What to Do
Click on a Part to the left. This will open a page with the sample videos and notes for that specific part. When it comes time to view a sample, you will be able to hide all the non-video parts of the page, making it easy to resize and share screen space with the SMS application.
Note to Users of Safari on OS X (Macs): Some of the sample videos served by YouTube are incompatible with this browser. If you run into this problem, download and use Mozilla Firefox or Google Chrome.
Click to View The Training Manual Introduction
The training program you will now see contains 47 speech samples, some of which you will see more than once. They are one, two, and three minutes in length. Some of the samples contain the speech of nonstuttering speakers, but most are from a diverse clinical population of persons who stutter. The speakers vary in age from 4 to 35 and are both male and female. The severity of stuttering ranges from very mild to very severe. Some people who stutter were recorded before they had received treatment, while others were recorded during the course of treatment. For some speakers, Standard American English was not their first language. Each person was videotaped while speaking in monologue in response to topic cards
and, occasionally, in response to the camera operator.
In terms of the measurement of speech rate and stuttering frequency, the basic requirements are: a method of rapidly noting the occurrence of each syllable as it's being spoken, and whether it is stuttered; a convenient way to note and sum the total number of syllables spoken and the total amount of speaking time; and a method for calculation of speaking rate in syllables spoken per minute and stuttering frequency in percentage of syllables spoken that are stuttered. In terms of the measurement of speech naturalness, the basic requirements are: an empirical method of periodically rating and recording the naturalness of an interval of speech, and a convenient way to average these ratings for an entire speech sample. We use a 1-9 rating scale, where 1 equals speech that has been judged highly natural, and 9 equals highly unnatural sounding speech (Martin, Haroldson & Triden, 1984).
The system that you're going to learn here accomplishes these requirements. It is the computer-aided counting and calculating software system known as the Stuttering Measurement System, or SMS (Ingham, Bakker, Moglia, & Kilgo, 1999). Now I'll give a brief demonstration of how this system works, and then you'll be ready to start your training.
On the computer screen before you is the SMS data acquisition screen. Assume you are listening to a speech sample spoken by a person who stutters and concurrently operating the SMS.
- First, depression of the left mouse button signals to the computer that one syllable has been spoken, so one syllable is counted. This is shown in real time on the computer screen in the SYLLABLE box, where the number of syllables counted is shown cumulatively.
- If you are using a laptop computer without an external mouse, press the corresponding button to record syllables.
- Depression of the right mouse button indicates that a moment of stuttering has been judged to occur. This is shown in real time on the computer screen in the STUTTER box, where the number of stutters judged to occur is shown cumulatively. The right mouse button should be depressed for the full duration of each moment of stuttering. When one occasion of stuttering is counted, one syllable is added to the syllable count as well.
- Again, for laptop computers without an external mouse, press the corresponding button to record stutters.
- In addition, periodically a tone sounds and a 5-second visual signal appears at the bottom left of the computer screen, signaling the listener to make a naturalness rating for the preceding interval of speech. This is accomplished by pressing a number from 1 to 9 on the computer keyboard, where 1 equals speech that has been judged highly natural, and 9 equals highly unnatural sounding speech.
- The clock at the top left of the screen displays cumulatively the length of time during which data are being collected, and it stops automatically at a preset duration, for example, at the end of 2 minutes.
When data collection for a given speech sample is completed, the computer screen displays the total number of syllables and stutters that were counted. In addition, the computer calculates and displays, for the entire sample, the client's speaking rate in syllables spoken per minute (SPM), the percentage of syllables spoken that was stuttered (%SS), and the average naturalness rating.
In the next paragraph of your workbook is this transcript of a 20-second segment of speech from a 4-year-old boy who stutters. Each syllable spoken is indicated by a dot above the vowel of the syllable. Each moment of stuttering is underlined. The location and duration of the naturalness rating signal, set to end at 20 seconds, is indicated within brackets. Now I'll demonstrate how the SMS is used for online recording of syllables, stutters, and naturalness while the clip of that sample is played. This will give you a picture of the way the system works and of the tasks you’ll be learning.
These data indicate that this boy uttered 39 syllables in this 20-second sample. In addition, six of those syllables were stuttered. Therefore, the speaking rate was 117 syllables per minute with approximately 15% percent syllables stuttered. Naturalness for the entire sample was judged to be 5 on the 9-point scale. These are the kinds of data that you will be collecting by the time you have completed this program.
Start learning how to use the SMS.
Let's set up the system for the first step in your training program. The SMS system offers many options to the user in terms of the kinds of analyses that can be made from these fundamental data collection procedures. These are described in detail in the appendix(below) and in the appendix of your workbook. However, here we'll just talk about the settings that are needed for you to begin.
- Start by bringing up the SMS program on your computer screen. I have mine set up with an icon on the desktop screen. (Be sure to note the names of the clever people who devised this system!)
