Screen Out dialog.
   • If you wish to set the message of the overall target, click Tailoring to open the Quota Control dialog.
3. Select Message.

4. If using a multi-lingual survey, select the language.
5. Enter the customised message. This message displays in the default Snap Online web page when the quota exceeds the target or the screen out triggers.

6. Click OK to save the changes.

Setting the web page URL

1. Select View | Quotas and Screen Outs. This opens the Quota and Screen Outs window. (Prior to build 12.10 this was on the Tailor menu.)
2. Open the quota or screen out as follows:
   • If you wish to set the message of a variable based quota, double-click the quota in the list to open it in the Quota dialog.
   • If you wish to set the message of a screen out, double-click the screen out in the list to open it in the Screen Out dialog.
   • If you wish to set the message of the overall target, click Tailoring to open the Quota Control dialog.
3. Select URL.
4. Enter the URL of the web page. This web page is displayed when the quota or screen out is exceeded.

7. Click OK to save the changes.

When there are multiple quotas and an overall target quota, the participant will be shown the message of the first quota that is exceeded.

The post Notifying the participant when a quota or screen out is exceeded appeared first on SnapSurveys.

First, assess the characteristics that are required to give the desired sample for the survey. Then create a series of questions that select these participants. Define these as screening variables by creating a Screen Out for each question or variable.

Creating a screen out based on a variable

There are three steps in the process of creating a screen out:

1. Identify or create the variable that is the basis for the screen out. This can be a question on the questionnaire or a derived variable if you need to set screen outs that are a combination of questions, e.g. Males under 25.
2. Create the screen out.
3. Enable the screen out.

Identify the screening variable

First, decide which question or combination of questions should be used as the basis of the screen out.

This can be:

• A single response question in the questionnaire; the screen out will be based directly on the question variable
• A derived single response variable for screen outs based on a number of questions. For example, one based on age and gender.

Place the screening question near the start of the questionnaire so the participant can be excluded as soon as possible. The participant will be notified early on and will not answer questions unnecessarily.

Create the screen out

1. Select View | Quotas and Screen Outs. This opens the Quota and Screen Outs window. (Prior to build 12.10 this was on the Tailor menu.)
2. Click New Quota or Screen Out then select New Screen Out. This opens the Screen Out dialog.
3. In Name, enter a name to identify the screen out.
4. In Variable, select the variable that is the basis of the screen out. When you select the variable, all the answer codes display in the Details table. By default, each category is included and shown as In.
5. Overall Target shows the overall target number of responses that are to be submitted for the survey. The quota control overall target is used as the read-only default. Selecting the Enforce target option, enables the Overall target and Required fields. Enabling the Enforce target option is only recommended when you are using screen outs only. When you have screen outs and quotas together they may interact. For example, if one of the quotas has codes that are over quota then the screen out can stop the survey on the overall target before the quotas are complete.
6. Required is a read-only field that shows the number of responses left to meet the overall quota target.
7. Select Out for the categories that screen out the participant. When the participant selects an answer in a category that is set to Out, they will be excluded from the survey. In the example shown, the survey excludes customers who are under 18. Customers who respond that they are under 18 will be excluded from the survey.

6. Once interviewing has started the Count and % show the latest counts of screened-out participants for each category.
7. When the screen out triggers, this directs the participant to a web page or shows a custom message. In multilingual surveys, select the language then enter the message for the selected language.

9. Click OK to save the screen out. The Quota and Screen Outs window lists the new screen out.

Enabling the screen out

You can enable Quotas and Screen Outs individually or at the survey level.

In the Quota Control dialog, select Enabled to allow all quotas and screen outs in the survey to run. Clear Enabled to disable all quotas and screen outs in the survey.

In the Screen Out dialog, select Enabled to activate the selected screen out. Clear Enabled to disable the selected screen out.

Screen outs vs routing

A screen out looks similar to routing a question to the end of the questionnaire when a specific answer is selected. However, a screen out differs from routing:

When the survey excludes a participant from the survey with a screening question, the participant exits the questionnaire without submitting their response. The response is not saved and does not count towards any quota target, including any overall target.

With routing the participant skips to the end of the questionnaire where they can submit the response. The saved response counts towards any quota target, including any overall target.

The post Setting up Screen Outs appeared first on SnapSurveys.

1. Identify or create the variable to be used as a basis for the quota. This can be a question on the questionnaire or a derived variable if you need to set quotas that are a combination of questions, e.g. “Males under 25”.
2. Decide the target quota for each possible variable answer. This can be set as a number or a percentage.
3. Create and enable the quota.

Identify the quota variable

First, decide which question or combination of questions should be used as the basis of the quota.

This can be:

• A single response question in the questionnaire; the quota will be based directly on the question variable
• A derived single response variable for quotas based on a number of questions. For example, one based on age and gender.

Place the quota question near the start of the questionnaire so the quota can be checked as soon as possible. The participant will be notified early on and will not answer questions unnecessarily.

Create the quota

1. Select View | Quotas and Screen Outs. This opens the Quota and Screen Outs window.
2. Click New Quota or Screen Out then select New Quota. This opens the Quota dialog.
3. In Name, enter a name to identify the quota.
4. In Variable, select the variable that is the basis of the quota.
5. When the variable is selected, all the answer codes are shown as categories in the Details table. The target for each category can be entered as a number or percentage. Leaving the target blanks allows unlimited responses (until any overall target is exceeded). When saved this is shown as na (not applicable).

6. The Overall field shows the overall target number of responses that are to be submitted for the survey. When you change the quota target for a category the overall target is also updated. This affects any other quotas and screen outs for this survey. The overall target can be updated by entering the target number, if required. Required is a read-only field that shows the number of responses left to meet the overall quota target.
7. Once interviewing has started the Actual and % show the latest response counts and percentages.
8. When the quota is exceeded you can set whether the participant is directed to a web page or shown a custom message. In multilingual surveys, select the language then enter the message for the selected language.

