Professional Documents
Culture Documents
Your Email Address Babbletype Translation Test 20200811
Your Email Address Babbletype Translation Test 20200811
This document is intended for translators. If this is not the test you were expecting, go back to the
BABBLETYPE TEST email and download the file containing the correct document.
Your goal is to fill out all of the information we require directly inside this document and then to return
the completed version to us. Be sure to read all of the instructions before proceeding. Don’t edit or
delete anything. All the content you see in this document now should still be there in the same form
when you submit it (with all questions answered and the translation test completed).
This document consists of five sections: information about the work we have available, instructions
about setting up your audio player, a questionnaire, the transcription test, and then next steps (how to
submit, how to review your test results, etc.).
Languages we support
Babbletype supports only a small number of languages which are commonly required in the market
research industry. These are the only languages for which we ever have work available:
French
German
Italian
Japanese
Korean
Mandarin
Portuguese
Spanish
One hundred percent of our language orders are from the source language into English (never from
English into another language).
1
Babbletype translation test v5 20200811
Nature of the work available
Babbletype specializes in audio-recording-to-text services, and all of the translation work we have
available is from an original audio recording in one of the languages above directly into a written
English-language transcript. We perform no text-to-text translation.
The majority of the translation work we perform is technical in nature, most often interviews dealing
with medical or pharmaceutical topics. Knowledge of, and willingness to do the necessary research on,
technical, pharmaceutical and medical terms is critical.
Much of the translation work Babbletype has available involves more sophisticated formats, such as
capture of responses against pre-defined question lists in tables. These formats take time to learn and
become fully familiar with.
Babbletype maintains very high standards in terms of the quality of the work we produce. This includes
the accuracy and completeness of capture, and the quality and accuracy of the English-language
transcripts we produce. We limit the people we work with to those able to meet and maintain those
standards.
Unlike English-language transcription work (which is we produce in large volumes daily), translation
work from any specific language is not continuously available. Demand varies by language, but rarely
exceeds a few hundred hours in any given language annually. In some cases, months can pass between
orders in a language, and the timing of when orders are received is not predictable. For this reason,
while Babbletype can contribute to your translation earning plans, we should not be your only client.
Note that Babbletype actively seeks out contractors able to both translate and to transcribe directly
from English (at the skill levels we require), since we can keep people with both skill sets busy much
more regularly.
Given the limited volume we work with and the sophisticated nature of our work requirements,
Babbletype’s approach is to work with a very small number of translators per language, who we can get
to know well, and who can get to know us well.
Get Express Scribe and Hot Keys working before taking the test
Efficient transcription requires you to work in two different programs at the same time: a word
processing program and an audio player. It can be incredibly time-consuming to constantly jump back-
and-forth between the two programs (to the audio player to control playback and to the word
processing document to enter content). To avoid all that wasted effort, transcription audio players have
a feature called Hot Keys which allow you to control the audio player while continuing to work in the
word processing program.
One important element of the transcription test, for both you and us, is to obtain an accurate
understanding of how long the file takes you to transcribe. You can’t do that without setting up and
learning how to use Hot Keys. So do that first, before starting on the test.
2
Babbletype translation test v5 20200811
Babbletype recommends Express Scribe by NCH software for this purpose. To get a copy, google, “How
do I install the free version of Express Scribe?” and go to the relevant page on nch.com.au. Download
and install the program.
Next, set up Hot Keys on your computer and practice using them before starting the test. You want to be
able to control audio playback (Play, Pause, Back, Forward) using the Hot Keys while typing in a separate
word processing document. Go to YouTube and search for “express scribe hot keys” to find information
on how to do that.
3
Babbletype translation test v5 20200811
Questionnaire
Babbletype looks for more than translation skill in the contractors we work with. The purpose of this
section is to ensure that you and Babbletype are a good fit for each other in all of those other attributes.
Please answer all questions, even if the answer is only “None” or “N/A”, etc.
4
Babbletype translation test v5 20200811
experience do you have, over
what period of time, in what kinds
of topics?
Do you have any direct experience
in translating from an audio
recording direct into a written
transcript (i.e., translation-
transcription)?
Let us know the level of
experience you have dealing with
technical and medical terms and
how you obtained that
experience. If none, just write,
“None.”
Your English language-related skills
Let us know about any experience
you have in writing, either in your
business or personal life. If none,
just write, “None.”
How many words would you say
you write in the course of an
average week for any purpose?
Let us know about the most
challenging writing project you
have ever worked on. If you don’t
recall, just write “Don’t recall.”
Let us know about any experience
you have in editing. If none, just
write, “None.”
Please provide a list of books
you’ve read over the last month. If
you don’t remember, just write
“Don’t remember”.
Your equipment and accounts
What kind of computer do you
have (e.g., Windows, Macintosh,
etc.)?
Let us know your computer’s
model, age, processor, memory
and storage capacity. If you don’t
know, answer “Don’t know”.
