Browse
 
Tools
Rss Categories

Export

Reference Number: AA-00230 Last Updated: 2017-11-23 11:09

Mobenzi Researcher supports 3 data export formats: 

Only submissions matching the applied filters will be included when exporting. If you have renamed any of the columns in the Grid, they will be used in the export file instead of the default question identifiers.


Large Exports

Should your data set contain more than 500 submissions, a pop-up will appear and you will be asked to provide an email address where a secure link to download the file will be sent once it is ready. Please wait for the download link to arrive in your email account before requesting further exports. Please note that the download link is unique and cannot be forwarded or shared with another user.

Stat/Transfer Format

Stat/Transfer is a widely used third-party utility which enables you to convert data files to over 30 formats including popular analysis programs such as Stata, SPSS, Epi Info and many more. You will need to purchase a copy of Stat/Transfer to make use of this export format.

There are two types of files which are important to understand for the Stat/Transfer export format: the data file (which has a ".dat" file extension) and the definition file (which has a ".stdd" file extension). The data file contains the response data and the definition file contains the associated meta data which describes the response data types, code values, etc.

When selecting the Stat/Transfer export format, you will be able to configure a few export options:

  • How to handle repeating sections
  • Whether to use option values or labels when generating variable names for multi-select questions


Repeating sections

  • By selecting "Export each section as a separate file", a separate data file and definition file pair will be created for each section. These will all be zipped up into a single archive which you can extract after downloading.
  • By selecting "Export repeating sections as separate files", all non-repeating sections will be merged into a single data file and definition file pair. Each repeating section will generate its own data and corresponding definition files. All files will be zipped up into a single archive which you can extract after downloading. 
  • By selecting "Merge repeating sections as additional rows into a single file", all sections will be merged into a single data file and definition file pair. A separate row will be generated for each iteration. For example, if there are two repeating sections in the survey - "Section A" and "Section B" - and for a particular submission "Section A" repeats 3 times and "Section B" repeats 4 times, the file will contain 4 rows for the submission.
  • By selecting "Merge repeating sections as additional columns into a single file", all sections will be merged into a single data file and definition file pair. For repeating sections, additional columns will be generated to accommodate the maximum number of repeats in the data and the column/variable headings will be suffixed with "_0", "_1", "_2" as needed to indicate each instance. Note that this is format replicates the legacy Stat/Transfer export format which was previously the only option available.

Multi-select option naming

There are also two options which can be used to determine the naming convention for multi-select question types. When exporting multi-select questions, a column/variable is generated for each available option which is then represented as "selected" (1) or "not selected" (0).

Assume you have a multi-select question "HOUSEHOLD_ITEMS" with 3 options: Television (value: TV), Radio (value: R) and Fridge (value: F): 

  • By selecting "Append option values to question name", the column/variable names generated for each option will be labelled by appending the relevant option's value to the question name. E.g. "HOUSEHOLD_ITEMS_TV", "HOUSEHOLD_ITEMS_R", "HOUSEHOLD_ITEMS_F".
  • By selecting "Append option labels to question name", the column/variable names generated for each option will be labelled by appending the relevant options's label to the question name. E.g. "HOUSEHOLD_ITEMS_Television", "HOUSEHOLD_ITEMS_Radio", "HOUSEHOLD_ITEMS_Fridge".

Single-select option naming

Likewise with single-select question types, there are two options which can be used to determine the naming convention for single-select question types.

Assume the example question above ("HOUSEHOLD_ITEMS" ) is a single-select question with 3 options: Television (value: TV), Radio (value: R) and Fridge (value: F): 

  • By selecting "Use the selected question's value", the column/variable name generated for the option will be labelled by appending the option's value to the question name. E.g. "HOUSEHOLD_ITEMS_TV" or "HOUSEHOLD_ITEMS_R" or "HOUSEHOLD_ITEMS_F".
  • By selecting "Use the selected question's label", the column/variable name generated for the option will be labelled by appending the option's label to the question name. E.g. "HOUSEHOLD_ITEMS_Television" or "HOUSEHOLD_ITEMS_Radio" or "HOUSEHOLD_ITEMS_Fridge".

Converting from Stat/Transfer to another format

Once the export has been generated, extract the zip file. Open Stat/Transfer and select "ASCII/Text - Stat/Transfer Schema" as the Input File Type. Then browse to the ".stsd" file you extracted from the export download.

Next, select the Output File Type (e.g. Stata) and specify where you'd like to save the output file. You can use the "Variables" tab to select which fields you'd like to include in the output file. For more advanced Stat/Transfer functionality, be sure to review their user manual.

SAS Users

Due to a SAS limitation, the longest variable label (question text) you can have is 256 characters. If you experience an issue processing the converted file using SAS, it may be that a variable label is too long and you will then need to set the SAS variable label to be the same as the variable name.

CSV Format

The CSV format is a widely supported format which can be read by many spread sheet applications such as Microsoft Excel, Numbers and Google Docs. Raw response data is exported into one or more CSV files. In addition to the raw response data the survey's code book, question reference list and submission meta data are also exported into separate CSV files for interpreting and analysing the raw response data:

  • The “Submissions” file contains a summary of all submissions which are included in the export. Only metadata is included.
  • The “Codebook” file lists every question used in the survey. Each question receives a unique ID from the system. For select-type questions, each option is also listed along with its associated value.
  • For reference purposes, a “Questions” file is generated which lists each question contained in the survey, the question's system-assigned ID, and the text displayed and the question type.