9. Click OK to save the quota. The new quota is shown in the Quota and Screen Outs window.

You can set up any number of independent quotas in this way. The first quota to be exceeded for any particular interview will stop the interview.

Enabling quotas

Quotas can be enabled individually or at the survey level.

In the Quota Control dialog, select Enabled to allow all quotas and screen outs in the survey to run. Clear Enabled to disable all quotas and screen outs in the survey.

In the Quota dialog, select Enabled to activate the selected quota. Clear Enabled to disable the selected quota.

The post Setting up Quotas appeared first on SnapSurveys.

This quota is not linked to any question or variable but can be combined with additional quotas and screen outs that set out further criteria for each response.

Quota controls

You can set the overall target quota in the Quota Control dialog.

1. Select View | Quotas and Screen Outs. This opens the Quota and Screen Outs window. (Prior to build 12.10 this was on the Tailor menu.)
2. Click Tailoring . This opens the Quota Control dialog.
3. In Target, enter the target number of responses required. When responses are received the Required number is updated to reflect the number of responses left to fill the quota.

4. Set the URL or message that is shown to the participant when the overall target number of responses is exceeded.

Changing the overall target

In a quota based on a variable, the targets counts and percentages must add up to the overall target and 100%. When the overall target is changed the target counts and percentages in the quotas are automatically updated to reflect the change.

An example of how the quota is changed is shown.

1. An existing quota has target allocations of 25 responses or 25% of responses for each category.
2. The overall target is changed in the Quota Controls from 100 to 302.
3. When the quota is opened again the Target and % have been automatically updated to the best fit for the new overall target.

The overall target can also be set when creating or editing a quota or screen out. Regardless of where the overall target is changed, the new target is used for all the quotas and screen outs in the survey.

Changing the quota targets

Usually when you create or edit a target, you know the targets for each category rather than the overall target. When each category target is changed in the Quota dialog the overall target is updated in line

1. Select Tailor | Quotas and Screen Outs. This opens the Quota and Screen Outs window.
2. Double-click a quota. This opens the Quota dialog.
3. Change the target for one of the categories and the overall target updates to the new target total.

4. Click OK to save the quota. A message is displayed warning that the change to the overall total affects all quotas and screen outs.

If you change the overall total, the field changes color as a warning that the category targets and % need recalculating. When the target totals match the overall target the color is cleared from the field.

The post Setting an overall target appeared first on SnapSurveys.

Quotas can be set up that:

• limit the total number of responses submitted
• allocate the total number or percentage of responses based on the answer to a question
• allocate the total number or percentage of responses based on a derived variable

A screen out allows you to select which participants can progress through the questionnaire based on their answers to the screen out questions. This does not limit the total responses submitted.

You can set up multiple quotas and screen outs for each survey. However, the overall target number of responses applies to all quotas and screen outs.

Survey design considerations

There are a number of design considerations to note when creating the questionnaire.

• The quota and screen outs are checked when the participant moves to the next page or submits the response. A page break should be set after each quota or screen out question.
• The quota and screen out questions are mandatory in order to calculate the number of responses.
• The quota and screen out variables must return a single response.
• Make sure you do not use routing that skips any quota or screen out questions.
• When a quota or screen out is based on a derived variable then all source questions are also mandatory and should not be skipped using routing.
• Place the quota or screen out questions near the start of the questionnaire so they can be checked as soon as possible.
• Mobile responses, collected using Interviewer, are not included in the screen outs or quota limits.

Quotas and Screen Outs window

The Quotas and Screen Outs window shows a list of all the quotas and screen outs that are based on a variable.

On the Snap XMP Desktop menu, select the Tailor menu then select Quotas and Screen Outs.

This opens the Quotas and Screen Outs window where you can:

• create quotas and screen outs
• clone quotas and screen outs
• delete quotas and screen outs
• open the Quota Control dialog to set the overall quota targets and settings
• view the quota or screen out properties
• view the sources and dependencies

Quota Control dialog

The Quota Control dialog contains the overall quota targets and settings.

In the Quotas and Screen Outs toolbar, click the Tailoring icon.

This opens the Quota Control dialog where you can:

• enable or disable all quotas and screen outs for the survey
• set overall target numbers of responses
• set the message or URL that is shown when the overall quota is exceeded

The post Introduction to Quotas and Screen Outs appeared first on SnapSurveys.

Static masking is used to hide a response option in all conditions of the survey. Meaning, once applied to a question the response option will always be hidden. This might be useful for grid questions where one option is not applicable for a specific row within the grid. When used, the response option that is desired to be hidden is referenced and prefaced with the word “not” (i.e. “not 5”).

Dynamic masking is used to conditionally show response options based on a response to a previous question. This type of masking can be used to customize a survey’s response options based on how a respondent is answering questions.

Below are two examples of dynamic masking. Static masking will not be covered directly in this work sheet.

Example One: Masking follow-up questions with the response options

1. First, there will need to be two questions with identical response options. The best way to do this is by cloning the first question to make the second. To clone a question, select it in the Questionnaire Design window .

2. Click Clone to clone the selected question which will appear just below the selected question. The text of the cloned question should be edited appropriately and the response type can be changed to single:

3. The masking feature should be applied to the cloned question; in this example that’s Q2. To apply masking, right click the question and select Variable Properties. This opens the Variable Properties window where you can set the Mask property.
4. To Mask the cloned question so that only the options selected in a previous question are shown, the following syntax should be applied: “Question Name” (i.e. Q1)

5. If the opposite masking is desired (show response options that were not selected), the syntax should be as follows: “not Question Name” (i.e. not Q1)
6. Once the masking is applied, click Save to save the questionnaire.
7. To preview the masking feature, publish the survey in preview mode .

