by Jack Katz, Ph.D.
Since I was a child I have had a Central Auditory Processing Disorder (CAPD), but we knew of no speech or reading services and CAPD was unknown. Despite my limitations, I got into college by good fortune, in part because of affirmative action to increase the number of males in that college. I made the cutoff by 0.25 of a point.
In my 3rd semester I had a remedial speech class in which I first learned that words had these little components to them (called speech sounds). Things improved from then on, especially when I took a phonetics class that was critical for me personally. Three years later I got into CAP work by chance, but for a long time did not realize that I was studying myself. It seems that while I was working to help others I was also helping myself.
At the age of 72 things started to go backwards for me. My listening was not as sharp and my memory issues were getting worse (no doubt as a result of aging). I tried to find someone who could give me therapy for CAPD, but that never quite worked out. In desperation I took an internet program (Posit Science) that is not specifically for CAP, but for a variety of central issues associated with aging. Sure enough I felt so much sharper, in general, which clearly made things better including my CAPD. For about 5 years after that I felt fine, but when I was 78 I again felt things were going downhill, particularly in noise. My decoding skills were getting weaker too, often on the words that the kids with CAPD got wrong; I was making the same mistakes.
I tried to get someone to do auditory training with me, but again it was to no avail. After about a year I realized that if I wanted therapy I had better do it myself. People who heard that I was doing this said, “Do the therapy for yourself?? That’s not fair!” I asked them do you think I would cheat to make myself look better? Nobody wanted me to improve my skills as much as I did. So I tried to figure out how to do this. What happened was totally unexpected. My auditory skills improved and I also learned how to be more effective in giving this therapy to others. We always used a variety of “repairs” when there were errors or bumps in the road. But, in this case I learned so much more about what would help and when. The therapy that I decided to use is the Words-in-Noise Training (WINT) program, and specifically the WINT-3 program that is used with an audiometer.
There are 600 primary words in the program, most of them monosyllabic, in groups of 10-items per sublist. It is true that I know the majority of the words, but I surely don’t know if they are on a specific list or not. However, to make sure that my familiarity with the therapy materials did not elevate my scores, my criterion was not only to get the word right but also to be able to identify each of the sounds. Quite often I was able to figure out the word but was unable to identify the component sounds! When I did not get each sound I considered the word in error.
For myself I repeated the same WINT items from session to session until I performed well and then went on to the next section of items. However, in working with others, to avoid learning the correct words, the therapy continued from one sublist to the next during the session and on the next visit to go to the next sublist instead of repeating the previous session.
The first sublist is given with no noise and the next 10 words at +10dB SNR. Then subsequent sublists were given with 2dB more of noise each time until the last sublist is presented at 0dB SNR. The speech was always set at the same comfortably loud level. When I made an error I would stop the CD, mark down the word, my error, and the noise level. Then I tried a variety of techniques and kept careful track of what I did and what the result was (not simply correct or not, but gradations). Pretty soon I was doing more of some repairs and less of others.
Not surprisingly, turning off the noise was the most consistently effective way to improve my image of the word. But that did not insure a measurable advantage in noise. What it does give is a clear notion of the target sounds and what the word should sound like. This would carry over at least for words at lower levels of noise. Sometimes simply replaying the word (especially once I knew what it was) at the same noise level was helpful. But I found that Modeling the word for myself (saying it out loud and emphasizing the error sound) just before it came from the CD was a big help. In the past, I often used these and other repairs with my patients, but now I use them more and more strategically.
I found if I anticipated the correct word it would be helpful and sometimes putting my lips and tongue in the position of the difficult sound I was even more effective in making the association with the sound in noise. All of these strategies work with my patients as well. Although the loudspeaker is only about 2 feet from me, a very good strategy to use when I can’t hear a particular sound is to put my ear about 3 inches in front of the speaker. Sometimes one repair strategy will work somewhat better with one person than another.
When I finally plotted my data I was very pleased to see how much I improved over the 9 sessions it took to finish the first section of WINT (see Figure 1). However, I was surprised to see a bump for session #6 and wondered why. On the response sheet I had written that I had a very poor night’s sleep. I have frequently seen this effect on speech-in-noise scores in my patients and also if they had a bad day at school or an argument with their parents.
An interesting observation is that if you ignore session #6 my curve of improvement looks much like the data for 60 children (5 to 16 years of age) on the WINT program. Those data are shown in Figure 2 (Katz, 2009). Note that Figure 2 shows 15 sessions in the first round of therapy. Also the children dealt with all 600 words instead of the 70 that my data were based on. Compared to the 22-to-4% error for myself the comparable scores for the children were 24-to-13% error. Considering my advantages over these young children the two curves seem quite similar.
Importantly I wondered whether my benefits of therapy on the first 70 words were carried over in the following sessions with different words. Indeed they did. None of the remaining 6 sections of WINT had more than 11 errors and typically it took just 2 sessions to complete the 70 words in those sections. However, 2 sections of WINT took 4 sessions to complete which is still far better than the 9 sessions at the beginning of WINT. The average percent of error for the remaining 16 sessions was just 6.2% with a standard deviation of 2.5. This was far beyond my expectations! I will have to rethink my therapy approach. But importantly, how much can we extrapolate from my self-therapy of WINT to others who are seen for CAPD therapy?