The Walking Around Project

Transcription Guidelines for PILOT #1

(revised July 26, 2011; for more information, contact Karla Batres)

I. Working with the recordings

NOTE: I recommend using HEADPHONES when listening to these recordings!

0. Download and install Audacity (freeware): http://audacity.sourceforge.net/

1. Open one file (.WAV) in audacity:

Audacity > File > Open…

2. Zoom in seven times so that you can focus on a short section at a time:

Audacity > View > Zoom in …. (7 times)

3. Go to the beginning of the conversation, by moving the cursor on the blue bar of
the Audacity window all the way to the left edge of the window (see small circle
in Audacity screenshot below):

4. Find the beginning of speech – you can tell if there is speech or silence from more
the waveform window:

1
o Silence is a straight line in the middle of the window (no action)
o Speech (or noise) shows us as various vertical lines extending up and
down of the horizontal line in the middle of the window (see the larger
circle in the above Audacity screenshot)

o Click just before the beginning of a speech portion and highlight a few
seconds but moving the cursor to the right and letting go when you’ve
highlighted enough speech (aim for about 4 seconds)

o Click the green “play” button (in the top left corner of the Audacity
window), to listen to the highlighted section of the recording, and
transcribe what you hear according to the guidelines below.

o (You can stop playing the recording by clicking on the square orange
“stop” button, three buttons to the right of the “play” button.)

o To find out the time point of the beginning of an utterance, look at the
very bottom of the Audacity window: there is a horizontal bar with
information. You will see “Selection: …”, which indicates the time point
from the start to the end of your selection. In parentheses, you will see the
duration of the selected speech portion (see circle in screenshot below):

o If you need to know the exact beginning of this speech portion (if this is a
new speaker, overlapped speech or unintelligible speech) write down the
exact TIME STAMP in this formation: m:ss.SSS (m= minute, s=second,
S=millisecond),
e.g. <0:07.482>

2
 o You can also simply click at a point for which you need the time stamp. At
the bottom of the Audacity window, you will see the time stamp for the
location of your “cursor”.

o Highlight your next section and proceed in this manner…

NOTE: please do not transcribe any directions you might hear me say at the beginning of
the recordings, such as “Did you hear that message?” etc. Thanks!

II. Transcription Guidelines:

Speech is transcribed and annotated with time and speaker information. The structure of
transcript will look like this:

<time> <<speaker info>>

transcription of this speaker’s entire utterance
<time> <<speaker info>>
transcription of this speaker’s entire utterance etc.

1. SPEAKER:
We will use <<speaker1>> and <<speaker 2>> for each conversation
Direction-giver (in lab) = speaker1
Direction-follower (walking around on cell phone) = speaker2

2. TIME:
o Every time the speakers switch, we will include a new timestamp; Time is
indicated in this formation:
m:ss.SSS (m= minute, s=second, S=millisecond),
e.g. <25:09.072>
a. We will transcribe the entire utterance of a speaker – we consider an
utterance finished when the speaker stops speaking or makes a long pause
(a pause of more than 1 second), or the other person starts speaking.

3. OVERLAPPED SPEECH:
Overlapped speech (when both speakers talk at the same time) is marked with an
ASTERISK “*” at the beginning and at the end of each speaker’s overlapped
utterance.

For example:

<0:01.362> <<speaker 1>>

hello, how *are you*?
<0:02.118> <<speaker 2>>
*I am fine* and you?

3
  This transcription example indicates that “are you” (from speaker 1)
and “I am fine” (from speaker 2) occurred at the same time.

Use double asterisks (**) for additional overlapped speech within the same
utterances.

4. CAPITALIZATION:
We will capitalize proper names, but will NOT capitalize the beginning of
sentences.

Examples:
a. he took the car on Saturday.
b. Jane was walking along Walnut Street when I met her.