Example Two: Conditionally masking using derived variables.

In this example respondents will first pick the location where they work. Then, based on the location they picked, the response options of a latter question will only show the options available at that location. This would be useful in situations where response options for a question need to be grouped together and filtered out based on previous responses.

1. Create a question that will represent each group of masked options. In this example there will be three office locations:

2. Now create the question that will contain masking. This question should contain all of the possible response options. For example:

3. Next, the derived variable that determines the logic of the applied masking should be created. Click Variables and select the question being masked from the variables list, Q2.
4. Click Clone to create a new variable based on the selected question, Q2.
5. Edit the new variable so that the Type is set to Derived and Response is set to Multiple.
6. Set the code values for each response option in the derived variable. Each code value response references the masking question, for example Q1=(1,2,3)
7. The following shows the values for the departments at each office location:
   • London has Customer Services and Training
   • New York has Accounting, Customer Services, Human Resources, IT Support, Research, Sales and Training
   • Paris has Accounting, Customer Services, Data Center, Development, Human Resources, IT Support, Research, Sales and Training.

8. Click Save to save the derived variable.
9. In the Questionnaire window, select the masked question, right click and select Variable Properties .
10. In the Mask property enter the name of the derived variable, V1, as the masking value:

12. Click OK to save the masking value.
13. Once the masking is applied, click Save to save the survey.
14. To preview the masking feature in action, publish the survey in preview mode .

Alternative masking derived variable definition

There is an alternative way to define the derived variable used for masking. When the Toggle Definition is selected the definition fields are shown, including the Initial Value field. The Initial Value field can be used to enter a calculation that provides the basis for the values in the code list. This simplifies the code values when using a long calculation. The diagram shows the equivalent derived variables; one with an initial value and one without. Both these derived variables will behave in the same way.

If there is a topic you would like a tutorial on, email to snapideas@snapsurveys.com

The post Masking code boxes appeared first on SnapSurveys.

You can use the read only property for:

• seeding the question with values from a participant database that you do not want the participant to change
• the question contains an initial value that is based on answers to previous questions in the survey that you do not want the participant to change

You can set a question as read only in the Variable Details or the Variable Properties windows. Note that the Read Only property is not available for Paradata or Note variables.

In the Variable Details window, the select the Read only field to set the variable as read only. Clear the Read only field to allow the respondent to edit the variable. The Read Only field is available when you click Toggle Definitions.

In the Variable Properties window, the Read only property is set to Yes to set the variable as read only or No to allow the variable to be edited.

The post Creating a read only question appeared first on SnapSurveys.

The initial value is set in the Variable Details window or Variable Properties window.

In the Variable Details window, the Initial Value field is shown when you click Toggle Definitions .

The Initial Value property is also available in the Variable Properties dialog opened from the Questionnaire window.

Note that the Initial Value is not available for Paradata variables or Note variables.

Using the Initial Value property

In this example, the questionnaire is set up to ask about the number of people in the participant’s party. The values that the participant enters can be used to set the initial value for a question about the number of people in the party who ordered food.

In this questionnaire, Q2 is an Open Series question asking for the number of adults and children in a group of people at a restaurant and Q3 is an Open Ended question asking how many people in the party ordered food. Usually everyone orders food so the default for Q3 is the total number of people in Q2.

The questions are created using the following step by step instructions:

1. Open the survey in the Questionnaire window with Design mode set.
2. Click New Question   on the Questionnaire toolbar to insert a question.
3. Select the Open Series style from the Style Selection Box on the toolbar.
4. In the area marked “Click here for text”, enter “How many people were in your party today?”
5. Click Variable Properties on the Questionnaire toolbar. This opens the Variable Properties dialog.
6. In the Variable Properties change the
   • Response: Quantity
   • Label: Adults
7. Insert another entry with the properties
   • Response: Quantity
   • Label: Children.

8. Insert another Open Ended question to be used to show the number of people who ordered food. Again set the Response to Quantity.

9. Click Variable Properties on the Questionnaire toolbar. This opens the Variable Properties dialog.
10. Type Q2a+Q2b in the Initial Value to show the total number of people in the party. When an expression is used this automatically sets the Read Only property to Yes. Setting the Read Only property to Yes prevents a respondent overwriting or answering the question.

11. This question needs to be on the following page so insert a page break above this question.

12. When the questionnaire is running the participant enters the numbers in each question.

13. On the next page of the questionnaire the total number of people is used as the default.

The post Setting an initial value appeared first on SnapSurveys.

If conditions for single and multiple response questions

The conditions used for routing must return true or false.

These conditions are mainly for use in single or multiple response questions.

If conditions for any question

These conditions are for use with any type of question.

Examples of If conditions

These examples show possible conditions applied to single or multiple response questions

The post Routing rule expressions appeared first on SnapSurveys.

During the interview the question codes are re-arranged in the specified order. The code numbers remain the same and the transformations in the background record the response against the correct answer code.

Setting code rotation on a question

1. Select the question in the Questionnaire window.
2. Click on the Variable Properties button .
3. Select the rotation option from the Ordering drop-down list.

• None: the codes will always appear in the order set in the questionnaire.
• Inverse: the codes will be presented to the respondent in the normal or inverted order at random, For example given codes “Good”, “Ok”, “Poor”, they may appear in the order “Good”, “Ok”, “Poor” or “Poor”, “Ok”, “Good” at random.
• Forward: the codes will start at a random code and then be presented in forward order.
• Random: the codes will be randomly re-arranged for each interview.
• Alphabetic: the codes will appear in alphabetical order.

