Day in the Life of a Speech-Language Pathologist
Three speech-language pathologists wrote down everything they did on one ordinary workday. No interviews, no prompts. Just the day as it happened.
These characters are composites, built from dozens of real accounts, interviews, and community threads. The people aren't real. The experiences are.
What you'll learn
- A specific day in three SLPs' working lives, broken down hour by hour
- How the job differs across schools, hospitals, and pediatric clinics
- The documentation, the small victories, and what the job feels like on the inside
Irina's Tuesday
I get to school before the buses. The therapy room is a converted storage closet between the gym and the teachers' lounge. It has a small table, four chairs, a shelf of games and flashcards, and a poster of mouth positions that I laminated myself. There's no window. I set up articulation cards on the table, open my laptop, and check IEP deadlines on Frontline. I have 14 IEPs due this month. Three of them haven't been started. I drink coffee from a travel mug that says "SLP: Saving Little Phonemes." It was a gift from my college roommate. I didn't choose it.
First group of the day. Three kindergarteners working on /k/ and /g/ sounds. I walk to their classroom, collect them, and walk them back. One of them, a girl named Lina, starts crying because she doesn't want to leave art class. The art teacher, who is mid-demonstration with tissue paper flowers, gives me a look that says "this is your problem now." I tell Lina we have stickers today, which is true. She stops crying. We walk to the therapy room. For 30 minutes we practice words: "cat," "go," "cup," "game." I use a minimal pairs approach, contrasting words they're substituting. Lina says "tat" for "cat" every single time. The other two are making progress. I take data on a sheet, marking correct and incorrect productions in little boxes. Percentage correct: Lina 20%, Marcus 65%, Amira 80%. Amira is close to dismissal criteria. I'll bring it up at the next IEP meeting.
Individual session with Mateo, a 2nd grader with childhood apraxia of speech. CAS is the diagnosis that keeps me up at night. The brain knows what it wants to say. The motor plan to say it doesn't cooperate. Mateo has been working on multisyllabic words for months. Today we're doing three-syllable words with varying vowels. "Elephant." He gets it. "Banana." He gets it. "Butterfly." He pauses. I model it slowly. He tries. "Buh-ter-fly." Clean. All three syllables, correct stress pattern, accurate vowels. I give him a high five so hard my hand stings. He grins. He says it again: "butterfly." Still clean. I write it down in capital letters in my data sheet because moments like this deserve capital letters.
Group session. Three 3rd graders, language therapy. We're working on inferencing from short passages. I read a paragraph about a boy who puts on a raincoat and grabs an umbrella. "What can we figure out about the weather?" One kid says "it's raining." Another says "he thinks it might rain." The third says "he likes umbrellas." All three answers tell me something different about where each student is with inferencing. The first is literal. The second is inferential. The third is a 3rd grader who would rather talk about umbrellas. We work through four more passages. I track their responses on a rubric. By the end, all three can identify at least one inference per passage, but only one can explain how she knew. Explaining the reasoning is the harder skill. That's next month's target.
IEP meeting for a 5th grader named Destiny. Present: me, the special education coordinator Mrs. Fuentes, the classroom teacher, a school psychologist, and Destiny's parents. Destiny has a language-based learning disability that affects reading comprehension and written expression. She's been on my caseload for two years. Her parents want to increase her speech-language minutes from 90 to 120 per month. Mrs. Fuentes wants to reduce them to 60 because Destiny's standardized scores improved. I pull up Destiny's progress monitoring data. Her scores improved on structured tests, yes. But her classroom performance, which I tracked through work samples and teacher interviews, shows she still struggles with paragraph-level comprehension when the text isn't simplified. I recommend maintaining 90 minutes with adjusted goals that target grade-level text. The parents agree. Mrs. Fuentes agrees reluctantly. The meeting was scheduled for 30 minutes. It runs 50.