5. PUNCTUATION:
We will not make use of punctuation, unless it is necessary.

o We will NOT use periods (for punctuation).

o Commas "," are used to separate clauses (esp. when it would be confusing to
read this part of the transcript without a comma), and for list intonations (e.g.
go to the top of the stairs, then turn left, then take two steps, then turn right)
o Question marks "?" are used to indicate questions with a clear rising
intonation (voice going up)

6. COLLOQUIALISMS:
Transcribe what people say; don’t correct their grammar if you think it would be
better to say it differently!

7. PARTIAL words:
Partial words are indicated with a dash (without any spacing between the dash and
the word):

a. absolu-
b. -tion

8. NOISES:
Sound phenomena such as laughs or coughs, are indicated by means of special
brackets {}, as follows:

a. {laugh} c. {sneeze}
b. {cough} d. {yawn}

4
9. INTERJECTIONS:
English interjections have the following standardized spellings:

o eee o mhm o whoa

o ew o oh o whew
o jeeze o okay o yeah
o huh o uh-huh
o hm o uh-oh

10. FILLERS:
English “hesitation sounds” have the following standardized spellings:

o er o um
o erm o uh

11. (Short) PAUSES:

Pauses/hesitations that are shorter than one (1) second will be indicated with dots in
parentheses, as follows:
o One dot (.) for a short pause;
o Two dots (..) for longer pauses,
o Three dots (...) for even longer pauses (but still less than 1 sec)

12. NO SPEECH:
Any amount of silences (over 1 s) should be indicated as “NS” (no speech).
Include timestamps for the beginning and end of this NS portion, as follows:

<time>
[[NS]]
<time>

13. UNINTELLIGIBLE SPEECH:

a. If you’re not entirely sure what a person said, indicate your best guess as
follows:

((text))

b. If you cannot make a guess at what was said, leave one single space
between the parentheses, isolated from the rest of the text. Include
timestamps for the beginning and end of this speech portion, as follows:

5
 <time>
(( ))
<time>

14. NUMERALS:
Write out all numerals. Only hyphenate numbers between twenty-one and ninety-
nine

o twenty-two o nineteen ninety-five

o seven thousand two o nineteen oh nine
hundred seventy-five

15. Individual LETTERS:

Individual letters that are pronounced as such should be written in caps and
preceded by a "~" symbol:

o I got an ~A on the test.

o his name is spelled ~S ~I ~M ~P ~S ~O ~N.

16. ABBREVIATIONS:
When abbreviations are used as part of a title, they can remain as abbreviations:

o Mr. Brown
o Dr. Jones

III. Naming & saving transcription files:

1. Use word for transcriptions.

2. Please save files as .doc (not .docx).

3. Save the file with your transcription according to the following convention:

“WAC transcript_Pilot1.X_ABC.doc”

 Replace “X” in Pilot1.X with the session number of the recording you are
transcribing. This information can be obtained by looking at the file name of the
recording, e.g. the following recording is from session 3 of Pilot study #1:

“WAC_Pilot#1.3_S1B&S3A_032311_teleoconf_all.wav”

6
  The file name of the transcript would therefore include “Pilot1.3” in the name,
as follows:

“WAC transcript_Pilot1.3_ABC.doc”

Finally, the last couple of letters should be your initials (two or three letters).
Using my name as an example (Katharina S. Schuhmann), my initials are “KSS”
and the file name of the transcript would be as follows:

“WAC transcript_Pilot1.3_KSS.doc”

4. If you are reviewing someone else’s transcription, please add another underscore
followed by “r” and your initials. If your name was “Adam Smith”, this would
look as follows:

“WAC transcript_Pilot1.3_KSS_rAS.doc”

5. The beginning of your transcription file should contain the following information
(keeping this format, but of course replacing names/dates etc.):

WAC transcript
Recording: “WAC_Pilot#1.3_S1B&S3A_032311_teleoconf_all.wav”

Transcribed by: Katharina Schuhmann (KSS)

Date: June 2, 2011
Checked by: Adam Smith (AS)
Date: June 3, 2011

….