10 Reports
- Coding reports
- Coding comparison
- Code frequencies
- View graph
- Code relations
- SQL statements (run SQL queries)
This report gives a list, or matrix, of coded data based on your selections. One or more codes need to be selected. If a category is selected all codes in that category are also selected. Multiple categories can be selected using the mouse and the ‘ctrl’ button.
The Search button (Button containing cogs) will present the results.
Coding report showing results of coded text
Coding results can be narrowed down by using the File selection, Case selection or Attribute selection buttons. Only codings within the selected files or cases will be presented. When case selections are used, file selections are ignored. Also when file selections are used, case selections are ignored.
If text is entered into the Search text field, only codings which contain the matching text will be presented. The search text also looks for matching text in memos for coded image areas and coded audio/video segments. The search text function must be used in combination with File selection, Case selection or Attribute selection.
** Important coded text, images, segments There is a tick box to select only the Important codings. This is useful for quickly finding those important exemplars that you may use in a written report.
The Attribute selection button opens a window where you can select attributes for files and cases, for example: only interview files for people aged > 60.
There is a problem with applying case attributes. This most commonly occurs with survey imports. The case attribute selection option will select the entire text file, rather than the sections assigned to cases.
One option for now is to use Reports > Database queries. In the upcoming 3.6 version onwards there is a default query that will do this.
In earlier versions do this below, create and run a SQL query like the following example:
SELECT case_text.caseid, attribute.name as 'Case attribute', attribute.value, case_text.pos0,
case_text.pos1, code_name.name, code_text.pos0, code_text.pos1, seltext
from
case_text join attribute on attribute.id=case_text.caseid join source on source.id = case_text.fid
join code_text on code_text.fid=source.id join code_name on code_name.cid = code_text.cid
where
code_name.name = 'code1' -- example code name
and attribute.name='District' -- example Case attribute name
and attribute.value='north' -- example case attribute value
and case_text.pos0 <= code_text.pos0
and case_text.pos1 >= code_text.pos1
f you are creating a report and selecting one of the matrix options, then the results pane will be split into two panes. The left pane shows the coded data in a list, and the right pane shows a matrix of rows and columns. The column types can be codes, categories, or top-level categories.
The panes can be widened or narrowed by dragging with the mouse on the bar that splits each section.
Reports can be exported to text, open document (ODT), html or csv files.
When cases are selected, another pane is displayed showing a table of cases (in rows) and categories or codes in columns
- Shortcut H. Pressing h will hide and unhide the top controls section of the window.
- Check box to add summary statistics to the report.
- Matrix options. So have codes/categories by file or by case. The matrix can be transposed, using the check box.
- Check box for coding in context for coded text segments. This shows the surrounding 250 characters of text to help put the coded section in context in the report. The coded text segment is shown in bold.
- Memo display. A drop down box gives options to include no memos, various memo option, or to display text file annotations for selected text files.
This option shows the overall similarities and differences between two coders across all text files. Select two coders and click the run comparisons button. The statistics created here are only for coded text, not coded image nor media files.
For each code:
Agreement % shows agreement for a combination of coded and non-coded text characters.
A and B % shows agreement for the only the coded text characters divided by the total characters in the text.
Not A and Not B % shows the total of the uncoded text divided by the total characters in the text.
Disagree % shows the percentage of all the coded and non-coded text that did not match between coders. It is the same as 100 – the Agree %.
Agree coded only % shows the percentage of all the dual coded text divided by single plus dual coded text. (In version 3.1 +)
Cohen’s Kappa is calculated based on the information in Wikipedia https://en.wikipedia.org/wiki/Cohen%27s_kappa
This option shows the overall similarities and differences between two coders in ONE selected file. Select one file, one code and two coders. Then click the run comparisons button. The statistics and comparisons for audio/video coded segments are in development. Overall agreement, disagreement and Cohens Kappa are calculated for text and images. An image file is displayed in a separate window showing the two coders selections in yellow or blue boxed areas. Mouse hover provides additional information.
This window shows all the codes and categories and the frequency of use for each coder. Code frequencies can be selected for specific files.
This dialog shows the relations between two or more codes. This can only be applied to text files. The relations are:
- Proximity – Two codes do not overlap. The distance in number of characters is shown.
- Overlap – Two codes partially overlap. The lowest and highest character positions of the combination are shown. The union of the overlapping section is shown in character positions.
- Inclusion – One code is included within another code. The lowest and highest character positions of the combination are shown. The union of the overlapping section is shown in character positions.
- Exact – Both codes match in their start and end positions. The lowest and highest character positions of the combination are shown. The union of the overlapping section is shown in character positions.
A csv file can be exported.
Click on a file to get a summary of that file. It describes media metadata, characters and word frequency. Attributes. It summarises the count of codes for that file. For text codings is summarises the average text length. For image codings it summarises the average pixel area. For audio/video codings it summarises the average segment length.
Click on a code to get a summary of that code. The coders who have used the code. The number of codings in text, image and audio/video media. For text codings is summarises the average text length and the most common words. For image codings it summarises the average pixel area. For audio/video codings it summarises the average segment length.
Charts are displayed in the default web browser. Various chart types and filter options are available.
Pie, Bar, Sunburst and Treemap charts can be created.
Heatmaps of files or cases against codes can be created. These are limited to 40 rows and columns for quicker rendering and better visibility.
Chart selections
Bar chart
Sunburst chart
Word cloud
A word cloud is available with different options for background colour, foreground colour or colour range, rectangle size, number of words (n-grams) from 1 to 4, and maximum number of words to include. The maximum words is determined by the most frequent to the least frequent and changing this option will change how crowded or sparse the word cloud appears.
There are a list of stop words that are used to exclude common words from the word cloud display (e.g. an, a, the). This can be overridden by creating your own stopwords.txt file with one word on each line. This file should be put inside the .qualcoder folder. This folder is located in the users home directory.
This window contains three panes. The top pane is where SQL statements are entered and the bottom pane contains the results of queries. The left pane contains tables and field names. Double-clicking on a field name adds it to the SQL statement. Results can be exported to a delimited file. If you are not familiar with SQL take care as you will be able to update and delete the data as well as select data. Note: Some Unicode symbols are not converted to plain text and are ignored.
There are several prepared statements listed:
- Case text
- Coded text using a case attribute selection (v3.6+)
- Codes fileid and coded text
- Get coding table - an implementation of an RQDA function that returns the code name, table, text length and text start and end positions.
- Coded text with each case
Most table fields are text. The following fields are integer: anid, avid, attrid, caseid,catid, cid, fid, id, imid, jid, pos0,pos1, x1, y1, width, height.
Right-click in the results table for filtering options.
Right click in the SQL window give you various options such as select all, copy, paste.
