StoryEngine Production Guide How to create an evidence narrative from a raw transcript

Create a space for reflection

Throughout the editing process, prioritize the participant/interviewee experience. Keep it easy and simple for them and focus on creating opportunities for reflection.

The editing process is an extension of the interview — you are still working to describe and document the participant’s reality and perspectives. Editors help interviewees articulate and clarify what they want to say.

Produce a good-enough Q&A-format source text

Edit for flow and clarity in order to produce a good-enough Q&A-format source text. This is your particpant’s evidence narrative, which they will then use for other purposes (see below).

Don’t be afraid to make big changes as you edit! Your role is more of a ghostwriter than an editor. Stay true to the interviewee’s voice and ideas, but feel free to move chunks of text around, suggest additions, and cut and change words.

Use the comments feature in Google Docs to communicate your changes with participants. Flag any sections where you interpreted or may have changed the meaning, need clarifications, or have any other substantive concerns. Focus on:

  • Transcription errors like typos, homonyms, inaudible sections, and ideas/thoughts broken between sentences or paragraphs. If the content does not make any sense, add text based on your understanding of what the participant is trying to say and flag that section for close review via a comment (Example: “It was hard to hear what you said here, so I added this text. Is this correct? If not, please feel free to change it!). Checking the original audio should be a last resort because it it time consuming.  
  • Removing overused, qualifying, or superlative words such as “actually” / “really” / “very” / “pretty” / “little” / “just” / “that” / “basically”. When doing the final proofread, use your software’s find command to locate these words and delete any that are unnecessary
  • Removing repeated use of “I feel” / “I think” / “for me” / “personally” — sometimes OK to leave it in, but usually stronger to remove: “I think a healthy internet is, for me, one that…” versus “A healthy internet is one that…”
  • Editing terms people use in conversation that are not used in writing, such as using the word “like” as a substitute for “said”. For example, if the transcript reads “she was like yeah let’s do it” the editor could replace that with “she agreed” or “I was able to secure agreement from my colleagues”
  • Places in text where people are pretending to quote another person as a way to express their own ideas or experiences. This is another technique people use in conversation. Edit text to communicate these thoughts into the participant’s own words, without quotes. (This is related to the point above, using “like”.)
  • Reduce shifting tense or perspective — Especially between 1st, 2nd, and 3rd person (them/you) or past/present/future. For example: “That gives you a certain amount of credibility when you’re going and give these talks.” could be edited to read “I build my credibility by speaking at conferences.”
  • Unclear antecedents for words like “people” or “project” or “this”. Editors may need insert the specific thing the participant is referring to so that the meaning is clear and the reader does not get lost.
  • Make the actor/decider clear. Edit out use of the passive voice — these are often places that need clarification. For example, “it was decided” does not tell us much, but “the project team lead decided that…” tells us that there was a collaborative process and that decision-making was via consensus
  • Copy edit for consistent style. Based on your style guide and word list, ensure consistency in the document and across story collections. Style examples include:

Em-dashes, replace hyphens and ensure spacing — like this — before and after
internet, lower-case “i”
nonprofit, without a hyphen
wifi, lowercase
numbers, write out all numbers up to nine as words
bullet lists, ensure consistency within the list, full stops only when lists contain more than one sentence, no semicolons after each list item

Depending on the team and how you are working, the following may also be important:

  • Drafting or editing a short bio and/or context statement to go at the beginning of the story. We recommend 80-word bios and 30-word boilerplate context language, such as “This story is part of an evaluation of our Fellowship Program…”.
  • Adding helpful links up front and throughout the text, especially when the participant refers to something the reader may not know much about.
  • Assigning document style sheets — Style sheets are your friend because they structure your text, make updating/changing formatting super easy, and allow you to export stories to different applications. For example, on this website we use H3 for questions.

What’s next for this story?

Once the story is approved. Here are some of the ways it can be used:

  1. Sensemaking — From traditional qualitative data analysis to group sensemaking via collaborative online annotation, a workshop, etc.
  2. Creating a story bank — Full stories can be shared using a Creative Commons open license, so that colleagues and allies can re-use and remix content to advance their work. The StoryEngine website is an example of this.
  3. Generating learning, advocacy, and communications products — Full stories can be further edited, chunked up,. combined, and packaged to create learning, advocacy, and communications materials. They are especially useful when you need to provide examples to back up an argument or rationale, or to illustrate findings, complex dynamics, and theories of change. Examples: Quotes for reporting, presentations, and proposals. Digital media, such as blogs, podcasts, videos, infographics, or social posts. Printed collateral, such as posters, books, or merch.