|This page describes a feature in AmCAT 3.4
|View other version: 3.3 - 3.4 - 3.5
After your articles have been coded, you can download the codings so you can analyze them outside of AmCAT, for example in Excel, SPSS or R. To do so, first go to the codingjobs overview ('Coding Jobs' in the dropdown menu under the 'Codings' tab).
Here, you can first check whether all articles have indeed been coded. Have a look at the complete and the todo rows in the table on the lower half of the screen. Are there indeed 0 more articles to do (code)? If so, you can download the codings.
From the table, select the codingjobs you want to export, and click ‘Export codingjobs’ to the upper right of the table.
From the dropdown menu, select the level of codings (sentence and/or article codings) you want to export.
In the next screen, you can select variables to include in your datasef as well as its output format.
Selecting variables for your dataset
When exporting coding results from AmCAT, you first have to specify the format of the dataset and the variables to include.
- Include duplicates indicates whether articles that are included in more than one codingjob, should also be exported more than once.
- Currently, CSV, XLSX and json are supported export formats. Although SPSS is not supported, you can import a CSV or XLSX file into SPSS.
- Metadata such as date, headline and medium are coded automatically by AmCAT. Select here which metadata to include in the dataset.
- Under ‘Field options’ – the manually coded variables -- you can select which variables you want to include in the dataset. By default, all variables are included.
- Furthermore, for codebook variables, you can select whether you only want to export its ID (the number given to the code by amcat), its labels (entered for each code in the codebook editor) or both. If you use SPSS or Stata to analyze the data, you might want to download (also) the IDs and then attach the labels to the IDs ('values') in SPSS/Stata itself.
- ‘Export # parents’ is for hierarchical codebooks. Each ‘# parent’ creates an additional variable for a higher level in the hierarchy. For example, the variable Topic, the last variable in the screenshot to the left, has a hierarchical codebook, shown to the right. Two possible topics you can code are Terrorism in Europe and Terrorism in the Middle East, which, in this codebook, are both considered types of terrorism. That is, they are 'children' of the 'parent' terrorism. If you choose to export 0 parents, you export the codes exactly as they are coded: Terrorism, Terrorism in Europe, and Terrorism in the Middle East. If you instead choose to export 1 parent, AmCAT adds an additional variable to the dataset which aggregates both Terrorism in Europe and Terrorism in the Middle East to the level of their parent, meaning that the value for both of these codes is 'Terrorism'. This saves the trouble of recoding variables into aggregate categories!
Click Submit, and in the next screen Download results to download the codings in the chosen file format.