Preparing textual data for ATLAS.ti V.5
Factors which affect how you work with text in ATLAS.ti
The software allows the semi automatic and automatic (with) selection of text to certain units of text, either by double clicking or when using of the Autocode tool.
the word 1 (double click)
the sentence (via the colon : or any single or multi occurrence of the full stop/period . (2 double clicks - but this works variably)
the paragraph 3 double clicks
the whole document - 3 double clicks
So full stops, exclamation marks questions marks and colons are useful. So are paragraphs.
In structured data
PARAGRAPHS: There are two levels of Paragraph in textual data (rtf) within ATLAS.ti
Paragraphs separated by a "single hard return"
Paragraphs separated by more than one ("multiple") hard returns
For 'structured data' e.g. formal interview frameworks you will need to keep speaker identifier or question identifier attached to paragraph they belong to e.g. text or speech begins on the same line or the very next line as the identifier. This will allow you to do an autocode search for all Q1's across all data files and save the whole paragraph/paragraphs 1 following each hit (but using the 'multiple hard returns' option to stop the coding at the first double hard return or empty line) at a new code linked to all Question 1's.
Q1: Tell me about your history?
AL:...well it was like this...
Tell me about your history?
AL:...well it was like this
Or (e.g. in a focus group
CAROL: I don't know about anything
BOB: I know such a lot of things
NOT like this as the extra carriage return between the Q1: and the actual question will cause problems in Atlas.ti.
Tell me about your history?
In semi or unstructured data
It will probably not be desirable or possible to use shortcut methods of coding (as with the autocode tool mentioned above) so there is no need to adopt any special paragraph structure as above - simply break data up into paragraphs as you would normally. However, if all paragraphs are relatively short it would be useful to keep each Speaker ID and the speaker section belonging to it within the same Single hard return paragraph , and only have multiple hard returns between the end of each speaker section and the start of the next.
Save the file in one of two formats
Save in Word format (as long as you do not want to edit your data file once assigned to ATLAS.ti Hermeneutic Unit (project)
Save as Rich text Format (if you will want to be able to edit data once the data files (transcripts? field notes? ) have been assigned to ATLAS.ti. Note that it will be much safer to save your textual data files as you normally would, in word, for use in ATLAS.ti, just because you cannot edit them. See RECOMMENDATION below.
NOTE:ATLAS.ti has no Spell checking
Function: Use Edit tools in Word, as much as possible to clean up data, before assigning data files to ATLAS.ti
Making changes to the files in ATLAS.ti after the file has been assigned to an HU is a serious matter and you should always be careful to save the individual edits you make, and Save the Hu - before exiting the HU or you may have problems re-opening the HU and connecting to that file if you have not been careful about saving changes.
RECOMMENDATION: Get the files right in Word , and save in Word format. this means you cannot edit the individual data files in ATLASti, and therefore will not experience difficulties connected with unsaved edits made in ATLAS.ti.
Also refer to General transcription guidelines
Old data that was prepared for version 4.2, will have been saved as Text only with line breaks - these line breaks are identified as sentence ends when using autocode tool or double clicking to select sentences.
You may need to convert your old text to the new paragraph model
(In Edit mode with file open in main PD window - convert to new paragraph model)