You can also set code rotation in the Variable Details window.

1. Double click the variable in the Variables window.
2. Click on the Definitions button to display the rotation option
3. Select the order from the Ordering drop down list on the right hand side.

Excluding the final code(s) from the rotation

There is often a requirement to leave some answer codes in their original positions. For example, the question shown below has answer 8 as “Other”.

You can exclude the final codes from the rotation in the Variable Details window or the Variable Properties dialog.

In the Variable Details window:

1. Double click the variable in the Variables window.
2. Click on the Definitions button to display the rotation options
3. Click the Ex button to open the Ordering Excluded Codes dialog containing a list of all the code labels.
4. Select the codes that you want to exclude from the ordering. These codes will always appear at the end for alphabetic ordering.
5. Click OK then click Save to save the changes to the variable.

In the Variable Properties dialog:

1. Select the question in the Questionnaire window.
2. Click on the Variable Properties button .
3. In Ordering Exclude, enter the codes that you want to exclude from the ordering. For the example given, type 8 to apply ordering to the first 7 codes while leaving code 8 fixed. If you want to exclude more than one code then enter the codes as a comma separated list; for example, ‘7,8’.
4. Click OK then click Save to save the changes in the questionnaire.

Enabling code rotation for manual data entry

When entering data entry manually, you can rotate codes by selecting the option in Data Entry Tailoring.

1. Click to open the Data window.
2. Click to open the Data Entry Tailoring dialog.

3. Check the Rotate Codes option and click OK.

The post Code rotation appeared first on SnapSurveys.

Question routing is where respondents are only asked questions relevant to them when completing the questionnaire. Questions that are not relevant to them are skipped. You use Routing Rules to set up routing in a questionnaire.

There are three types of routing rules that can be used in a Snap questionnaire.

1. Conditionally Ask Question
2. Goto On Answer
3. Goto After Question

A rule can either be added at the beginning of a question, within a question code or at the end of a question.

Routing can be hidden or displayed in any of the publication types, for example, a web questionnaire containing routing would ensure respondents only view the questions they are asked as long as Always is selected in the visible list box. This is located in the Routing Rules dialog. The 3 options available are:

Always – routing is always displayed on the questionnaire in all publication methods.

Never – routing is not shown on the questionnaire in any form of publication method.

Paper only – routing is shown on paper questionnaires, but not web surveys.

Routing Rules dialog

The routing rules dialog is displayed if you click on the Questionnaire window toolbar or right-click and select Routing Rules from the context menu. It displays all routing rules for the currently selected variables

Routing Rules Summary Table

The following table explains where the rules of each type appear in a question and how they relate to other questions in the questionnaire.

If a question includes a combination of rules, each rule is treated in the order that it appears in the question. For example, if the first rule asked is a Conditionally Asked Question, only those who meet the condition are asked the question. If the next rule in the question is a Goto On Answer, those remaining respondents who match this rule move to their target question. Any remaining respondents are asked the next rule and this continues until the last rule has been asked. As each condition is met, the respondents are moved to their target question immediately.

Routing rules tutorial

The description of routing uses an example survey, Activate. The survey asks about membership in a leisure club. The example below shows the Activate questionnaire with routing applied.

• Existing members would be asked Q1, Q2 and Q5
• Non-members who are interested in joining would be asked Q1, Q3, Q4 and Q5
• Non-members with no interest in joining would be asked Q1, Q3 and Q5.

Routing rules are added to this questionnaire to ensure respondents are asked the relevant questions.

Setting up a Conditionally Ask Question routing rule

The Conditionally Ask Question rule is set up at the beginning of the question and is situated within the question text. Snap looks at the rule before the question is asked and the question is only asked if the conditions are met. If the conditions of the rule are not met, Snap skips this question and moves onto the next question.

• one condition is allowed per question
• the condition can reference more than one question
• the questions referenced must be previously asked question on the questionnaire

If you create a condition that will not work, such as referring to a later question, the rule is highlighted in red.

The Activate survey contains Q4 “Please give your name and address” This question only needs to be asked to respondents who ticked “Yes” to Q3 “Would you be interested in becoming a member here?”

In order to set up this scenario, you must create a routing rule within Q4 to identify who needs to be asked this question.

1. Click anywhere within the question you wish to set routing on (here, in Q4 “Please give your name and address”).
2. Right-click and select Routing Rules from the menu. You can also click on the Questionnaire window toolbar.

3. Click Add to create a new routing rule. The example below shows that no routing rules have been set up for Q4 as yet.
4. Select Conditionally Ask Question as the Type and click OK.

The Routing Rules dialog expands and fills with information. (You can also see this by clicking Details>> on the Routing Rules dialog and hide it by clicking Details<<.)

The cursor is located in the If box.

1. Type your routing condition. In this example, “Q3=1”. The “1” refers to code 1 of Q3, which is the first answer “yes”. Further details on routing rules can be found in Routing rules expressions.
2. Type ‘If ‘Yes’,‘ in the Text box. This text can be displayed on the questionnaire
3. Select in which publication mode, if any, you would like the routing to be displayed from the Visible drop-down list.
4. Click OK to save your rule.

The rule has been set up so:

• when the If condition is true (yes is ticked in Q3), respondents will be asked the host question (Q4)
• when the If condition is false, the host question (Q4) will be skipped and a response of Not Asked will be recorded

The text you created within the Routing rules dialog has been inserted within Q4 and is denoted with an arrow highlighted in blue. Double clicking anywhere within the highlighted text will allow you to edit.

Setting up a “Goto on Answer” routing rule

The Goto on answer rule is set up within the question codes (answers) of the question. This rule will direct a respondent to a target question within the questionnaire if they tick a particular answer code. For example, if they tick yes to Q1 they go to Q2, if no is ticked, they go to Q3.