I missed my 10:00 group because the IEP meeting ran long. I walk to the 2nd-grade classroom to apologize to the teacher, Mrs. Pak. She's in the middle of a math lesson. She looks at me over her glasses. The look says: "This is the third time this month." She's not wrong. I tell her I'll reschedule the group for Thursday. She nods. I write myself a note to reschedule and also a note to bring Mrs. Pak coffee on Thursday. Relationships with teachers are half the job. If they don't want to release kids to you, your schedule falls apart.
Two back-to-back articulation groups. Six kids total across two 25-minute sessions. /s/ blends in the first group, /l/ in the second. I run a dice game where they roll and say a target word that many times. The kids like the game. I like it because it generates a high number of productions in a short time, which means more data. The /s/ blend group averages 78% accuracy at the sentence level. The /l/ group is at 55% at the word level. Very different places in the therapy process. I adjust my data sheets and move on.
Lunch. I eat a turkey sandwich at my desk while writing present levels of performance for three upcoming IEPs. Present levels are the section of the IEP that describes where the student is right now: what they can do, what they struggle with, how their disability affects their access to the general curriculum. Each one takes about 20 minutes if I have my data organized, 40 minutes if I don't. Today I have my data organized for two of the three. The third student's data is in a notebook in my car. I write the two I can and make a note to get the notebook at dismissal. I finish the sandwich at 11:50. I did not taste it.
Push-in session in a 1st-grade classroom. This is collaborative teaching: I work alongside the classroom teacher during literacy centers. Today she's running guided reading groups and I'm at the vocabulary station with four students, two of whom are on my caseload and two who aren't. We're working on tier-two vocabulary from a book about ocean animals. I pre-teach words like "enormous," "predator," and "shallow" using picture cards and kid-friendly definitions. The kids on my caseload get embedded language targets within the activity. One of them, who's working on describing, tells me that a whale shark is "big, spotty, and doesn't eat people." That's a three-element description with accurate content. I write it down. Push-in is my favorite service delivery model because the kids don't miss class and the language targets happen in context. It's also the hardest to schedule because I need a teacher willing to share their classroom time.
A kindergarten teacher stops me in the hallway. She's concerned about a student who substitutes /w/ for /r/ and /th/ for /s/. She wants him evaluated. I go observe the student during a reading circle. He's 5 years old. I listen for about 15 minutes. The /w/ for /r/ substitution is a developmental pattern that typically resolves by age 6 to 7. The /th/ for /s/ is called a frontal lisp and is also common at this age. Neither pattern is disordered at 5. I screen him informally, checking stimulability for both sounds. He can produce /s/ with a model. I send a note home to his parents recommending monitoring and re-evaluation at the start of next year if the patterns persist. The teacher seems unsatisfied. She wanted an eval. But evaluating a developmentally appropriate pattern wastes everyone's time and adds a child to an already overloaded caseload. I explain this. She nods but I can tell she'll bring it up again in January.
Three more pull-out sessions back to back. A language group for 4th graders working on narrative structure. An articulation session for a 1st grader with a lateral lisp on /s/ and /z/. An individual session with a 6th grader working on social communication: reading facial expressions, understanding sarcasm, adjusting language for different listeners. The 6th grader tells me his classmate said something was "fire" and he asked if there was an actual fire. We talk about figurative language. He writes "fire = good" in a notebook he keeps for slang translations. The notebook has 34 entries. He started it himself. That's initiative I didn't teach him.
Last session. Individual with Andre, a 7th grader who stutters. Andre is 13. He's been in speech therapy since he was 4. He hates it. He told me this in our first session and I told him I respect that. Today we're working on voluntary stuttering, which is a technique where you intentionally stutter on a word to reduce the fear of it happening involuntarily. It sounds counterintuitive. It is counterintuitive. Andre thinks it's the worst idea anyone has ever had. But he agreed to try it this semester. Today we practiced in a hallway conversation with Mr. Davis, the custodian, who I prepped beforehand. Andre walked up and said, "Hey Mr. D-D-Davis, do you know where the, the, the recycling bin is?" He did it. Voluntary prolongation and repetition, right there in the hallway. He looked mortified. Mr. Davis answered normally. Andre exhaled. We walked back to the therapy room and he said "that was terrible." I said "you did it though." He said "yeah." That's the win. Not fluency. Willingness.
