In this article
Respondent data
The question types and their limits are as listed below:
- Open text - Unlimited
- Open text with field width specified in properties - Limited to the number of characters specified
- Numeric - 38 digits
Note: Excel has a display limit of 15 digits for numbers, so if a numeric question larger than this is exported to Excel then the numbers will be rounded and displayed in exponential notation. This is an Excel limitation so cannot be controlled by Forsta Plus. Two viable work-arounds are available here:
- Export the file as a text file and then import it to Excel in text format.
- If the numeric question is shown on a hitlist within Reportal, exporting the hitlist will result in the limitation. In this case, change the question type to Open text in Authoring, re-launch the survey and force the BitStream files.
- Coded questions (single, multi and grid) - 32 characters (default), but field width up to 40 characters may be set
- Multi with open text property - (same as open text)
- Numeric List - (same as Numeric)
- Other specify-element ("other" property set in answer list) - Unlimited
- Loop - (same as coded questions)
Note: The respondent table has a maximum limit of 1024 columns. However users should not exceed 900 columns as Forsta Plus also has a varying number of system-derived respondent columns and in subsequent releases may add to these. If the total column count exceeds 1024 the survey launch will fail.
Note: The respondent table for legacy databases has a maximum limit of 8060 characters per record (adding all maximum allowed 1024 columns together). This limit also includes the system variables, therefore the actual row size limit per record is less than 8060 characters. The system variables can take a maximum of 1221 bytes if all values are specified. This situation is however rare, and a realistic value is approximately 600 bytes. In the event the 8060 characters limit is exceeded, an error message will be displayed.
When using the optimized format (go to Optimized Database Format for more information) there is no limitation on the number of characters when using an open text question without specifying a length. If your respondents are hitting the limit, you are recommended to change the database to the optimized format.
Meta data
- Question title - 255 characters
- Question text - 32767 characters
- Question instructions - 32767 characters
- Answer/scale texts - 2000 characters
- Question ids - 50 characters
- Codes - 32 characters (default). Only alphanumeric characters and underscore (_), no white space. The survey administrator can increase this limit to a maximum of 50 characters for individual questions (the Field Width property in the Question Properties sheet). Note that this maximum will not be a problem for single questions, however problems will arise during launch if grid and/or multi questions use the full 50-character codes - see the note below.
Note: In the database, the column headers are constructed as a combination of the question id and the code, and are limited to 50 characters. If this limit is exceeded for a header, then an error message will be displayed when you attempt to launch the survey. Care must therefore be taken when allocating codes for grid and multi questions if the code maximum has been increased above its default setting.
Scripts/masks
- Script nodes - 32767 characters
- Validation code - 32767 characters
- Masks (code mask, scale mask, column mask) - 255 characters
Recoded variables
Same as for ordinary question types, except for the extra element
- Expression - Unlimited
Respondent list
- Values uploaded in the respondent list - 255 characters
Note: If the “Use data types of background variables for fields in respondent data” option is selected (go to General Fields for more information), you can set up questions in the respondent list that will accept more than 255 characters. However for open text fields with no field width set in the background variable, or with the field width set higher than 4000, the respondent table will be given a field width of 255.
Data Processing
The service supports various data processing operations:
Data Processing Rules
Data Central Server Rules.
A data export/import wizard, which initiates Data Processing and Data Central Server rules.
Although Forsta has designed the Service to handle the data processing strain generated by the most advanced activity of its clients, extreme use may cause performance issues or errors. Forsta shall not be held responsible for issues relating to extreme use. The definition of "extreme use" will depend the data processing operations performed and the volumes of data being processed. The volume will be a factor of the number of records, number of variables, variable types and how densely populated the data is (i.e. how may of the data cells that have values). The following are general guidelines on the limits for the various data processing tasks, based on data sets of different sizes:
For data processing operations involving data sets with up to 100 variables:
- Data sets run through Data Central Server Rules (including SPSS export) should be limited to 750 000 records.
- Data sets run through Data Processing Rules (including Triple-S and delimited text export/import) should be limited to 1 000 000 records.
For data processing operations involving data sets with up to 1000 variables:
- Data sets run through Data Central Server Rules (including SPSS export) should be limited to 75 000 records.
- Data sets run through Data Processing Rules (including Triple-S and delimited text export/import) should be limited to 100 000 records.
For data processing operations involving data sets with up to 2000 variables:
- Data sets run through Data Central Server Rules (including SPSS export) should be limited to 37 500 records.
- Data sets run through Data Processing Rules (including Triple-S and delimited text export/import) should be limited to 50 000 records.
Translator
- The number of categories and/or scales that can be viewed in the translator interface is limited to 1300 per scale/answerlist.
Archiving
When using the Database Cleanup task, a report with a single data source is soft-deleted a preset number of days after the survey is archived. For SaaS users in the Forsta servers this is set to 14 days. For On-Premise users, the system administrator can adjust the system setting DaysToKeepSurveyDatabasesForDeletedProjects as required.