This routing type can only be used in question codes and should be used with questions containing a small number of codes. In other cases, use the other two routing rules to decide whether a question is asked or not.

Goto on answer rules are normally only used for single response questions. If a multiple response question has Goto on answer rules on several codes going to different question numbers, Snap follows the routing that goes furthest down the questionnaire.. For example, if a question has routing on one code to go to Q7 and another has routing to go to Q8 and the respondent selects both codes, Snap will automatically jump to Q8.

In the Activate survey, Q1 asks respondents ‘Are you a member of Activate leisure club’. Q2 asks ‘For how long have you been a member’ and Q3 asks ‘Would you be interested in becoming a member here’. This example shows how to set up a Goto On Answer routing rule on the Yes and No codes of Q1 to direct the respondents to the relevant questions.

1. Click anywhere in the Yes answer code of Q1.

1. Right-click and select Routing Rules from the menu. (You can also click on the Questionnaire window toolbar.)
2. Click Add to create a new routing rule. The example below shows that no routing rules have been set up for Q4 as yet.
3. Select Goto on Answer as the Type.
4. Select Answer 1 (Yes) from the On.

5. Click OK.

6. Select Q2 (For how long have you been a member?) in the Goto list. This is the question that you want the respondent to go if Answer 1 (Yes) has been ticked.
7. The text Go to Q2 appears in the Text box. You can edit the text by clicking and typing inside the box. Q2 is highlighted in blue, which means you can change the information that appears by double clicking on it to open the Goto Target’s Name dialog.

You can select from Full Name, Display Name and Grid Name. Display Name is the default and will display the name of the question as shown on the questionnaire. Full Name will display the full name of the question even if not shown on the questionnaire. Grid Name will display the full grid name as specified in the numbering topic.

8. The Style section enables you to alter the number of columns the codes display in. This is useful if the text will be squashed by the Goto instructions.
9. Click Add to create a new routing rule for the No code on Q1.
10. Select Goto On Answer is in the Type field.
11. Select Answer 2 (No) from the On field and click OK.

12. Select where you want the respondent to go if answer 2 (No) has been ticked from the Goto box. In this case Q3 has been selected as the question to go to, and the default text Go to Q3 appears in the Text field.

13. Click OK.

Modifying routing within your questionnaire

Using the Questionnaire window for adding routing is a very easy way to add and modify routing rules.

A routing rule can be moved around the questionnaire by clicking and dragging on the highlighted text.

1. Click and drag to move the rule and drop once positioned in the right place within the questionnaire.

2. To edit the rule, double-click anywhere within the blue shaded area.
3. The routing rules dialog is displayed. You can edit the rule by clicking Details>>, delete it by clicking Remove, or create a new rule by clicking Add.

4. If questions are moved or copied within the questionnaire, the routing expression that refers to that variable will automatically update.

To help you track the routing rules you have set up in the questionnaire, a summary of this information can be viewed in the Variables window.

Setting up a “Goto After Question” routing rule

This rule is placed at the end of a question and is used to direct respondents to the next appropriate question, to a new section or to the end of the questionnaire. Any number of Goto After Question rules are allowed per question. Once a question is asked the respondents will jump to the next relevant question as directed by the contents of the rule and the answer they have given.

In the Activate survey, Q2 asks respondents “For how long have you been a member?” After they have answered this question we want those respondents to go straight to Q5.

To set up this example:

1. Click anywhere in the question where you wish to add routing (Question 2 in this example).
2. Click on the right mouse button and select Routing Rules from the menu. Alternatively, click on the button on the Questionnaire window toolbar.
3. Click Add to create a new routing rule. The Details>> button allows you to view the content of existing routing rules that have already been created.

4. Select Goto After Question from the Type box and click OK.
5. The next step is to select where you want the respondent to go from the Goto box. In this example, select Q5 “Do you have any comments you would like to make?”

6. Your cursor is located in the If box: this is where you can type in an IF condition if required. In this instance all of the respondents need to be directed to Q5 so leave the If box blank. Further details on routing rules can be found in Routing rules expressions.
7. You can alter the text shown on the questionnaire by typing in the Text box. This text will be placed in a footnote at the end of the question.
8. Click OK. The routing is set up as shown in the example below.

Routing limitations

Routing is quite straightforward if only one thing is happening (for example, if people answer “yes” to question 1, jump to question 3). However, when you start doing more complicated things, there can be limitations on what is possible.

Because of this, the different editions have different routing capabilities

Online routing limitations

Browsers and devices running Snap Offline Interviewer support all forms of routing. To ensure that the pages with routing on can be built successfully, page breaks may be inserted before a question with routing on it.

Page breaks are inserted if there is text substitution or complex routing (routing that depends on other routing) on that question. Snap has to insert a page break before it to ensure that all the substitutions and routing dependencies are set up before the question is displayed.

Page breaks are not inserted if the question has one level of “Not Asked if” routing.

You can check whether you have simple or complex routing by viewing the routing dependencies. You will only see which page breaks have been inserted when you publish your questionnaire.

Checking your routing dependencies

1. Click on the Snap toolbar to open the Variables window.
2. Select the question variable that has routing and the Sources button on the Variables window toolbar to open the Sources/Dependents window
3. The Sources pane displays a list of the other variables that this question needs to know the answer to (or whether they have been answered).

4. If there is more than one source variable listed, a page break will be inserted before this question. If there is only one source variable, a page break will not be inserted unless on-page routing is disabled or unavailable.

Routing using paradata

You can use routing to provide different respondents with different surveys

For example, you can route on the ID.site paradata variable to make your survey look different if they have come to it from a different site. This is done by using special links to connect to the same online survey. When a respondent clicks the link, it identifies the survey site, and sets up this information in the Id.site paradata variable. You can then use survey routing to make questions appear or disappear, create derived variables, and so on.