- On the top row you will see several labels. Click
ADD SUBJECT
; type in 1 as the First Name and sample as the Last Name (for Sample 1); click on “Add” and thenClose.
Now you have added a file that will store the data for this sample. - Next, click on
SELECT SUBJECT.
Look forSample 1 SMS
, the sample label you just added, and highlight it by just clicking on it. Then clickOK.
Now you have opened the file in which you want to store the data. Footnote 1 - Next, click on the Run SMS label in the top row, and select
Half Screen Web Mode.
Notice that a warning appears reminding you that no other programs should be operating while you are using the SMS. After all other applications (except this browser) are closed, clickOK
, which will advance you to the next screen of the SMS called theSMS Run Configuration
screen. (The other labels on the top row are described in appendix.)
SMS settings that are relevant for your first task.
- Run Length. The SMS Run Configuration screen allows you to select the parameters you would like to use for a given data collection occasion. A
run
refers to the duration (in seconds or syllables) for which you would like to record data. For now, set this parameter at 60 seconds by typing60
in the box (or by clicking on the arrow and selecting 60). - Display Timer. Click in the box so that a check mark appears, indicating that this feature is
on.
This will allow you to see the clock on the screen as data are collected. Run Summary Printout. Ignore this item for now. (Its function is described in the appendix.) - Pause Control and Trigger Interval, will also be ignored for now, although these are important features for data collection during conversation samples, so check the appendix for their description.
- Standard Mousebutton Configuration. Click in the box so that a check mark appears. This establishes the left/top mouse button for counting syllables and the right/bottom mouse button for counting stutters. If you are left handed then you can reverse the mousebutton configuration, but be aware that all further descriptions for the mousebuttom operations will be reversed.
- Start/Stop Tones. Click the box so that it produces a check mark. This item determines whether tone signals will be heard at the beginning and end of a run.
- Use Syllable Mode and Syllable Count Target. These will be ignored for now. You may review this important option in the appendix. Display Syllable Counts. Click in the box labeled
Continuously.
This will allow your ongoing syllable counts to be displayed on the screen. - Display Stutter Event Counts. Click in the box labeled
Continuously
so that your ongoing stutter counts will also be displayed while you are counting. - Display Stutter Interval Counts. Ignore this item and consult the appendix for its function. Display Naturalness Rating: Ignore this item for now. We’ll return to it in Part Four.
- Rate Naturalness and Rating Period and Tone Prompt. Ignore for now. These features will also be described in Part Four.
- Periodic Display Parameters, Update Period and Display Duration. Ignore these features because we have already chosen to display our data continuously.
- Use Intervals and Interval Length. Ignore this item, but read about its function in the appendix.
When all of the parameters have been set to your liking, click OK.
This advances you to the data collection screen, which you will see as you collect data. At the top left of the screen is the clock, showing :00. At the top right is Sample 1, SMS (the sample name) and the RUN number, which for the time being is number 1. Each time you start another under the same file name, the run number will advance. You can also see that the display is set so that you will continuously see the number of syllables and stutters that you are counting. Leave your computer with this screen showing while you read the instructions that follow. You will come back to it shortly.
Now you are ready to begin Part One, Counting Syllables. For the remainder of the program all of the instructions you will need are contained in the workbook. We hope you enjoy learning to use the SMS and that you acquire an appreciation for the information it provides regarding important speech characteristics of people who stutter. Take this time to read the section on Part One. The workbook will tell you when you are ready to begin Sample 1. At that time, click here and begin the next phase of your training.
Footnote 1In the future, when you
want to return to this file, all you have to do is click Select Subject
from
the top row, find Sample, 1 in the list of files that have been stored,
click on it to select it, and then click on OK.
That way, all data
collected under this label will be kept in the same file for easy retrieval.
Return
APPENDIX
This appendix describes all of the options for data collection available in the Stuttering Measurement System (SMS), only a few of which have been utilized in the preceding training program. However, all of the skills acquired in the training program are fundamental to using the various data collection options. These options make the SMS extremely flexible and accommodating to a variety of clinical and research measurement goals.
To fully utilize the following descriptions, it will be helpful to have the SMS on a computer screen accompanying your reading. Begin with the first screen (after the credits have faded).
FIRST SCREEN
OK
or touch the Enter key.
NOTE: Throughout the program hot buttons
, which can be
activated either by clicking on them or by touching the Enter key, are noted
by a darkened border.
(Clicking on Cancel
will close the directory without a name being selected.)
Once you have selected a person's name, you are able to proceed to use of the
SMS to collect additional data, which will be recorded in that person's file,
or to view previously collected data. (See BROWSE below.)
OK
or touch Enter.
(Items on the Run Configuration screen are explained below. For now, touch Escape [Esc] to
return to the first screen.)
