I really didn't get good guidance on this in graduate school. The supervisors I had in my clinical placements all did it differently, so I felt like I didn't get a good grasp on it when I graduated. I thought I needed data on every kid after every session. I remember feeling like I couldn't really teach because I was worried about my data sheet. Of course we all have activities to work on each student's goals, but focus on the data instead of the student, are you really able to give quality instruction?
Data is important and we must take it, but what really is the best way? After twelve years in the school systems with a TON of large and mixed groups, I feel like my system is a good balance of taking accurate data AND providing quality instruction.
My theory is you should be teaching concepts/skills with a variety of modalities and cueing systems while focusing on the students (not your data sheet). While you are teaching, you shouldn't be taking data on responses. If you did, those responses would not be raw and would be influenced by your teaching method, cues, prompts, or even responses/comments from the other students in the group. If you are busy tallying for each student, it can take away from the lesson and the child's attention (I had one student who got nervous if she saw me tallying during a session). You can always make notations during the session on their performance but taking good "raw" data for progress monitoring should be done differently in my opinion!
When I was a clinical supervisor for a graduate program, it seemed that many of my student clinicians wanted to just drill a skill and take percentages during each therapy session, not leaving room for really teaching of concepts in depth and ensuring the child is understanding (which can lead to memorization instead of really learning it). This is why I think we should re-think our data taking process!
There are two main ways I do this.
#1 DATA SAMPLING: This method is great for SLPs who are required to have data for each session (district policy, Medicaid billing, insurance billing, etc...)
This is taking raw data with no cues or teaching to see exactly where the child is independently with the skill. I tend to do a rich teaching experience with the skill beforehand and then at the end of the session (last 5 minutes) I will do a quick data sample of the skill for each student. During the data sample, I won't cue them like I would during the teaching session.
Some prefer to do this before or after the instructional period.
What it looks like:
Take data on one student at a time at the beginning or end of the session (short bursts- 20 trials). Use the rest of the session for meaningful teaching, cueing, and modeling.
Example: Mixed group of articulation and language students. You are working on /r/ and compare/contrast, and antonyms with a book unit and related craft. During the lesson you will be teaching the concepts with the language students (using visuals, graphics organizers, highlighting text, ect..) and you will be providing articulation or phonological approaches for the articulation student (using visuals, mouth models, videos, mirrors, etc...). At the end of the session, have a set group of words or pictures to review (or data sample) and record how the student did.
Example of what data sample could be:
Articulation: A set group of 20 words for the student to say at the end of the session (do not provide cues and tally their % of accuracy.
Compare/Contrast: Have a set of 10 vocabulary pairs and request 1 similarity and 1 difference and record how many the were able to do independently.
Antonyms: Have a set of 10 pictures or words and request them to name an antonym and record correct responses.
"So what are the other students doing during data sampling then"? I usually have little challenges (different depending on ages) such as a drawing challenge, play dough building challenge, or lego building challenge ready for them. You can also simply have a a toy, coloring page, or app for them to play with while they wait their turn.
#2 DATA ROTATION: This method really is the best for large groups in the schools. I still basically use a data sampling technique with this (because remember, we want our therapy sessions to be rich instruction time, not busy tally taking). The key thing to remember with this method is to ensure you get good data on each student for progress notes. If you have a group of 5 students 1X per week and you report progress every 4 weeks, this wouldn't work. Most schools report progress anywhere from 6-12 weeks, so usually you can get them all in (and hopefully you don't have too many groups of 5!)
This is taking data on only one student per session and then rotating each session.
What it looks like:
You will only be taking data on one student during the session (but you will still be targeting everyone’s goals). This allows you to really focus on one student's data at a time to ensure the most accurate data. The key thing to remember is your teaching and cueing (is your data influenced by your cues or visuals?). This is why I like the data sampling even while doing data rotation.
ANOTHER NOTE ON DATA: If we take data using cues or prompts, it can be very subjective. Each therapist may interpret this differently. If you use language such as "minimum cues/maximum cues", a student may be scored differently depending on that therapist's interpretation of minimum and maximum. I also see goals that don't specify what types of cues the student needs (visual? auditory? tactile?) or they will list all of them (which is not really individualized). I typically do not write goals this way because if my student transfers, would the new therapist use the same cueing? Most of my goals say something like this: "independently with ___% accuracy." Meaning they were able to do it at that rate with no help. This lets me know they are truly mastering the skill. When I see goals that say "80% accuracy with min cues," again.... what does that even mean? My min cues may look different than your min cues. The only exception to that is if the child needs a visual aid, a visual can be a static or concrete thing (you don't want to put anything in a goal that could have multiple interpretations). This is why I LOVE data sampling. I use all my cues and prompts during teaching, and then I take a sample of what they can do without it (which is great raw data to use for progress tracking).
I would love to hear your thoughts on this or how you take data in large groups.
Before you go I have a AWESOME sample for you. I actually created this for data sampling purposes. They are called Anywhere Articulation and Language Strips. They are quick sets of targets (for almost every goal) to bring out when you want to take some quick data. They come in color and black & white. I cut mine out and put on rings but you can also put them in a binder for easier assembly (there is a quick video preview on the listing). Click here to view it on TPT
Click here to download a sample of the /l/ articulation page and a categories language page. If you liked this freebie and want to share it, please send the link to this blog post to your friend (not the freebie itself because that is a violation of my terms of use). Thank you! I hope you enjoy it!
Thanks for reading!
-Mandi
Panda Speech Therapy
Follow me on Instagram! (my preferred social media).
Oh and if you are in need of a resource for students with ARTICULATION and VOCABULARY goals, check out this product! It's also great for mixed groups. Speech Sound Vocabulary Challenge