This example assumes that you are running a website satisfaction survey on three different sites, which represent three different branches of your organisation. The same survey will display a different title, depending how you arrived. The example includes brief instructions on writing an HTML file to test that the links to the different versions of the survey work.

Viewing routing

The procedure described below will give you an overview of which variables contain routing. If you set up a Conditionally Ask Question routing rule in the Questionnaire window, a Not ask Value is also created. If Go To On Answer and Go To After Question routing rules are set up, relative information is automatically set up in the Goto and Skipped by columns.

1. In the Variables window click on to access the Variable Tailoring dialog box.
2. Either Change the Show Counts option from Always to Never and click on Keep or Use OR select the routing tick box.

3. The Not Asked column reveals any routing in variables. The Goto column displays the target questions and the Skipped by column displays the question where the routing originates. The Valid, No Reply and Errors columns are not displayed.

The post Adding routing appeared first on SnapSurveys.

You can use a static mask to always hide code boxes, for example, if you have a grid question and wish to hide some responses for some grid elements.

You can use a dynamic mask if you want to hide code boxes according to the response to previous questions.

Applying a static mask

1. Select the question containing code boxes you wish to mask in the Questionnaire window. Right-click and select Variable Properties from the menu to display the Variable Properties dialog.
2. Enter the mask expression in the Mask field. For example, to hide code five, type not 5. To show codes 2 to 4 and hide all other codes, type 2~4 in the Mask field.
3. Click OK to save your settings.
4. Click to preview your mask.

Note: the Mask will not be visible in the Questionnaire Design Window.

Applying dynamic masking

Dynamic masking is frequently used to hide or display responses that have not been mentioned in a previous question.

The example below deals with questions with identical codes. The second question (Q2) will have codes hidden depending on the responses selected on the first question (Q1).

1. Right-click the question that you wish to apply a mask to.
2. Select Variable Properties from the menu to display the Variable Properties dialog.
3. Enter a mask expression in the Mask field. You can use any logical expression as
   • To hide all codes selected in Q1, type not Q1 in the Mask field.
   • To only display codes that have been selected in Q1, type Q1.
   • To display codes that have been selected in Q1 or Q2, type Q1 or Q2 in the Mask field.
   • To display codes that have been selected in Q1 or Q2 but hide codes that have been selected in Q3, type Q1 or Q2 and not Q3.
4. Click OK to save your settings.
5. If it is likely that there will be only one code available to choose, and if so, the respondent must select that one code, you can enable Auto answer. This means that if there is only one code available, that response will automatically be selected and teh question will not be displayed.
6. Right-click the question that you wish to apply auto answer to. (This is usually a question with a mask applied).
7. Select Variable Properties from the menu to display the Variable Properties dialog.
8. Select Yes in the drop-down list by Auto answer.

Creating and applying a conditional mask using a derived variable

You can create a dynamic mask where the possible responses displayed depend upon your answers to previous questions.

The example below explains how to define a derived variable that is used to mask departments, depending which office someone works in. It assumes that you have a survey where Q1 asks which office someone works in and Q2 asks which department they work in.

Q1 is assumed to have three codes:

The mask variable will be a multi-choice variable with the same number of codes as the variable (Q2) that it is masking.

Creating the mask

1. Click to open the Variables window.
2. Select the variable to which you wish to provide a mask for.
3. Click to clone the variable.
4. The new variable will open. It will have the same codes as the variable it is based on. Change the Type to Derived.
5. Change the name of the variable to make it clear that it is a mask variable, (e.g. DepartmentMask) and change the Label to explain what you are going to use it for.
6. Change the Response to Multiple.

Set up the code Values. For example, if they have selected the code for London in Q1, only the Customer services and Training departments are visible.

5. Click to save the changes.

Applying the mask

1. Select the question to mask in the Questionnaire window.
2. Right-click and select Variable Properties from the menu to display the Variable Properties dialog.
3. Enter the mask variable you have created in the Mask field.

4. Click OK to save your settings.

You can check your mask by previewing the survey.

The post Using masking appeared first on SnapSurveys.

In data entry, patterns are used to

• Ensure that respondents or data entry personnel enter the correct format for information such as postcodes, zip codes or National Insurance numbers. Pattern matching is used to check whether the answer is a possible match in the pattern.
• Convert the data into a standard form, such as adding currency symbols in quantity data.

In filters and derived variables, you can

• Use the pre-defined patterns to temporarily change data, for example, converting literal responses to lower case
• Use one part of a pattern made up of several defined parts for example, the area code in a phone number, and use that part to analyse the data.

Snap XMP Desktop is supplied with a library of pre-defined Patterns. You can also create your own using the Patterns dialog from View|Patterns.

System patterns

Snap XMP Desktop is supplied with a library of pre-defined patterns that include patterns for:

• Literals such as email address, postcode and zip code
• Dates such as date in words
• Quantities such as currency and decimal number
• Components that can be used to build your own patterns, but cannot be used directly in variable details or as filters. For example letter or digit.

For each pattern there is

• A unique pattern name
• A form which identifies which types of variable can use the pattern
• A label describing how the pattern is used.

Applying a pattern in the questionnaire window

1. Open the survey in the questionnaire window.
2. Select a question, right-click and choose Variable Properties.
3. Ensure you have the right Response selected for the question. Patterns are only available for open-ended questions, including date, time, literal, and quantity questions.
4. Click the Source Pattern list to view the existing patterns for use with the variable.

5. If the pattern you were expecting to see is not showing you might need to change the Response type for the question. Only those patterns that are relevant to the response type are shown in the list of Source Patterns.
6. Click OK to apply the pattern to the variable or Cancel to abandon any changes that you have made.
7. Select in the Questionnaire window to save the Pattern and any other changes to the questionnaire.

