ENNI scores needed to use the information on this website are: Story grammar, Mean Length of Communicative Unit (MLCU), and Number of Different Words (NDW).

Story grammar scores are obtained by reading the transcript and filling out the scoring sheet. Instructions and scoring sheets for story grammar can be downloaded here: http://www.rehabmed.ualberta.ca/spa/enni/story_grammar.htm

Transcription of narrative samples takes place in CLAN, and CLAN also has the analysis programs. Using CLAN, language sample transcripts can be analyzed for sentence length (MLCU) and expressive vocabulary (NDW). How to use the CLAN commands for the narrative samples is described below.

How to calculate MLCU

Step 1. Add a morphological tier:

In order to run the MLU command, CLAN requires that a %MOR tier is added to the transcript. Add a morpheme tier using MOR command. Select all 6 narrative files for analysis using the File In button. The @ symbol will appear at the end of the command line indicating that the files were selected.

e.g. mor +t*CHI @

This command will generate 6 new files with a MOR tier added after each of the child’s utterances that contains part-of-speech tags for every word. These files will have the extension .mor.cex (e.g. narrativeA1.mor.cex, narativeA2.mor.cex, etc.). If there is no folder selected next to the output button, the file will be created in the input (working) folder (i.e. the same folder as the transcript being analyzed).

This command automatically parses the words in the transcript, excluding words marked as hesitations and fillers.

If you receive error messages, check to ensure that CLAN is looking for the library files (i.e. the library folder labeled English) in the right place.

Here is what a file with a %MOR tier looks like:

Top of Page

Step 2. Calculate MLCU in words on all 6 new files:

Since MLCU is calculated in words rather than morphemes, the command line should include the option -t%mor, which tells CLAN to disregard MOR tiers added to the original file: MLU +t*CHI -t%MOR. Make sure to run the analysis on the 6 new files (files with the extension mor.cex) generated by the MOR command.

e.g. mlu  +t*CHI  -t%mor  –s”[+ bch]” +u @

• Select the 6 files for analysis using the File In button in the command window (the @ symbol in the command line indicates that files were selected).
• The –s”[+ bch]” option tells CLAN to disregard lines in the transcript that end in [+ bch], i.e. utterances that were not relevant to the narrative.
• The +u option tells CLAN to run the MLU command on all six narratives and calculate the total MLU (if you forget to include +u, CLAN will generate a separate MLU value for each narrative).

The “ratio of morphemes over utterances” in the output window is the MLCU score.

Here is what the output window looks like.

Top of Page

How to calculate number of different words (NDW) and total number of words (TNW)

Step 1. Morphemization in ENNI

NDW refers to the number of unique or different words in the transcript (also called types in language sampling). TNW refers to the total number of words in the transcript (also called tokens). In order to calculate NDW and TNW, you will need to mark word boundaries for grammatical morphemes by placing hyphens between the stem and grammatical morpheme.

For this analysis, we recommend saving a copy of the original transcript as a separate file (e.g. filename_copy.cha) in order to have two versions of the transcript – with and without hyphens.

The grammatical morphemes plural -s, third person singular –s, possessive -‘s, present progressive –ing, and past tense –ed are preceded by hyphens in the transcripts as in the examples below. Note that spelling should be altered as well in the case of doubled or dropped letters, so that the program could recognize the stem of the word.

Examples:

"bounces " bounce-s

"bouncing " bounce-ing

"bounced" bounce-ed

"drops " drop-s

"dropping" drop-ing

"dropped " drop-ed

"falled" fall-ed”

"felled" fell-ed”

"pockets" pocket-s

"berries" berry-s

"centses" cent-ses”

"giraffe’s" giraffe-‘s”

We did not mark derivational morphemes (e.g. –ful in beautiful), because it is not clear that children recognize the relationships among words with and without such morphemes. For instance, do children consider beautyand beautiful to have the same root word? Instead, we allowed the CLAN program to count them as two different words.

Top of Page

Step 2. Calculating Number of Different Words and Total Number of Words using CLAN

Both NDW and TNW are calculated by doing a frequency analysis using the FREQ command:

e.g. freq +t*CHI +r6 –s”[+ bch]” +s”*-%%” +u @

• Select all six narratives for analysis using the File In button. The @ symbol will indicate that the files were selected.
• The +r6 option is used to exclude all repetitions from the counts (i.e. words followed by [/] or [//]).
• The –s”[+ bch]” option tells CLAN to disregard lines in the transcript that end in [+ bch], i.e. utterances that were not relevant to the narrative.
• The +s”*-%%” option tells CLAN to disregard letters after the hyphen, so e.g. play-ing and play-edwill be counted as two instances of the word play.
• The +u option tells CLAN to run the MLU command on all six narratives at once.

Here is the FREQ output for one short story:  

This command generates a list of words. The number beside each word indicates how often that particular word occurred in the transcript. A summary of the number of types and tokens is provided at the bottom of the word list.

In this example, NDW = 22 and TNW = 44. Note that this is an example of NDW and TNW scores for a very short file. Scores for all six stories will be higher.

Top of Page