Students dismissed. The building gets quiet fast. I stay at my desk and write SOAP notes for today's 22 sessions. Each note needs to include: the goal addressed, the activity used, the data collected (percentage correct, number of trials, level of prompting), and a brief narrative about the student's performance. If I'm efficient, each one takes about 90 seconds. If a session was complicated, like Destiny's IEP or Andre's voluntary stuttering, the note takes longer because I need to document parent communication, decisions made, or clinical reasoning. I type fast and don't edit much. These notes exist for compliance. If anyone ever audits my caseload, these paragraphs are the proof that I saw each student and worked on their goals. The work itself already happened.
Finish notes. Check email. Four parent emails: one asking when her daughter will be "fixed" (she won't be fixed, she'll make progress, and I need to find a kind way to reframe that), one confirming an observation time, one asking if speech therapy is "really necessary" (yes), and one thanking me for working with her son, which I read twice because those emails are rare. Two new teacher referrals. One from administration about an upcoming compliance review that will require me to pull documentation for 10 randomly selected students. I respond to the urgent ones and star the rest for tomorrow.
Drive home. Call my mom. She asks how work was. I say "Mateo got butterfly." She says "that's nice." It is. Mateo has been working on that word for two months. His mother will hear it at home and probably won't know it's new. That's okay. The best outcomes in speech therapy are the ones that disappear into normal life. The kid just talks. Nobody notices. That's the goal. That a word like "butterfly" stops being hard and starts being just a word.
Wendell's Thursday
I get to the hospital before shift change. The SLP office is a small room on the second floor with two desks, a mini fridge, and a bookshelf full of assessment manuals nobody opens anymore because everything is digital now. I check the census board. Eight consults came in overnight. Three say "evaluate for dysphagia." Two say "cognitive-linguistic assessment." Two say "speech and language evaluation." One says "eval for communication" with no other detail, no diagnosis, no history. Just "eval for communication," like ordering a pizza with no toppings specified. I'll figure it out when I get there.
Chart review in Epic. I pull up the first patient: Mrs. Darby, 76, left hemisphere stroke three days ago. CT confirmed a middle cerebral artery infarction. She has nonfluent aphasia, meaning she understands most of what's said to her but struggles to produce words. G-tube was placed yesterday for nutrition. The nursing notes say the family has been asking when she can eat. The family always asks when they can eat. It's the question that defines the first week after a stroke. Not "will she walk again" or "will she talk again." When can she eat. Because eating is human in a way that hospital stays try very hard to erase.
Bedside swallow evaluation on Mrs. Darby. Room 412. Her daughter is sitting in the corner chair with a crossword puzzle she hasn't started. I introduce myself. Mrs. Darby looks at me and tries to say something. What comes out is "I... the... want..." and then she stops, frustrated. Her daughter says "she wants to eat." I tell them that's what I'm here to figure out. I start with ice chips. I place a small amount on her tongue and watch. She moves it around, initiates a swallow. Timing looks adequate. No wet vocal quality after. I try pudding-thick consistency. She swallows it. A slight delay in initiation, maybe a half second longer than I'd expect, but no overt signs of aspiration. No cough, no throat clearing, no change in voice quality. I try nectar-thick liquid. She takes the first sip, swallows it. Second sip: cough. Not a big cough, but a reflexive one, right at the moment of swallow. I stop. That cough tells me something might be entering the airway with thin and near-thin liquids. But a bedside eval can't see what's happening below the vocal folds. I need imaging. I document: "Bedside evaluation suggests possible aspiration risk with thin liquids. Recommend modified barium swallow study for further evaluation." I tell the daughter we need one more test. She nods. She's been nodding for three days.