Applying a pattern in the Variables window

You can set the pattern for a variable using the Variable Details dialog.

1. Click to open the Variable window.
2. Select the variable to which you wish to apply a Pattern and double click to open the Variable Details dialog.
3. Click to toggle definitions of the variable and you will see that the Variable Details now display information including Pattern.
4. Click the Pattern drop-down and a list of patterns, applicable to the response type, is shown.

5. Choose the required pattern and click to save the changes.

Patterns during data entry

There are two ways of using patterns in data entry.

• Check the data against the requirements. If it does not match, display an error message.
• Check the data against possible matches and change it to a standard format. If the data entered conforms to a number of entry options (for example, “two”, “2” or “£2”) included in the pattern the data is accepted then converted to a standard format (for example, all of the above responses are converted to “2” for analysis).

Composition of a pattern

Patterns are composed of components and matches.

• matches are text strings that a response must match
• components consist of one or more text strings that part of a response must match

Using components within matches means that:

• specific parts of a respondent’s answer can be identified and scrutinised at the analysis stage of the survey process
• a consistent style of result can be output and analysed

In the example below, the pattern has one match and four components.

In addition to the text string used to match the response, you can add further details about the pattern.

• you can add data in the Reference field about where you accessed the information for the pattern; for example, a website address; a database or a reference book
• you can specify what type of response the pattern can be used within the Used In field (select Other Patterns Only for patterns that will only be used when creating another pattern)

The example below shows two Matches for year; {4 digits} and {2 digits}, signifying that there can either be 4 digits or 2 digits representing the year part of the date.

The Matches for the month component can be seen below. There are 5 matches within the component, signifying that either 12, 11 or 10 is acceptable or a zero followed by a digit, such as 05 or a single digit, such as 5 (to represent May).

The separator component of this DD/MM/YYYY pattern has been set up to accept dashes, obliques and full stops as the divider between the various parts of the date.

Creating a new pattern example

This example shows how the DD/MM/YYYY pattern can be created. This ensures that the respondent enters a date with day-month-year format, where the year part can include a two-digit century.

The separator between day/month and month/year can be any of “/”, “-” or “.”.

All these are valid dates: 06/08/2020; 6-8-2003; 06.8.2020; 6/08/20 for this pattern.

Examples of invalid dates are: 6^th August 2020; 06/Aug/2020.

1. Select View | Patterns to open the Patterns dialog. (Prior to build 12.10 this was on the Tailor menu.)
2. Select New in the Patterns dialog to display the Pattern Properties dialog.
3. Choose a name for the pattern and type it in to the Name box. Try to make the name as specific as possible to its purpose. For example, rather than write it as Date write it as DD/MM/YYYY so that its format can be identified easily by others using it.
   • Enter a Label giving more information. This will be displayed when the pattern is selected on the front page of the Patterns dialog.
   • You can enter a reference in the Reference giving specific details about where you accessed the information for the Pattern: for example, a website address; a database; a reference book.
4. Select the response type that the pattern will be used with from the Usage drop-down list. Select Other Patterns Only for patterns that will not be available directly to variables, but only for use within other Patterns.

5. You can create patterns by using a plain text string as a match, however, it is sometimes more effective to build the match up from components.

Creating components to use in your match

1. Click More to display the Components part of the Pattern Properties dialog.

2. Click New to the right of the Components area to display the Component Properties dialog.

3. Type in a Name for the Component you are writing. For example for DD/MM/YYYY you might call one of the components ‘Day’, which you would type in the Name space. In Label you might write ‘Day part of Date’.
4. Select New to create a Match for the Component.
5. The Match Pattern dialog displays.
6. You build up the pattern to match from text and components. You can use the buttons on the dialog to select items, or right-click the Match field to display the context menu.
7. Either type a text string into the Match box or double click one of the Global Patterns to select it.
   • For the day component you would want one digit or two digits recognised, as respondents would be likely to type in either one or two digits when answering this part of the question. For example: 01 or 1 when referring to the first day of the month.
   • If you are looking for a particular character click Characters… to show more options in the Character Selection dialog.
   • Highlight the characters required and click OK. If a letter is selected, both the lower and upper case version will show in the match unless Case Sensitive is ticked.

8. For the first two-digit match of the Day example:
   • either double-click digit in Global Patterns twice
   • or double-click digit and then use Repeat to permit multiple copies.

The Repeat button gives options for each part of the Component. Ensure that the part of the Component you want to affect is highlighted before selecting a Repeat command. Each of the commands creates constraints which the data must adhere to in order to fit the pattern.

9. Click OK to return to the Match Pattern dialog and OK again to return to the Component Properties dialog.
10. Once satisfied with all the Components, return to the Pattern Properties dialog.

Deleting components

If you wish to delete any of the Components, highlight it and click Delete. You are shown a warning message and asked to confirm that you want to delete the component.

Creating the match from the components

1. On the Pattern Properties dialog you will see that you now have a list of Components but no overall Match for the pattern. Select New under the Matches part of the dialog and the Match Pattern dialog will display again

2. The newly built components will show in the left-hand column. You can now create a match from both your components and the existing Global Patterns. Characters and Repeat work in the same way as previously explained.
3. Select the year (4 digit) component, then month and day.
4. Click OK when you have finished to return to the Pattern Properties dialog.

Testing a pattern

1. Once you have created, cloned or modified a pattern, you can test it by clicking Test on the Pattern Properties dialog. The test dialog appears.

2. Type in a sample of data into the Text box. For example, for the date pattern you might type in 20201231.
3. If the pattern works correctly it will show that it has matched the data, the result that will display in Snap and the Values of each of the component parts.