- number of syllables spoken (SYL),
- number of stutters recorded STUT),
- number of intervals during which stuttering occurred (SI),
- average number of syllables spoken per minute (SPM),
- average number of syllables spoken per minute during stutter-free intervals (SFSPM), \
- percentage of syllables stuttered (%SS),
- percentage of intervals containing stuttering (%SI),
- and average naturalness rating (NAT).
A minus one (-1 or -1.00) will appear in any column for which data were not collected. For example, if the data were not collected in intervals, -1 or -1.00 would appear in the columns labeled SI, SFSPM, and %SI. (Further description and calculation of each of these treatment efficacy measures are included as the last item in this appendix.)
The lower right hand corner of the display offers the user the opportunity to print out any or all of the data stored in the computer for the particular subject by indicating the starting and ending run numbers and clicking on PRINT. (Naturally, your computer must be connected to a printer for this application to function.) CLOSE will closed this data file and return you to the original screen.
Add
stores the person's name in the directory. Close
closes this file. Once a name has been added, it is necessary to open the directory and select the name (SELECT SUBJECT) BEFORE CONTINUING FURTHER. So that you will be able to follow the SMS instructions that follow, add a subject name, perhaps your own name or a John Doe
, now.
Delete
, or touch the Enter key. Then close the file.
RUN CONFIGURATION SCREEN
Now we are ready to learn about all of the data collection options available in the SMS. To open the Run Configuration screen, turn off all other applications on your computer, select a subject (as described above), click on SMS at the top of your screen, and indicate OK.
The Run Configuration screen will then appear. The top line indicates the name of the subject selected, that this is the SMS program, and the run number that will be recorded.
run.
The SMS program has two methods for automatically predetermining the length of a run. The Run Length option allows the user to preset the duration of time for which data will be collected. For example, if a speaker were to be providing a 5-minute long reading sample, the Run Length would be set a 5 minutes (300 seconds). When the reading begins, the person recording the data would press any key on the keyboard to start the clock and begin counting syllables and stutters, etc. until the time expired. The timer can be set for any duration. Clicking on the arrow in the Run Length box opens the box (indicated by the blue color) so that any number of seconds can be inserted by typing in the desired Run Length. It is also possible to select one of the predetermined run lengths y clicking on the up or down arrow and then clicking again to save the desired number (by turning the blue background to white).
trigger
the clock stoppage. Any interval from 0.1 to 5.0 seconds can be used by typing in the Trigger Interval box the desired number of seconds, or by selecting a number by clicking on the arrow and then clicking on the number until it occurs in the Trigger Interval box and the blue background has been removed. We typical elect 1.3 seconds as our Trigger Interval (therefore, this is the SMS's default interval), based upon findings by Bruce Ryan indicating that naturally occurring pauses in ongoing speech are typically less than 1.3 seconds in duration (Ryan, personal communication, 1993).
(NOTE: Some computers may not be fast enough to stop the run precisely at the selected Syllable Count Target, especially if syllable button presses are occurring very rapidly. One or two extra syllables may be counted.)
3
indicating a marginally normal ratings. If the speech sample being rated is relatively short, one naturalness rating made at the end of the completed sample may be adequate. More often, naturalness ratings are made periodically during an ongoing speech sample; the longer the sample, the more appropriate it is to make several naturalness ratings throughout. A given naturalness rating reflects the rater's perception of how natural-sound was the speech of the preceding interval, that is, the period of speech occurring since the last naturalness rating.
TO START THE PROGRAM:
After all the desired options have been selected, click OK
(or press Return). The data collection screen will appear and the program will begin when any key on the keyboard is pressed. Therefore, press a start
key just as the speech sample to be analyzed begins.
At the completion of a run, the summary data will automatically appear on the screen. At the same time, in the upper right corner a dialogue box will appear that allows the user to:
- Make another run with the same settings (the default selection),
- Make another run with new settings, or
- Return to Main Menu.
If Make another run with new settings
is selected (by clicking in the circle or moving the marker to the appropriate circle key by touching the down arrow on the keyboard and pressing OK
), the Run Configuration screen appears and changes can be made in preparation for the next run to be scored. If Return to Main Menu
is selected, the first screen reappears. Then the user may add or delete a subject, select another subject, browse the data files, print data files, or shut down the program by clicking on the x
in the top right corner of the screen.
At any time during data collection, pressing Esc will end a run. When this occurs, the above-described dialog box will appear, also giving the user the option to save or delete the data collected during the partial run.
THE SMS EFFICACY DATA
The data displayed at the end of each run are the same data that are stored in the speaker's name under BROWSE on the first screen of the SMS. A variety of calculations is available, depending upon the options selected from the Run Configuration screen before data collection for a given speech sample. All of the options and calculations are described below.
on
so that the denominator in this formula represents the speaker's actual talking time, excluding pauses of significant length.
This completes the summary of the data recording and analysis calculations available in the Stuttering Measurement System. Questions and comments may be sent to Janis Costello Ingham via email. Please refer to the printed manual for the address. It is found at the end of the appendix there.