How do I conduct a Sentiment Analysis?

How do I conduct a Sentiment Analysis?

Sentiment Analysis uses an AI language model to determine whether text data in your project has positive, negative, or neutral sentiments. Sentiment Analysis ultimately produces quotations defined by sentiment codes that you can use in later analysis.

You can choose to refine your analysis to documents, document groups, or quotations containing codes that you specify.

To use Sentiment Analysis on ATLAS.ti 23 Windows:

Select the Search & Code tab and from there Sentiment Analysis.
Toggle the documents or document groups you want to analyze OR select Codes at the top and toggle the codes you want to analyze.
Select whether you want Sentiment Analysis to code paragraphs or sentences- toggle the sentiments (Positive, Neutral, or Negative) you want to analyze and label each sentiment code if desired- review each of the resulting quotations to see if the sentiments match the text, then click the + sign on the proposed code to apply the coding

You can also use the buttons at the top to apply all proposed sentiment codes or all codes belonging to a particular sentiment. Additionally, you can also select desired quotations to apply custom codes if desired.

To use Sentiment Analysis on ATLAS.ti 23 Mac:

To open the tool, select Analysis > Sentiment Analysis from the main menu.
Toggle the documents or document groups you want to analyze OR select Codes at the top and toggle the codes you want to analyze.
Select whether you want Sentiment Analysis to code paragraphs or sentences- toggle the sentiments (Positive, Neutral, or Negative) you want to analyze and label each sentiment code if desired- review each of the resulting quotations to see if the sentiments match the text, then click the + sign on the proposed code to apply the coding