Add Maple
← All help articles

Single Sentiment Score from Text

This workflow converts text responses into one sentiment category per response, then into a single numeric score (mean).
It is useful when you want one headline sentiment metric, but still keep full category detail.

1
Single sentiment score walkthrough

Where to find it

  1. Open a Text column.
  2. Click AI Coding / Tagging.
  3. Click Get started.
  4. Select I want to analyze the text into themes or other categories.

Step 1: set sentiment categories

For this workflow, use a manual code list so you control the sentiment scale.

  1. Click I have codes.
  2. Add sentiment labels (one per line), for example:
    • Very negative
    • Negative
    • Neutral
    • Positive
    • Very positive
  3. Click Continue.
  4. (Optional) refine code descriptions to improve consistency.

Tip: Use clear text labels first. You can map them to numbers later.

Step 2: apply one code per response

  1. In the review/apply step, select Apply One Code Per Record.
  2. Choose your model (for example, Fast for speed).
  3. Click Apply Codes.

This creates a new categorical sentiment column with one sentiment label per row.

Step 3: convert sentiment labels to numeric values

  1. Open the new sentiment column.
  2. Click Assign numbers.
  3. Map labels to an ordered scale (for example, Very negative = 1 ... Very positive = 5).
  4. Click Save.

Now the sentiment column can be summarized as a numeric score.

Step 4: view the single sentiment score

  1. Switch from distribution to stats view:
    • In many charts this appears as Box/Dot Plot | Bins (choose Box/Dot Plot), or
    • In some layouts it appears as Key Stats | Ranges (choose Key Stats).
  2. If shown, choose Mean in Box Plot | Mean.
  3. Turn on Show numbers on chart if you want exact values shown on the chart.

The mean value is your single sentiment score.

Compare groups or view overall

  • By group: Add a pivot (for example, by segment, region, or customer type) to compare average sentiment scores between groups.
  • Overall: Remove pivots to show one overall sentiment mean for the full selection.

Why this workflow is useful

  • You get one easy-to-track numeric metric.
  • You still keep the full label breakdown (Very negative to Very positive).
  • You can switch between detailed categorical view and summary score view in the same column.

Related guides