If the test fails either the text does not fit the pattern or the pattern needs editing.

Result and Modify Case only alter the way the data is displayed in the analysis.

Modifying, cloning and deleting patterns

1. Select View | Patterns to open the patterns dialog. (Prior to build 12.10 this was on the Tailor menu.)
2. Highlight the required Pattern and click either Clone or Modify to access the Pattern Properties dialog box and then follow the instructions for creating a pattern provided earlier in this section.
3. If you wish to delete a pattern, click Delete. You are shown a warning message and asked to confirm that you want to delete the component.

You are unable to delete any of the System patterns provided with the software.

If you delete a component, you will need to modify any matches that contain the component, otherwise the associated pattern will not work correctly.

Always ensure that the longest match string is at the top of the list to improve the search efficiency when the pattern is in use. For example in a list of Matches of {digit} {digit}, {digit} {letter}, {digit} they should be listed in this order to identify answers with two digits before classifying them as errors from the other two matches.

Copying patterns between surveys

If you already have a survey containing patterns that you need, you can copy these to another survey by saving them in SnIF format or to the clipboard.

1. Open the survey containing the patterns you wish to copy.
2. Select View | Patterns to open the Patterns dialog. (Prior to build 12.10 this was on the Tailor menu.)
3. Click Export to display the Pattern Export dialog box. The format will be set to SNAP Xml Format (SnIF Xml).

4. Set the Destination as Clipboard. If you are saving a very large amount of data, you may need to save to a file instead.
5. Enter the names of the patterns you wish to export in the Content field (leave blank to export all user-created patterns).
6. Click OK to export the selected patterns. A confirmation message appears showing the number of patterns exported (this will not include standard patterns provided in all surveys).
7. Open the survey that you wish to copy the patterns to.
8. Select View | Patterns to open the Patterns dialog. (Prior to build 12.10 this was on the Tailor menu.)
9. Click Import to display the Pattern Import dialog box. The format will be set to Snap Interchange Format (SnIF).

10. Set the Source to the file or Clipboard that you previously exported the patterns to.
11. Click OK to import the patterns. A confirmation message appears showing the number of patterns imported.

The post Introduction to patterns appeared first on SnapSurveys.

The expression must be a list or range of possible values.

These can be

• limited range For example, 1 TO 5 or 1~5
• unlimited range using relational operators >(greater than), < (less than), >= (greater than or equal to), <= (less than or equal to), <> (not equal to). For example >10 would include all values greater than 10
• ranges defined by variables in the questionnaire. For example <Q9 would include all values less than the response to Q9
• a list For example f, F, m, M or 1,2, >6

You can set the valid range:

• by opening the Variable properties dialog from the questionnaire window and editing the value of the Valid field

• by opening the Variable details dialog from the Variables window and editing the Valid field.

To use the Valid property for open responses in a Web surveys, you must also check the Validate Open Responses option in the Interview section of the Questionnaire properties.

Note: You can use a variable within a limited range (e.g. Q9 to 15).

If you wish to use a relational operator within a limited range (e.g. >Q9 to 15), you cannot do so directly. Instead you must create a derived variable (e.g. V1) that contains the result of (Q9 + 1), and then set the valid range to be V1 to 15.

Setting a valid range for a number

1. Open the survey in questionnaire mode.
2. Select the question that you wish to set a valid range on.
3. Click on the window toolbar to open the variable properties dialog. (You can also press Alt+Enteror right-click the question and select Variable properties from the context menu.)
4. If the Valid field is not visible, scroll the list of fields to display it.

5. Enter the valid range. For example 1 to 100 or <101. This would set a maximum of 100 for a question using percentages.
6. Click OK to save the changes.

Setting a valid entry to match a previous entry

If you wish a respondent to type an identical entry twice, such as, to confirm an email address, you can set a Valid response of the previous variable.

1. In the Questionnaire window select the question that you wish to set a valid range on.
2. Click on the window toolbar to open the variable properties dialog. (You can also press Alt+Enter or right-click the question and select Variable properties from the context menu.)
3. If the Valid field is not visible, scroll the list of fields to display it. Enter the variable name that you wish to match.

4. Click OK to save the changes.

The post Setting valid ranges on a question appeared first on SnapSurveys.

• setting the response type (if you did not do so when creating the question) to specify whether respondents can enter a single choice, multiple choices, free text, dates etc.
• setting up routing to direct respondents through the questions in different ways according to how they asked previous questions
• setting up patterns to limit what responses they can provide for a question
• seeding the survey with data such as names and addresses for a mail merge or user names for an online survey
• setting up question and code rotation so that respondents are presented with questions in a random order

Not all of these options are available for all outputs, for example, you can’t perform question rotation on a paper questionnaire.

Controlling how the questionnaire is delivered

Before you publish your survey, you can choose how the questions will be presented to the respondents. You can:

Change what appears in the questionnaire

• change the questions presented using question routing (skip logic) so that questions are asked or not asked depending upon respondent’s previous answers
• change the question text depending upon previous answers (this is known as dynamic text or piping)
• display calculations carried out on previous answers using in-survey calculations (derived variables)
• change which question codes are visible using masking
• reduce the effects of position bias by randomly rotating the order of codes and questions
• provide several versions of the same survey tailored to different topics

Limit or require answers

• force respondents to answer questions
• limit the possible responses by applying a valid range of answers
• apply patterns to make sure an answer is valid
• send an email when you receive a particular answer

You cannot change questions for each respondent for paper questionnaires, but you can check the answers using patterns when entering the data.

You can also provide a tailored questionnaire that people can complete a number of times on different subjects by using group questionnaires (or iterative surveys).

The post Introduction to survey logic appeared first on SnapSurveys.