Second consult. Reggie, 34, traumatic brain injury from a motorcycle accident six days ago. He's in a step-down unit. The consult says "cognitive-linguistic assessment," which means someone noticed he's not thinking clearly and wants me to quantify it. I walk in. Reggie is sitting up in bed. His wife, Shayla, is next to him holding his hand. I introduce myself and explain what I'm going to do. I start with orientation: "Reggie, can you tell me where you are?" He says "the hospital." Which one? He thinks. "Baptist?" Close. It's Vanderbilt. I ask him the date. He says March. It's October. I ask who the president is. He gets it. Oriented times three means person, place, and time. He's oriented times two: person and situation, but not place or time.
I move to attention tasks. I read him a series of numbers and ask him to repeat them back. He gets five digits forward, which is low-normal. He can't do three digits backward. I try a sustained attention task, listening for a target letter in a string of letters I read aloud. His accuracy drops sharply after about four minutes. He starts looking at the TV, which is on mute. I note: attention fades at the four-minute mark. I move to memory. I tell him three words and ask him to remember them. Five minutes later, he recalls one. Shayla, who has been watching, starts finishing his sentences when he pauses. "He means the..." and "What he's trying to say is..." I understand why she does it. But I need to hear what Reggie can do, not what Shayla can fill in. I ask her, gently, to let Reggie try, even if it takes longer. She nods and squeezes his hand harder. I spend another 20 minutes on problem-solving and reasoning tasks. Reggie is trying. He fatigues fast. My assessment: moderate cognitive-linguistic deficits affecting attention, memory, and executive function. I'll write the full report this afternoon and recommend daily therapy.
Down to radiology for Mrs. Darby's modified barium swallow study. Dr. Anand, the radiologist, meets me there. Mrs. Darby is wheeled in. The MBSS is the gold standard for evaluating swallowing. The patient swallows food and liquid mixed with barium while we watch in real time on fluoroscopy. It's an X-ray movie of the swallow. I start with thin liquid, the consistency that concerned me at bedside. Mrs. Darby takes a sip. On the screen, I watch the liquid move through her mouth, pool in the valleculae briefly, then spill over into the pharynx. The epiglottis inverts. The liquid enters the airway to the level of the vocal folds but does not pass below them. Penetration, not aspiration. She doesn't cough. That's worse than it sounds, because it means her airway isn't protecting itself. Silent penetration.
I try nectar-thick liquid. She swallows it cleanly. No penetration. Puree consistency: safe, timely swallow, full clearance. Mechanical soft: safe with small bites, a slight residue in the pyriform sinuses that clears with a second swallow. I have enough data. My recommendation: International Dysphagia Diet Standardisation Initiative level 5, which is minced and moist, with nectar-thick liquids. No thin liquids. She can eat, just not everything, and not the way she used to. I tell Dr. Anand my recommendations. He agrees. I'll go talk to the family.
Family meeting with Mrs. Darby's adult children. Her son, James, drove in from Clarksville this morning. Her daughter, Cora, has been here since the stroke. They sit in the family conference room, which has a round table, fake flowers, and the kind of lighting that makes everyone look tired. I explain the MBSS results. I use the word "aspiration" and then immediately say: "That means food or liquid going into the airway, toward the lungs, instead of into the stomach." James asks "so she can never drink water again?" I explain: this is where she is right now, three days after a stroke. The brain is still healing. We'll reassess in a week. Many stroke patients improve their swallowing function as they recover. I can't promise she'll get back to thin liquids, but I can say it's possible and we'll work on it. Cora asks about Thanksgiving. It's October. Thanksgiving is seven weeks away. But Cora is already imagining her mother at the table, not eating the meal. I tell her we'll take it one week at a time. James nods. Cora cries quietly and says "she makes the cornbread dressing." I don't have a clinical response to that. I just sit with it for a moment. Then I walk them through the diet textures: what she can eat, what it looks like, how to make sure pieces are small enough. Cora writes it all down on the back of a Vanderbilt parking receipt.
