| Q. |
How do you transcribe something that is not obligatory but sounds better if is included? e.g. The child says "Looks like he fell down.", instead of saying "It looks like he fell down." |
| A. |
You could mark the initial It as an omission (e.g.
"*It looks like he fell down.") but people often omit
"it" in everyday usage. Omissions are indications of errors and this is not an
error. So it should not be transcribed as an omission. Just transcribe what the
child said (e.g. "Looks like he fell down.").
|
| Q. |
How do you transcribe over generalization errors? e.g. The child says "We goed to the park." or "We wented to the park." or "That's the bestest one." |
| A. |
These words should be coded as errors. The SALT Reference Database uses only two
types of error codes: [EW:______] to mark word-level errors and [EU] to mark
utterance-level errors. (You have the option, of course, of developing a more
elaborate coding scheme.) We would transcribe the first example as "We goed|go[EW:went] to the park.", the second example as
"We wented|went[EW:went] to the park." and the third
example as "That/'s the bestest|best[EW:best] one."
There are two things to notice in these examples. First we do not slash the
bound morpheme (e.g. go/ed, went/ed, best/est) since we do not want to give the
child credit for two morphemes due to the over generalization error. Second the
vertical bar followed by the root word (e.g. |go,
|went, |best) is included to
identify the root word so that "goed" is counted as an instance of the word
"go", "wented" as an instance of the word "went", and "bestest" as an instance
of the word "best". This is important when counting the number of different
words and when listing the words in the word root tables.
|
| Q. |
How do you mark errors for which there isn't a transcription convention, e.g. pronoun reference errors, pronunciation errors, inappropriate responses, and topic initiation/continuation problems? |
| A. |
Although there isn't a transcription convention for these errors, you may want
to mark them using codes so that you can track their occurrences over time. You
would do this by developing a set of codes that you insert in the transcript at
the position of these errors. Be consistent. Use the same codes to mark the same
errors across all your transcripts.
|
| Q. |
How do you describe "errors" in standard form due to ethnic dialect? |
| A. |
Differences due to ethnic dialect are not errors and should not be coded as
errors. Instead, develop a coding scheme for marking dialectal differences and
attach a code to the word or utterance. The SALT Reference Database uses the
code [D:___] to represent phonological or dialectical differences. Consider the
example where the child, who speaks Black Vernacular, says "He go to the
store.". This would be transcribe as "He go[D:goes] to the
store.".
|
| Q. |
How do you transcribe word-position errors? e.g. The speaker says "I see the dog big." |
| A. |
This should be coded as an error. The SALT Reference Database uses only two
types of error codes: [EW:______] to mark word-level errors and [EU] to mark
utterance-level errors. (You have the option, of course, of developing a more
elaborate coding scheme.) Since there isn't an error with the words, we would
use the utterance-level error code as in "I see the dog big
[EU]."
|
