Skipper Autocode leverages AI to automatically detect themes in your response data and code them accordingly, or to bulk-add codebooks to individual questions in your survey. 

Enabling Skipper Autocode

On the Survey Editor page, when you first add an Open-Ended Question, Form Question, or Open-Ended answer option, a message will appear at the bottom of the page.

• Keep the box checked to leverage Skipper Autocode during your survey analysis.

If you did not enable Autocode before launching your survey and wish to use it, open the stats page of your survey:

1. Navigate to the Coding tab in the left sidebar menu.
2. Click Unlock Autocoding. You will be redirected to the checkout page to purchase this feature.
3.  After completing the purchase, Verbatim Translation will be available for use on the Stats page.

Using Skipper Autocode

To analyze and create codebooks using Skipper autocode, open the Coding pane in the left menu.

1. Click the drop-down menu beneath Autocoding and select the desired question.
2. Click the autocode drop-down menu and select autocode.
3. Click the Run autocode button and Skipper will analyze all verbatim responses, creating a codebook with relevant codes, and assign those codes to responses accordingly.

Note: Skipper autocode can review open-ended responses written in any language, and the corresponding codes displayed in the sidebar will be in English. 

Autocode Existing Codebooks

Allow Skipper to analyze responses and assign codes to each response. Use an already existing codebook or manually create a codebook:

1. Click Add a new codebook.
2. Type a title into the New Codebook text box.
3. Click OK.
4. Click the drop-down menu beneath Autocoding and select the desired question.
5. Click the autocode drop-down menu and select Apply Codebook.
6. Select the desired codebook from the Codebook drop-down menu.
7. Click Apply autocode to apply the codes. 
    

Note: Filtering to a particular question in the Autocoding Pane will not necessarily reflect that question in the main Open Ends report—you must use the main Open End report drop-down to change the page view.

Guided Autocode

Use Guided autocode to expand and refine an existing category. This process also allows you to create a codebook with a specific name for the category you wish to explore for a particular question. Use an already existing codebook or manually create a codebook:

1. Click the drop-down menu beneath Autocoding and select the desired question.
2. Click the autocode drop-down menu and select Guided autocode.
3. Select the desired codebook from the third drop-down menu.
4. Click the Run Guided autocode to confirm the command. Skipper will compare the existing codebook structure to the responses in the selected question, and create a brand new codebook.

Note: The new codebook may contain some of the same codes, but will also contain other similar themes detected in the comparison.

Editing Existing Codebooks and Codes

As always, if you dislike the way Skipper handled something, you can edit or delete items as needed.

Codebooks:

1. Click the settings wheel to the right of a codebook to:

• Rename the codebook
• Clone codebook
• Copy codebook structure
• Delete codebook and codes from all reports

   2. With a specific question selected from the Verbatim drop-down, additional options appear:

• Apply to Q#
• Run guided autocode on Q#

Codes:

1. Click the settings wheel to the right of a code to:

•  
  • Remove the code from the comments on the page
  • Remove code from all found comments
  • Rename the code
  • Assign to multiple comments
  • Delete the code itself from all reports

Sentiment Analysis

LLM-powered Sentiment Analysis enables Skipper to interpret and apply sentiment to unstructured data in open-ended responses. It can determine when sentiment analysis is appropriate, and gives users the ability to rerun and categorize the entire set of open-ended responses.

1. From the Results page, click the Verbatim icon at the top  to view all responses to the open ended questions included in your survey. The colored bands along the left side of each comment box indicate the sentiment related to the response.
2. Hover anywhere inside a response to see the corresponding emoticon, and click the assigned emoticon to view the spectrum of sentiments.
   • Green (smiling face) represents a positive response.
   • Orange (unsure face) represents a mixed response.
   • Red (frowning face) represents a negative response.
   • Blue (neutral face) represents a neutral response.
   • Gray (no mouth face) represents none.
3. Click the Coding icon in the lefthand navigation to view any codebooks you’ve created, as well as codes associated with each sentiment.
    
4. Click the Sentiments folder to view the sentiments as codes.
5. Click the gear icon on the right of the Sentiments folder and select Rerun Sentiment Analysis. This will reprocess the data, reassessing which questions warrant sentiment analysis, and redistributing sentiment codes. You can still manually adjust sentiments if needed.

Note:  To unlock the enhanced LLM-powered sentiment analysis, your account must have AI and Skipper enabled. Additionally, this feature is not applied retroactively to past surveys. If you'd like to apply the new LLM version to a previous survey, please contact our team.

Learn more about Sentiment Analysis here.

Autocode in Virtual Questions

Autocode is available in Virtual Questions as a Codebook preset that automatically generates and assigns codes for efficient correlation analysis.

1. On the Results page, click the Data Lab tab in the left menu.
2. Click the Virtual Question bar to add a VQ to your report, or drag it into place between questions.
3. Click the Optional Presets drop-down menu, and scroll to Codebooks.
4. Select the desired codebook and Generate Visualization.

Note: If a user renames a code, it will automatically update in Virtual Questions.

Autocode in Correlation Finder

1. On the Results page, click the Data Lab menu tab, and drag the Correlation Finder into place in your results.
2. Click the pencil icon to edit the name that appears in the Question Header.
3. Click the drop-down and select SigTest (t/z test) or Crosstab (chi²).
4. Click the Choose a banner drop-down to select a banner from the list of demographic traits, questions, codebooks, or Virtual Questions. Click the second banner drop-down to refine your variable.
   Note: The banner should be the independent variable that you are studying; the underlying factor that can lead to statistical significance. 
5. Click the Choose an attribute drop-down to select an attribute to check for correlations.
   Note: The attribute should be the dependent variable that you’re using to explore statistical significance.
6. Click Generate data to populate the chart. Once a Correlation Finder is added to your stats report you can export the chart and data table to PowerPoint or download a static image of the chart.

Note: If a user renames a code, it will automatically update in Correlation Finders.

Logic for Codes and Codebooks

1. To capture the frequency of use of a single code on a single question:
   • Q1 tag = 'codebook name: code name' is the same as tag Q1 ='codebook name: code name'
      
2. To capture the same as above but on multiple questions:
   • tag Q1,Q2 = 'codebook name: code name' is the same as tag Q1 = 'codebook name: code name' or tag Q2 = 'codebook name: code name'

     • The , works as or.

3. To express and:

   • tag Q1 = 'codebook name: code name' and tag Q2 = codebook name: code name'
     • Note: For single questions, the 'Q1' can go before or after tag, but for multiple questions, it can only come after tag.
4. The same structure applies to codebooks:
5.  
   •  
     •  
       • Codebook = 'name'
       • Codebook Q1 = 'name' is the same as Q1 codebook = 'name'
       • Codebook Q1,Q2 = 'name' is the same as codebook Q1 = 'name' OR codebook Q2 = 'name'
       • Q1 = 'name' AND Q2 = 'name' is the same as codebook Q1 = 'name' and codebook Q2 = 'name'

Exporting Verbatim Responses

There are two ways to download verbatim responses, along with their Codebook and Codes: By individual questions or for the entire survey.

To export by individual questions:

1. Scroll an open-ended question from your quant report.
2. Click the Export icon and a drop-down menu will appear.
3. Select the desired report:
   • Download all verbatim (CSV)
   • Download all verbatim with codes (CSV)

To export verbatims for the entire survey:

1. Open the Verbatim page of your stats page.
2. Click the Export icon in the left menu to open the Export Comments pane. From here you can can select any desired export.

Note: To export verbatim responses along with their codebook and codes, select Excel Coded Maps, Raw CSV, and Excel RAW.