Last month we talked about SALT and language sample analysis (LSA) for very young children. Typically, we think of language sample analysis as a tool for evaluating children. The truth is, language sample analysis should be a part of every evaluation where we are targeting expressive language skills, adults included! What is more functional for speech language rehab than being able to have a conversation? This month we would like to show you how SALT and LSA can be used with adult clients. 

A quick note up top: only reports from the Analyze Menu will be used when assessing adults. The SALT Reference Databases contain samples from children; we cannot use them when assessing adults. However, reports generated from the Analyze Menu are incredibly useful. 

When we think of our adult rehabilitation clients, we typically are working on relearning language for those affected by stroke, traumatic brain injury, or other impairments. The process for taking a language sample with an adult patient is pretty much the same process that is used for taking a sample from children, just with an “adult rehab” mindset.

1. Make a plan

Start by making some decisions about what exactly you want to measure or explore, and over what time period. This will drive decisions about your sampling context, transcription protocol, the analyses you run, and repeated sampling.

Below is a table listing some functional language skills that can be useful when analyzing language for adult rehabilitation, and the corresponding measures generated in SALT: 


Functional Language Skill Measure
Vocabulary and generating longer sentences Mean length of utterance
Re-acquiring vocabulary Number of different words
Word retrieval/anomia Pause times
Utterance formulation or word retrieval, perseverations     Mazes
Code specific to patient (e.g., neologisms, paraphasia) Errors


2. Elicit the sample

You are generally going to want to elicit a conversational sample when working with adults. If you need a refresher on how to do that, take a look at our free online course here. You can find some of our tips for eliciting a language sample ia a previous post: “It Makes a Difference”.

Hint: We strongly recommend you record the sample with a digital recorder – your smartphone should do the trick – so that you can focus on your client, not your notes!

3. Transcribe the sample

I admit this is the hard part. Start by going back and referencing your plan; use this to decide what you want to include in your transcript. You may not need to include examiner utterances, which could save time on transcription. You may not need to mark bound morphemes, since the client might not have any challenges with grammar.  However, it is usually worth the time to mark pauses, mazes, and unintelligible segments. A small number of additional coding will give you a great deal of useful information that you can use to show improvement.

Your needs might vary, though. You can find a good overview of the options in one of our earlier blog posts, “Transcription Shortcuts.” And, of course, we have tutorials and free online courses to help.

Hint: take a look at the free SoundScriber program for “walking” audio playback: “Walking plays a small stretch of the file several times, then advances to a new piece, overlapping slightly with the previous one. With this feature, it is possible to transcribe continuously without having to manually pause or rewind the recording.”

4. Analyze the sample

I recommend that you start with the Standard Measures Report. 

The example SMR report below shows two linked conservational samples taken a few months apart while the client was receiving therapy post stroke with mild aphasia. Note that measures on the SMR provide excellent data for tracking change over time.

Example Standard Measures Report (SMR)

In this sample, measures that show change over time include mean length of utterance (increased from 6.75 to 8.09 in words) and speech intelligibility. Overall pause times decreased significantly from 14% of total time to 8.9% of total time. Communication challenges that showed minimal or no improvement include: maze words and abandoned utterances.

How it Works in Practice

In my first hospital setting I was covering inpatient and outpatient populations in a small rural hospital. For adult outpatients that I was treating for dysarthria or aphasia, I would usually include a conversational language sample as part of my initial evaluation and would then repeat the sample when the patient was getting close to discharge. I would use my digital tape recorder and record a sample, transcribe and analyze and include measures in my report such as pause time, number of different words, mean length of utterance and a gross measure of speech intelligibility.

There are two things that I loved about using language samples with my adult patients. First, I had great data for reporting progress. “Hey you are using more words now,” or “Your speech is much more intelligible!” When dealing with adult clients, it can be easier to explain progress with data and numbers rather than a feeling of “you seem better.” The data also helped when requesting more visits from insurance companies. They need proof that therapy is working in order to continue payment! 

The second, and most rewarding part of using LSA, was playing back the initial conversational sample for the patient towards the end of therapy. I still recall one patient who had a stroke and was affected with dysarthria and word retrieval impairments. He was feeling somewhat discouraged because his progress was slow and he potentially would not make it back to his baseline. Trying to reassure him that he really had made progress, I played his conversational sample from the first session. He was in disbelief! He didn’t even think that the recording was of him! He said, “I guess I do sound better. I can’t believe I was that bad.” In this interaction I had audio and data to encourage the patient and show him his own progress in a very concrete way. 

You can find a step-by-step on how to link T1 and T2 samples and run relevant reports in one of our earlier blog posts: “Linking Transcripts.”

On Custom Coding

Yet another more detailed way to use SALT for the adult population is to utilize special coding. SALT allows users to create custom codes for any skills or features that SLPs are interested in analyzing. 

One example that might be feasible is marketing neologisms, verbal or phonemic paraphasias  in a patient who is being treated for Aphasia. The SLP takes a conversational sample initially and then again later on in therapy. I’ve also transcribed the language generated when having patients describe the “Cookie Theft Picture” from the Boston Diagnostic Aphasia Examination picture just to get an idea of time one/time two. For example, if the patient says “nicpic” for “picnic,” that could be coded as nicpic|picnic[NEO]. The [NEO] codes that are marking neologisms could then be brought up in a report.  Any feature can be coded and then analyzed later.  The SLP has great data on progress over time. 

We have a free online course on custom coding. You can also see a step by step example of custom codes in use in one of our earlier blog posts: “My Dog Elton.”


I hope that I have convinced you that language sample analysis is incredibly useful when working with adult populations. It’s worth the time to transcribe and run analysis. So, don’t forget to use your knowledge of language sample analysis and SALT, even when working with adult clients! 


Joyelle Divall-Rayan, M.S., CCC-SLP
SLP: Vancouver, Washington
Director of Education & Training: SALT Software