When selecting the CSV export format, you will be able to configure a few export options:

  • How to handle repeating sections
  • Whether to use option values or labels when generating variable names for multi-select questions


Repeating sections

  • By selecting "Export each section as a separate CSV file", a separate CSV file will be created for each section. These will all be zipped up into a single archive which you can extract after downloading.
  • By selecting "Export repeating sections as separate CSV files", all non-repeating sections will be merged into a single CSV file. Each repeating section will generate its own CSV file. All files will be zipped up into a single archive which you can extract after downloading. 
  • By selecting "Merge repeating sections as additional rows into a single CSV file", all sections will be merged into a single CSV file. A separate row will be generated for each iteration. For example, if there are two repeating sections in the survey - "Section A" and "Section B" - and for a particular submission "Section A" repeats 3 times and "Section B" repeats 4 times, the file will contain 4 rows for the submission.
  • By selecting "Merge repeating sections as additional columns into a single CSV file", all sections will be merged into a single CSV file. For repeating sections, additional columns will be generated to accommodate the maximum number of repeats in the data and the column/variable headings will be suffixed with "_0", "_1", "_2" as needed to indicate each instance.

Multi-select option naming

There are also two options which can be used to determine the naming convention for multi-select question types. When exporting multi-select questions, a column/variable is generated for each available option which is then represented as "selected" (1) or "not selected" (0).

Assume you have a multi-select question "HOUSEHOLD_ITEMS" with 3 options: Television (value: TV), Radio (value: R) and Fridge (value: F): 

  • By selecting "Append option values to question name", the column/variable names generated for each option will be labelled by appending the relevant option's value to the question name. E.g. "HOUSEHOLD_ITEMS_TV", "HOUSEHOLD_ITEMS_R", "HOUSEHOLD_ITEMS_F".
  • By selecting "Append option labels to question name", the column/variable names generated for each option will be labelled by appending the relevant options's label to the question name. E.g. "HOUSEHOLD_ITEMS_Television", "HOUSEHOLD_ITEMS_Radio", "HOUSEHOLD_ITEMS_Fridge".

Microsoft Excel Format

When selecting the Microsoft Excel export format, you are able to select an earlier Excel version if you don't have Microsoft Excel 2007 or later. Please note that the Excel 2003 format may be discontinued in the near future.

Each section is exported as a separate worksheet. The first 3 columns of each file/worksheet contain key metadata about the submission:

  • [Submission Id] – a system generated identifier which is unique to that submission.
  • [Fieldworker Name] – the name of the fieldworker who uploaded the submission (based on who was assigned the device at the time).
  • [Fieldworker Id] – a system generated identifier which is unique to that fieldworker. Two fieldworkers with the same name will still have unique identifiers.

If you would prefer, you can specify one or more fields to appear in every worksheet by using column fixing on the Grid view.


Repeating Sections

For repeating sections there are additional relevant columns in the relevant worksheet:

  • [Repeats On Question] – the name of the question on which the section repeats. If the section is not repeated, this will be displayed as “N/A”.
  • [Repeat Question Value] – the value which was captured for the question on which the section repeats. This is the number of times the section iterated and thus the number of rows which will be generated for that submission in that particular section.
  • [Repeating Index] – indicates the iteration number of the row in question. This value is zero-based, i.e. the first iteration is 0, the second is 1 and so on.

Multi-select option naming

There are also two options which can be used to determine the naming convention for multi-select question types. When exporting multi-select questions, a column/variable is generated for each available option which is then represented as "selected" (1) or "not selected" (0).

Assume you have a multi-select question "HOUSEHOLD_ITEMS" with 3 options: Television (value: TV), Radio (value: R) and Fridge (value: F): 

  • By selecting "Append option values to question name", the column/variable names generated for each option will be labelled by appending the relevant option's value to the question name. E.g. "HOUSEHOLD_ITEMS_TV", "HOUSEHOLD_ITEMS_R", "HOUSEHOLD_ITEMS_F".
  • By selecting "Append option labels to question name", the column/variable names generated for each option will be labelled by appending the relevant options's label to the question name. E.g. "HOUSEHOLD_ITEMS_Television", "HOUSEHOLD_ITEMS_Radio", "HOUSEHOLD_ITEMS_Fridge".

In addition to the actual response data worksheets, each export includes a number of additional worksheets:

  • The “Submissions” worksheet contains a summary of all submissions which are included in the export. Only metadata is included.
  • The “Codebook” worksheet lists every question used in the survey. Each question receives a unique ID from the system. For select-type questions, each option is also listed along with its associated value.
  • For reference purposes, a “Questions” worksheet is generated which lists each question contained in the survey, the question's system-assigned ID, and the text displayed and the question type.
  • A “Raw Data” worksheet is produced which is useful for importing the response data into other databases or performing other types of transformations.

Rss Comments
  • There are no comments for this article.