Two more bedside swallow evaluations. Both are post-surgical patients, one after a cervical spine fusion, one after a cardiac catheterization. Neither has neurological involvement. I run a brief bedside screen: ice chips, water, crackers. Both swallow without difficulty. Both cleared for regular diet. Each evaluation takes about 15 minutes including documentation. These are the straightforward ones. They balance out the Mrs. Darbys.
Lunch in the SLP office with my colleague Ramona. She's eating a salad out of a mason jar, which she does every day. I'm eating leftover pasta that my wife Anita packed. Ramona tells me about the new EMR documentation template that corporate rolled out last week. It adds four additional clicks to every progress note. You now have to select a treatment setting from a dropdown, confirm the patient's diagnosis from a list that doesn't include half the diagnoses we actually treat, and check a box that says you reviewed the plan of care, which we already document in the note itself. Ramona calculated that four extra clicks times 12 patients per day times 250 work days per year equals 12,000 extra clicks per year. She submitted this calculation to the department manager. The department manager said she'd "pass it along." Ramona and I both know what that means.
Speech-language evaluation on Earl, a 68-year-old man with Parkinson's disease. He was admitted two days ago for a fall, but the real reason his wife pushed for a hospital stay is that she can't understand him anymore. His voice is the textbook Parkinson's presentation: low volume, reduced pitch variation, imprecise consonants, and a flat affect that makes him sound like he doesn't care about what he's saying even when he does. I start with a Lee Silverman Voice Treatment assessment, which measures sustained vowel phonation, pitch range, and maximum volume. Earl's maximum volume during sustained "ah" is 62 decibels. Conversational speech hovers around 55 decibels. For reference, normal conversational speech is about 65 to 70. He's literally too quiet to be heard in a normal room.
I ask Earl about his background. He tells me, in a voice I have to lean in to hear, that he was a high school football coach for 30 years. Coached in Murfreesboro. His teams won two state championships. He used to stand on the sideline and yell plays across the field and every player could hear him. His wife, sitting behind him, says quietly: "Now I can't hear him from across the dinner table." That detail lands on me. I write it down. I keep my face clinical. I explain LSVT LOUD, an intensive voice treatment program: four sessions a week for four weeks, focused on training him to speak at a louder volume and recalibrating what "normal" feels like. He nods. He says, very quietly, "let's do it." I schedule his first outpatient session for next week.
Documentation. I have six evaluations from today that need full reports. Each report includes patient history, assessment results, clinical impressions, diagnosis, recommendations, and a plan of care. In the hospital system, these reports are legal medical documents. They go into the patient's permanent record. They're read by physicians, nurses, dietitians, case managers, and insurance reviewers. Each one takes 20 to 25 minutes. I type Mrs. Darby's first because her diet order needs to be in the system before dinner. Then Reggie's. Then Earl's. Then the two routine post-surgical screens, which are short. The last report, for the "eval for communication" consult, turns out to be a patient with a tracheostomy and a speaking valve question. That evaluation took 20 minutes and the report takes 15. I type until 3:30. My fingers ache. My coffee is cold. I drink it anyway.
Quick check on Mrs. Darby. I stop by room 412. She's sitting up, eating puree from a hospital tray. Mashed potatoes and something that might be chicken. She sees me and gives a thumbs up. Her daughter Cora is in the corner chair. She mouths "thank you" over Mrs. Darby's head. Mrs. Darby takes another bite. She is eating. Three days ago, the question was whether she would. Today she is. I watch for a few seconds to make sure her swallow looks safe, then move on. I'll reassess in a week.
