Antconc Tutorial
A. Generating Frequency table
1. Open AntConc.
2. Load your corpus via File > Open Corpus (if not already loaded).
3. Click on the Word List tab.
4. Click Start to generate the frequency list.
5. View the results — it shows words ranked by frequency.
6. Use Sort by Frequency or Alphabetically as needed.
7. Export if desired via File > Save Output.
Note: AntConc does not automatically filter content words (nouns, verbs, adjectives, adverbs), but you
can do it manually by:
Using a stoplist:
1. Go to Tool Preferences > Word List.
2. Load a stoplist file (a text file of function words to exclude like "the," "and," "is").
3. Click Apply and then re-run the Word List.
Manual filtering:
1. After generating the word list, manually scan for content words.
2. You can also export the list and filter it in Excel or another tool.
B. Generating Condordance table
1. Open AntConc.
2. Load your corpus:
Go to File > Open File(s) or File > Open Dir
Select your text files (plain .txt format)
3. Click the “KWIC “ or “Concordance” tab (top left).
4. Enter a search word in the search bar at the top (e.g., said, marriage, gentleman).
5. Click “Start” (or press Enter).
You’ll now see the concordance lines — the keyword centered in each line of text, with
surrounding context.
6. Adjust context width (optional):
Go to Tool Preferences > Concordance
You can change the number of words shown to the left/right of the keyword.
7. Export results (optional):
File > Save Output to export the table (e.g., as .txt or .csv)
C. Generating Collocations table
1. Same instructions from the previous steps for 1 & 2
2. Click the "Collocates" tab
3. Enter your search word (e.g., man, lady, marriage) in the search bar at the top
4. Set collocation parameters (optional):
Go to Tool Preferences > Collocates
Adjust the search window size (e.g., 5L–5R means 5 words to the Left and Right of your
keyword)
You can choose how the collocates are sorted (e.g., frequency, MI score, etc.)
5. Click “Start”
AntConc will list all the collocates around your keyword, with stats like frequency and
Mutual Information (MI) score.
6. Review the results:
You’ll see columns such as:
Collocate word
Frequencies
Position (L/R)
Statistical score (like MI or t-score)
7. Export (optional):
Click File > Save Output to download your results
Notes:
Use function words like "is", "the" with caution — they appear frequently but are not meaningful
collocates
MI score is best for identifying strong associations rather than just frequent co-occurrence.
Depending on your Antconc versions, you might see “Likelihood” or “Effect”
In newer versions:
MI (Mutual Information) and t-score are replaced or supplemented by:
Log-likelihood = useful for statistical significance. It shows how significant the collocation is.
Effect size = useful for practical strength. It shows how strong the association is, regardless of size.