What kind of internet connection
do you have and with what
5
Babbletype translation test v5 20200811
provider? If you don’t know,
answer “Don’t know”.
What is the current upload,
download, and ping speed of your
internet connection? If you don’t
know, answer “Don’t know”.
Are you able to stay connected to
the Internet continuously while
working?
Do you have, or are you able to
obtain, accounts with Google
(Gmail), Skype, and PayPal? In no,
please explain.
Your life situation and goals
Why are you applying for
translation work, and why now?
How did you hear about us? If a
website, which one? Please be
more specific than “the Internet”
or “Google search” or “YouTube”
if possible. Copy and paste the
address of a specific website if
possible.
What kind of earning power goals
do you have in working with
Babbletype, and how quickly do
you want or need to achieve
them?
Your work methods
All of the work Babbletype has
available is independent
contracting work only. Do you
understand the difference
between independent contracting
and full- or part-time
employment, and accept this
requirement?
Babbletype only works directly
with the actual independent
contractors performing our work.
We expect you to perform the
work we assign to you directly,
and to not sub-contract the work
to anyone else to complete. Do
6
Babbletype translation test v5 20200811
you accept this requirement?
Babbletype requires all work to be
performed only in online Google
word processing documents that
we supply. You may not work in
other tools (e.g., MS Word,
Express Scribe, etc., and then copy
and paste content into those
online documents). Do you accept
this requirement?
Babbletype requires all work to be
performed manually (i.e., by
listening and typing), and does not
allow the use of any machine
transcription, translation or similar
tools. We immediately stop
working with anyone we suspect
is using such tools. Do you accept
this requirement?
Your Express Scribe setup
Did you succeed in configuring
Express Scribe’s Hot Keys to work
as expected?
7
Babbletype translation test v5 20200811
Transcription test
Time and earning power
This test requires you to translate and transcribe a recording approximately 15 minutes in length. For
your own benefit, it’s important that you keep track of how long it takes for you to complete it.
If you are inexperienced with this work, it would not be unusual to spend quite a bit of time absorbing
instructions, dealing with Express Scribe or other tools. You can ignore that time investment as a one-
time thing. However, the time it takes you to actually translate and transcribe is another matter, and
that’s the time you should keep track of.
The reason for this is that it will give you an indication of your potential earning power doing this type of
translation-transcription work. This work pays not by your time spent, but by the number of recorded
minutes you translate-transcribe, and so how quickly you can do that provides an indication of how
much you have the potential to earn for the time you work.
For example, this work pays approximately $1.30 to $1.60 per recorded minute (depending on language
and format), which means that 15 minutes of audio translation-transcription is worth between $19.50
and $24. Use this range, and the time you take to do the work, to determine if you can perform this
work quickly enough, and earn enough from it to suit your needs.
Your goal in this test is to create an English-language written transcript from an original foreign-language
audio recording. Listen to the original recording in your source language, translate the content in your
head, then write down the full content of what was said in English. The document you return to us
should always be in English only.
To obtain the audio file to transcribe, click on the link for your language in the table below. When the
file opens online, download it and load it into Express Scribe (or similar transcription audio player).
While transcribing, work directly in this word processing file.
When you begin working, note the time that you started. When you finish, note the final time. Make
note of any time frames when you pause to do something else. After you completely finish transcribing
and checking your work, write down the total number of minutes the work took you to complete in the
time to complete field (below). Try to be as accurate as possible.
8
Babbletype translation test v5 20200811
What to capture
Capture everything everyone says throughout the course of the recording, as exactly as possible to the
way it was spoken. Do not remove, skip, change, otherwise edit what people say. Don’t invent or add
wording that isn’t hearable in the recording. Your translation does not need to be word-for-word precise
against the original recording, but all informational elements contained in the recording should also be
reflected in the transcript. Again, don’t skip or summarize.
What not to capture. The only thing you should not capture are sounds people use while talking, while
they are thinking through their answer, and which have no explicit meaning, such as “um”, “uh”, “ah”,
etc. On the other hand, if a sound has some meaning in it, it should be captured. For example, if a
speaker says, “Ah, I was thinking of buying one too,” then the “Ah” should be excluded. However, if a
speaker says, “Ah! I was thinking of buying one too!” then the “Ah!” has meaning (as an exclamation)
and should be kept.
Paragraph use. As the conversation proceeds, make a new paragraph every time the speaker changes.
While an individual is speaking, use just one paragraph regardless of the length of the speech. When
switching to the next speaker, add one blank line between the previous speaker and the next (as shown
in the example below).
Speaker naming. Identify the interviewer as “Interviewer”. Identify all other speakers by their first
names whenever possible. When not possible to identify the person talking, use “Male respondent” or
“Female respondent”. If it’s not possible to identify the gender of the speaker, use “Respondent”. If
more than one person answers, use “Multiple respondents”.
