Examples and Workflows

Examples and Workflows

Workflow 1: Basic Qualitative Analysis

Goal: Import interview transcripts, find topics, and visualize themes.

1. Import Data:

   crisp --source ./interviews --out ./corpus

2. Analyze (Topics & Sentiment):

   crisp --inp ./corpus --topics --sentiment --out ./corpus_analyzed

3. Visualize:

   crispviz --inp ./corpus_analyzed --wordcloud --ldavis --out ./viz

Workflow 2: Mixed Methods (Triangulation)

Goal: Analyze survey comments in the context of satisfaction scores.

1. Import Survey Data (CSV):

   # "comments" is the free-text column
   crisp --source ./surveys.csv --unstructured "comments" --out ./survey_corpus

2. Cluster by Satisfaction Scores:

   # Cluster based on numeric columns
   crisp --inp ./survey_corpus --kmeans --include "satisfaction,age" --out ./clustered

3. Analyze Text within Clusters:

   # Filter for Cluster 0 and find topics
   crisp --inp ./clustered --filters cluster=0 --topics --out ./cluster0_topics

Workflow 3: Temporal Analysis

Goal: See how topics change over time in a longitudinal study.

1. Import & Link Time:

   crisp --source ./field_notes --out ./corpus
   # Link docs to weekly periods
   crispt --inp ./corpus --temporal-link "sequence:timestamp:W" --out ./temporal_corpus

2. Analyze Trends:

   # Get top 5 topics per week
   crispt --inp ./temporal_corpus --temporal-topics W:5

Workflow 4: Semantic Search

Goal: Find relevant literature/docs for a specific query.

1. Search:

   crispt --inp ./corpus --semantic "patient mistrust of vaccines" --num 20 --out ./search_results

2. Export Metadata:

   crispt --inp ./search_results --metadata-df --out ./search_results_df