Supervision. I'm the clinical supervisor for Jolene, who's completing her clinical fellowship, the supervised year required after graduate school before you can practice independently. Jolene evaluated a laryngectomy patient this morning, a man who had his larynx removed due to cancer and is learning to use a tracheoesophageal puncture voice prosthesis. She brings me her evaluation report. Her assessment of the patient's TEP function is solid. Her recommendations are appropriate. I suggest she add a note about training the patient's wife on prosthesis care, since he's being discharged next week and someone at home needs to know how to clean and troubleshoot it. Jolene nods and writes it down. She's good. She'll be a strong clinician. I tell her that, because I remember how much it mattered when someone told me.
Drive home. Nashville traffic on I-40 is what it always is. Anita is making dinner when I walk in. She asks about my day. I say "Mrs. Darby can eat again." Anita says "good." I don't mention Earl. Not because I'm protecting Anita from sadness but because the feeling hasn't settled yet. Earl's voice, the one that used to carry across a football field, now barely crosses a dinner table. I'll think about Earl later, while I'm brushing my teeth or lying in bed. That's where the cases go when the day ends. Not into the notes. Into whatever part of your brain holds the people whose voices you're trying to give back.
Yumi's Wednesday
Arrive at the clinic. It's a converted house near Hawthorne Boulevard. The waiting room has a fish tank, a basket of board books, and small chairs that adults sit in anyway. I check the schedule: seven patients today, 45-minute sessions with 15-minute breaks between for notes. The schedule looks doable. Schedules that look doable usually aren't. There's always a late arrival, a crying child, or a parent who needs to talk after the session runs out. I pour coffee in the kitchen, where someone left a box of doughnuts with a note that says "Happy Wednesday, speech friends." I eat half a maple bar and open my materials for the first session.
First patient: Oliver, 3 years old, on the autism spectrum, minimally verbal. He uses about 10 spoken words and we've been building his communication through a LAMP-based AAC device, a PRC Accent with a 36-location vocabulary page. LAMP stands for Language Acquisition through Motor Planning. The idea is that every word lives in the same place on the screen, always, so the motor pattern becomes automatic, the way your fingers know where letters are on a keyboard. Today's targets: "more" and "help." I blow bubbles. Oliver watches them float. I wait. He reaches for the bubble wand. I point to the device. He looks at it. He looks at me. He presses "more." I blow more bubbles. He laughs. We do this four times. Each time, he presses "more" independently, without my pointing or prompting. Four independent requests. I mark the data. Then I hold the container lid on too tight and hand it to him. He struggles. I wait. He presses "help." I open it. We have a communication system. It's not speech. It's language. And right now, language is what matters.
Kaia, 5 years old, /r/ therapy. She's been coming twice a week for 8 months. The /r/ sound is the white whale of articulation therapy. It's produced in the back of the mouth with a tongue shape that most people can't see or feel, which makes it incredibly hard to teach. There's no simple visual model. You can't just say "put your tongue here." I've used tactile cues, visual feedback with a mirror, analogies about a "mountain tongue" and a "tiger growl," and an ultrasound biofeedback app that shows tongue position in real time. Kaia has been stuck at the word level for months. She can say /r/ in isolation, can get it in "run" and "right" about 60% of the time, but it falls apart in conversation. Today something clicks. We're doing a picture description activity. She picks up a card with a fire truck. She says "the red fire truck is going fast." The /r/ in "red" is clean. Not approximated, not distorted. Clean. Correct tongue position, correct resonance. At the conversational level, unprompted, in a spontaneous sentence. My face stays neutral. I've learned not to over-celebrate because it creates pressure. The kid doesn't need a standing ovation. She needs to believe it was always in her. But inside, I am doing cartwheels. Eight months. I write it down in my notes with the time: 9:47 AM.