Inaudible and unclear content. If something is said which you cannot make out at all, enter a question
mark surrounded by parentheses at that location in the text like this: (?). If something is said which you
can make out, but you are not sure that you have heard it correctly, surround the uncertain text with
parentheses, like this: “This is the part I am sure of, (and this is the part I am not sure of).” Note that in
our review, items which you have marked as unclear but that are in fact reasonably clear will be marked
as errors. Some unclears are normal, but large numbers of them.
Non-verbal content. If something occurs which is not made up of words, but which is material to the
events you are transcribing, capture them with a descriptive word of two in parentheses, for example:
(laughter).
English composition. Use correct rules of English composition throughout, including spelling,
punctuation and basic sentence structure.
Interviewer: OK, that’s interesting. Would you prefer if the service provider handled that for
you, or would it be OK if you had to do it yourself?
Joachim: Well, I would prefer it if the service provider could do that for me.
9
Babbletype translation test v5 20200811
Kihoko: I agree! I wouldn’t want to have to do that on my own.
Sarah: Same.
Transcript end
When you finish, go to the Next Steps section on the following page.
10
Babbletype translation test v5 20200811
Next steps
This section explains how to submit your application, and the process we use to review it.
File naming. As received, this document is named, “[Your email address] Babbletype English-English
transcription test {date}”. Replace the portion “[Your email address]” with the email address you will use
for submitting the test later. For example, if your email address is [email protected], then the new
filename would be, “[email protected] Babbletype English-English transcription test {date}”.
File submission. When your file is fully completed, create a new email, address the email to
[email protected], write TEST SUBMISSION as the subject line, attach the test file, and
send. You’ll receive an automated reply as soon as we receive it.
Time to review. We review all applications seriously and that takes some time. It can take at least a
week for us to fully complete reviewing translation applications and get back to you. Please be patient.
We do reply to everyone.
How we decide. Babbletype is extremely focused on fit. We only work with people who are a very close
fit to the attributes we are looking for. We look at the totality of your application to try to make this
judgment. Your results on the transcription test section help us understand skill level, and your answers
on the questionnaire section help us understand other important attributes. Final comments on both
can be found in the reviewer notes at the bottom of this document. Bear in mind that not being
accepted simply means that we don’t see the complete fit we are looking for.
How to read our comments. With regard to the transcription test, we use inserted word processing
comments to document any issues we see in your test. These comments are viewable in Microsoft Word
or in Google Docs; just ensure that comments are turned on so you can see them. To understand the
meaning of our markup codes, refer to the key below. Note that markup is not intended to be
comprehensive and that markup stops when a decision is reached.
Can you apply again if you are not accepted? Sorry, no. Babbletype does not allow multiple submission
attempts.
How to contact us. Feel free to contact us if you have questions, but only by email please. Write to
[email protected] and enter APPLICANT QUESTION in the subject line. Thanks for
considering Babbletype.
11
Babbletype translation test v5 20200811
Test review key
Our comments on your test include codes related to the types of errors we identify. Each code consists
of a letter defining the type of error, and a number defining the severity of the error. Refer to the table
below to understand the meaning of the codes we use.
Composition errors
SP, GR, P, T, Spelling, grammar, punctuation or typing errors
CAP 1: Typical error.
SS Sentence structure error.
1: Typical error involving one sentence that should be two, or vice versa.
2: Serious run-on sentences (one sentence that should be three or more).
5: Extreme run-on sentences involving large swaths of text.
Accuracy errors
M Mishear, incorrect content captured.
2: Minor mishear or wrong word choice not seriously impacting meaning.
10: Major mishear that changes meaning.
I Invented content not impacting meaning.
1: Typical error.
A Attribution error, wrong person attributed as speaking.
1: Typical error.
N Notation error, uncertain content which should have been marked as such.
1: Typical error.
CUC Clear unclear error, content marked as unclear which is in fact clear.
1: Typical error.
Completeness errors
S Skipped or missing content.
1: Minor, not impacting meaning.
10: Major, multiple words or a critical word.
NV Non-verbal content wrong or missing.
1: Minor errors which don’t significantly affect meaning.
Instruction errors
IN, F, C Instructions not followed. IN = general, F = format, C = capture.
(Examples: capturing words such as “uh” or “um”; failing to bold speaker attributions;
incorrectly notating laughter; no following font and spacing guidelines from the
example)
1: Typical error.
12
Babbletype translation test v5 20200811
Unclear count and error scores
Unclear count Composition error score Capture error score
Reviewer notes
Native English Review How understandable is the translation? How well can the conversation be
tracked from start to end? What impact do composition errors have on the
readability of the transcript?
[Reviewer initials/date]
Translator Review How would you describe the quality of the translation? What are the
strengths and the weaknesses of the translation?
[Reviewer initials/date]
13
Babbletype translation test v5 20200811