The add new volume dialog in AskiaSurf allows you to add a new volume (either a file or database) to the selected wave.
The dialog consists of three screens. The back and next buttons navigate between the three screens. You cannot move on from a screen until you have completed all its mandatory fields. When you have finished specifying the volume details on each screen, click done on the final screen to close the dialog.
This screen allows you to select the file/s or database/s that contain the question and interview data for this volume. It contains the following fields:
Name: Enter a descriptive name for the volume; this will appear in the campaign tree. You must complete this field, but the name does not have to be unique.
Source type: If you want to read data from a file, as opposed to a database, select the type of file here (QES, triple-S, or SPSS Dimensions). Note: in the present version of askiasurf, only QES files are supported.
Question source: In this section of the dialog, specify the details of the file or database that contains the survey question data.
Read questions from file: select this option if you want to read the questions from a file, as opposed to a database. Either type the path to the file, or click browse... and select the file.
Create working copy of the source file: If this option is selected, a copy of the source file will be made in the same folder as the campaign, and AskiaSurf will use this copy, rather than the original file. This protects you from accidental modification of the original source file, as AskiaSurf will be using the "safe" copy.
Read questions from database: select this option if you want to read the questions from a database, instead of a file. Enter a name for the database connection, and click connect. The connect to database dialog appears, allowing you to specify the details of the database in question.
Interview source: Interviews are stored with the questions: Select this option if the source file or database that contains the question data (i.e. the file or database you specified in question source; see above) also contains the answer data. If you have separate files/databases for question and interview answer data, de-select this option. The remaining options apply only if you have a separate file/database for interview data (i.e. if you de-selected interviews are stored with the questions).
Read interviews from file: select this option if you want to read the interview data from a file, as opposed to a database. Either type the path to the file, or click browse... and select the file.
Create working copy of the source file: If this option is selected, a copy of the source file will be made in the same folder as the campaign, and AskiaSurf will use this copy, rather than the original file. This protects you from accidental modification of the original source file, as AskiaSurf will be using the "safe" copy.
Read interviews from database: select this option if you want to read the interview data from a database, instead of a file. If you are connecting to a database that uses the old SQL Server table format (multiple tables per survey), rather than the new (single table) format, select interviews are stored using the legacy format. Enter the connection string for the database connection, and click connect. The connect to database dialog appears, allowing you to specify the connection string to the CCA and the connection string to the AskiaField surveys database.
File path format: Select the format in which you want to store the file pathnames (relative, absolute or UNC). By default, relative is selected, which is recommended for files on the local PC. For files stored on a network drive, we recommend you select UNC. Note: this option does not apply if you are loading both question and interview data from database/s.
This screen allows you to specify whether the volume uses inheritance, and if so, from which previous volume it should inherit.
This screen contains the following fields:
Not using inheritance: Select this option if you do not want this volume to inherit from a volume in the previous wave. However, if there is a compatible ancestor volume, we recommend that you use inheritance; this not only improves reconciliation performance, but also alleviates the need for you to re-make manual question and response matches from previous waves.
Inherits from another volume: Select this option if you want this volume to inherit from a volume in the previous wave. You should then select the appropriate volume from the previous wave; see volume list below.
Volume list: This list displays the volumes in the previous wave, indicating their names, the relative size of the volume being added (e.g. if your new volume has 10% more questions than the volume in the list, this figure would be 110%), and an estimate of their compatibility with the new volume (based on the number of matching question shortcuts). Normally, you would want to select the volumes with the highest compatibility.
This screen allows you to specify how the questions and responses in the new volume will be matched to the questions and responses already in the campaign. You must select at least one question matching method and one response matching method.
done to close the dialog.This screen contains the following fields:
Question matching: Choose the method or methods that will be used to match questions in the new data volume to those already in the campaign.
Select the methods from the following:
Internal code: to match by internal question codes
Shortcut: to match by question shortcut (this is selected by default)
Short caption: to match by question short caption
Long caption: to match by question long caption
Manual: askiasurf leaves the matching for you to do manually (not recommended as this removes much of the advantage of using askiasurf)
User data: to match by question user data
Order: to match according to where it appears in the project
To quickly select or de-select all methods, click the check-box next to matching method.
You can drag and drop methods into the order in which you would like them to be applied (see match using first successful method and match strictly using all matching methods, below).
Note: You must select at least one question matching method.
Response matching: Choose the method or methods that will be used to match responses in the new data volume to those already in the campaign.
Select the methods from the following:
Internal code: to match by internal response codes
Entry code: to match by data entry code (if your survey uses data entry codes, and they are frozen, using this option is recommended; if the entry codes are not frozen, do not use them to match responses)
Short caption: to match by response short caption (this is selected by default)
Order: to match by response order (i.e. the order in which responses appear)
Manual: askiasurf leaves the matching for you to do manually (not recommended as this removes much of the advantage of using askiasurf)
User data: to match by response user data
To quickly select or de-select all methods, click the check-box next to matching method.
You can drag and drop methods into the order in which you would like them to be applied (see match using first successful method and match strictly using all matching methods, below).
Note: You must select at least one response matching method.
Match using first successful matching method: Causes campaign questions to be matched to volume questions if any of the selected methods (see question matching, above) are successful (the matching methods will be carried out in the order in which they appear in the question matching or response matching list). For example, if shortcut and short caption are selected as matching methods, and the shortcuts of two questions are the same, the two questions will still be considered matched, so it will not matter whether the short captions are the same or not.
Match strictly using all matching methods: Causes campaign questions to be matched to volume questions strictly: all of the selected methods (see question matching, above) must be successful. For example, if shortcut and short caption are selected as matching methods, and the shortcuts of two questions are the same - but the short captions are not - then the two questions will not be considered matched.
Case-sensitive matching: Causes the case of shortcuts and captions to be checked (so Q1 will be considered different from q1).
Match using parent volume question and response values
Do not add any new questions: If this option is selected, when a question is found to have no match, it will not be added to the campaign.
Selective question matching: If match only the selected source questions is selected, then only the questions you select in the list will be checked for matching. If it is not selected, then all questions will be checked.
This screen provides further options that specify how the questions and responses in the new volume will be matched to the questions and responses already in the campaign.
Align interviews in this volume to the main volume by interview order: If you have multiple volumes in a single wave, the data about a given respondent in one volume needs to be matched with the data about that respondent in the other volumes. In other words, the data needs to be pulled together from the different volumes to make one complete record. If you select this option, AskiaSurf assumes that records appear in the same order across volumes. For example, the first respondent record in the first volume is assumed to be the same as the first respondent record in the second volume, and so on. This option is not available for the first volume in a wave, because at that stage, alignment is not applicable. It is available for the second or subsequent volume in a wave.
Align interviews in this volumes to the main order using a common question: If you have multiple volumes in a single wave, the data about a given respondent in one volume needs to be matched with the data about that respondent in the other volumes. In other words, the data needs to be pulled together from the different volumes to make one complete record. If you select this option, AskiaSurf assumes that records share a common identifier question, such as an ID number or a name. This question must be numeric or open-ended. AskiaSurf does not check that there is unique data in the specified question, so you must ensure that your identifier variable contains unique data for each respondent. In main volume questions and this volume questions, select which questions will be used to identify interviews. You can only select a question that is of open or numeric type. This option is not available for the first volume in a wave, because at that stage, alignment is not applicable. It is available for the second or subsequent volume in a wave.