Parent coaching session. Mila, 2 and a half, and her mom Jae-won. Mila is a late talker with about 15 words. No other developmental concerns, good receptive language, strong play skills. The research is clear: parent coaching is more effective than direct therapy for late talkers at this age, because the parent is with the child for 40 to 50 waking hours a week and I'm with her for 45 minutes. Today I'm teaching Jae-won "communication temptations," which are structured situations that motivate a child to communicate. I demonstrate: hold a desired toy where Mila can see it but not reach it. Wait. Model the word: "ball." Pause. Wait some more. Mila looks at the ball. Looks at her mom. Looks at the ball. "Ba," she says. Jae-won's eyes go wide. She does it herself with a different toy. Holds up a cup. "Cup." Waits. Mila says "cuh." Jae-won looks at me. I nod. She does it again with crackers. Mila says "cra." By the end of the session, Jae-won is running the technique perfectly, pausing at the right moments, modeling without demanding. Mila says "ba" for ball unprompted while Jae-won is putting on her coat. Jae-won tears up. I hand her a tissue from the box I keep on my desk because this happens more than you'd think.
Cancellation. The 11:30 family called this morning. Their son has a cold. Cancellations are a mixed blessing. I lose the session, which in a private practice means lost revenue (the clinic doesn't bill for no-shows unless it's a pattern), but I gain time. I use this hour to write progress reports. I have 12 due by Friday. Each one summarizes the last 90 days of therapy: goals targeted, data collected, percentage accuracy at each level, clinical impressions, and recommendations for the next quarter. They go to parents and, if the child has an IEP, to the school district. I finish three reports in the hour, which is fast but possible because I keep running data on all my kids and don't have to go digging through session notes. The clinic uses a practice management software that auto-generates some of the formatting, but the clinical narrative is all me. Three down, nine to go.
Lunch. I eat a rice bowl at my desk and scroll the ASHA Community forums. Someone posted a question about caseload sizes and there are 200 angry replies. School SLPs saying their caseloads are 75, 80, 90 kids. One person said 104. The thread is a mix of venting, solidarity, and people who left the schools for private practice or medical settings and are cautiously admitting they're happier. I close the tab. I've been in private practice for four years and I still feel guilty about it sometimes, like I abandoned the public school kids who needed me. Then I remember that the guilt was part of why I left. You can't pour from a caseload of 80.
Theo, 7 years old, expressive language disorder. His receptive language is strong. He understands everything. But putting thoughts into complex sentences is where it breaks down. We've been working on compound and complex sentences using conjunctions. For four months, the target has been "because." I model it. He practices it. We use visual sentence strips. We play games where he has to explain why things happened. Today, during a barrier game where he's describing a picture I can't see, he says: "I don't want to go outside because it's wet." Perfect. Subordinate clause, correct conjunction, logical causal relationship, and completely spontaneous. He wasn't repeating a model. He wasn't filling in a blank. He generated that sentence from his own thought. I write it down verbatim in my notes: "I don't want to go outside because it's wet." Spontaneous generalization is the goal of every language therapy session. It means the skill left the therapy room and entered the child's actual language system. That sentence is worth four months.
Evaluation. New patient: Sadie, 4, referred by her preschool for "unclear speech." Her parents, who are in the room, say she's hard to understand, especially with unfamiliar listeners. Her preschool teacher estimated about 50% intelligibility in the classroom. At age 4, a child should be roughly 90 to 100% intelligible to strangers. I administer the Goldman-Fristoe Test of Articulation, Third Edition, which is a standardized picture-naming test that samples all English consonant sounds in initial, medial, and final positions. Sadie names pictures while I transcribe her productions. She substitutes /d/ for /g/, /t/ for /k/, /f/ for /th/, and deletes final consonants in most words. I also administer the CELF Preschool-3, which tests receptive and expressive language skills. Her language scores are within normal limits: she understands and uses age-appropriate vocabulary, grammar, and sentence structure. The problem is isolated to speech sounds.
Results: her articulation score is 2 standard deviations below the mean for her age. That's a moderate phonological disorder affecting intelligibility. The pattern is consistent: she's simplifying sound structures in predictable ways. This is treatable. Most kids with her profile respond well to a cycles-based approach targeting one phonological pattern at a time. I explain all of this to her parents, who look relieved. The dad says "so it's not something worse?" I tell him her language and cognition are strong. This is a speech sound issue, and therapy will help. I'll write the full evaluation report tonight and schedule her to start next week.
Last patient: Ezekiel, 6 years old, fluency. He stutters. Repetitions and prolongations, mostly on initial consonants and at the beginnings of sentences. About 8% of his syllables are disfluent, which is moderate. We've been working on easy onset (starting words with a gentle, slow voicing) and light articulatory contacts (touching the tongue and lips together softly instead of pressing hard). He hates both techniques. He told me last week that they make him "sound weird." I validated that. I said: "You're right, they do sound different from how you usually talk, and it makes sense that feels weird." Today we negotiate. This is real clinical work, not just drills. I ask him where stuttering bothers him most. He says reading aloud at school. I ask where it bothers him least. He says recess. We agree: he'll practice the techniques during reading aloud at home, three nights a week, and I won't ask him to use them at recess. He also gets to pick the game we play during the last five minutes of each session. He picks Connect Four. Negotiating therapy goals with a 6-year-old is its own clinical skill. It's not in any textbook. But if the kid doesn't buy in, the techniques don't transfer. Ezekiel plays Connect Four. He wins. He stutters on "I win" and doesn't care.
Session notes for six patients. Each note includes the date, session duration, goals targeted, activities used, data, and a clinical narrative. The notes take about 40 minutes total. Then I start writing the evaluation report for Sadie. The report has sections: referral information, background history, assessment results (with standard scores, percentile ranks, and confidence intervals), clinical impressions, diagnosis, and recommendations. I type the first three sections, plug in Sadie's scores, describe her error patterns with phonetic transcription, and get halfway through clinical impressions before my eyes blur. I've been looking at a screen since 8:15 this morning. I save the report. I'll finish it tomorrow during my first cancellation, because there's always a cancellation.
Drive home. My partner Kay asks about the day. I say "Kaia got her /r/." Kay says "finally!" because Kay has been hearing about Kaia's /r/ for 8 months. Every week: "Kaia was close today." "Kaia almost had it." "Kaia's tongue is doing something new." And now, finally: "Kaia got it." The progress in this job is measured in months, sometimes years. A child starts therapy with a sound they can't make, a sentence they can't form, a word they can't say. You work on it. You take data. You adjust your approach. You wait. And then one day, in the middle of a picture description activity at 9:47 on a Wednesday morning, they say "red" and it's clean. You tell the person closest to you, and they say finally, and that's enough.
Frequently Asked Questions
What does a speech-language pathologist do all day?
It depends on the setting. School SLPs run back-to-back therapy sessions with students, write IEPs, attend team meetings, screen referrals, and document progress on caseloads that often exceed 60 kids. Hospital SLPs evaluate and treat swallowing disorders, aphasia, cognitive deficits from brain injuries, and voice disorders, spending significant time on bedside evaluations, imaging studies, and family education. Pediatric clinic SLPs work one-on-one with children on articulation, language development, fluency, and augmentative communication. Across all settings, the work involves far more documentation, parent or family communication, and coordination with other professionals than most people expect.
How many patients does an SLP see per day?
School SLPs typically see 18 to 25 students per day in individual and small group sessions lasting 20 to 30 minutes each. Hospital SLPs see 6 to 10 patients, but evaluations can take 45 to 90 minutes and often involve imaging studies and family conferences. Outpatient pediatric SLPs see 6 to 8 patients in 45-minute sessions. In every setting, documentation time often equals or exceeds direct patient time. A school SLP who sees 22 students also writes 22 session notes. A hospital SLP who completes 6 evaluations writes 6 full reports. The patient count is half the picture.