Goal Attainment Scaling (GAS) is a method for measuring individual progress toward personalized goals, most commonly used in rehabilitation, mental health, and education. Instead of comparing a person’s performance to a population average, GAS defines what success looks like for that specific person, then scores their actual outcome on a standardized five-point scale. It was developed in 1968 by clinical psychologist Thomas Kiresuk and biometry graduate student Robert Sherman at the Hennepin County Medical Center in Minnesota, originally as a way to evaluate community mental health programs.
How the Five-Point Scale Works
Every goal in GAS is built around a five-point scale that ranges from −2 to +2. The center of the scale, zero, represents the “expected outcome,” meaning the level of progress a clinician and patient agree is realistic within a set timeframe. The levels above and below capture how far the actual result deviates from that expectation:
- −2 (much less than expected): The starting point or baseline, sometimes called the worst anticipated outcome.
- −1 (less than expected): Some progress, but noticeably short of the goal.
- 0 (expected outcome): The person hit the target that was set at the beginning.
- +1 (more than expected): Progress exceeded what was predicted.
- +2 (much more than expected): The best realistic outcome imaginable for this goal.
Each level is defined in concrete, observable terms before treatment begins. For a child working on handwriting, for example, −2 might be “copies fewer than 3 letters legibly,” while 0 might be “copies 8 letters legibly within one minute,” and +2 might be “copies a full sentence with consistent letter size.” This specificity is what makes GAS different from a vague progress note. Everyone involved knows exactly what each level looks like.
Why Personalized Goals Matter
Standard outcome measures, like a walking-speed test or a depression questionnaire, compare everyone against the same benchmark. That works well for research, but it can miss what matters most to an individual patient. A person recovering from a stroke might care far more about being able to button a shirt than about grip strength measured in kilograms. GAS captures those personal priorities.
This design also sidesteps a common measurement problem called floor and ceiling effects, where a test is either too hard or too easy to detect meaningful change. Because each scale is custom-built for the person, the levels are calibrated to their realistic range of progress. Studies in pediatric rehabilitation for conditions like cerebral palsy have confirmed that GAS avoids floor effects and shows minimal ceiling effects, making it sensitive enough to detect change even in populations where standardized tests often plateau.
Setting Goals Collaboratively
The GAS form is designed to be filled out together by the clinician and the patient (or a parent or caregiver acting as proxy). This collaborative process is central to how GAS works. The clinician brings expertise about what’s medically realistic, while the patient identifies what outcomes actually matter in their daily life. Goals can also be weighted by the patient based on importance or difficulty, so that the final score reflects their priorities, not just their clinician’s.
Good GAS goals follow the same logic as SMART goals: they’re specific, measurable, and tied to a clear timeframe. The most common mistake in writing them is making the levels too vague. “Patient will improve mobility” tells you nothing. “Patient will walk 50 meters with a cane in under 3 minutes” gives you a clear target you can observe and score. Each of the five levels needs that same precision, with enough separation between them that you can reliably tell one from another.
Calculating an Overall Score
When a person has multiple goals, GAS combines them into a single summary number called a T-score. The formula converts the raw −2 to +2 ratings into a score where 50 means the person achieved exactly what was expected across all their goals. Scores above 50 indicate better-than-expected progress overall, and scores below 50 indicate the person fell short.
The calculation accounts for two things: the weight assigned to each goal (reflecting how important it is to the patient) and an assumed correlation between goals, which Kiresuk and Sherman set at 0.3 for most practical purposes. If all goals are weighted equally and a person hits the expected outcome on every one, the formula returns exactly 50. In infants with motor delays, a T-score of 50 has been shown to represent clinically significant change, confirming that “expected” in GAS terms genuinely maps onto real-world improvement.
You don’t need to calculate this by hand. The formula exists mainly so that researchers and program evaluators can compare outcomes across patients or across entire programs in a standardized way, even though every patient’s goals are unique.
Where GAS Is Used
GAS started in mental health but has spread into dozens of fields. In neurorehabilitation, it tracks recovery after stroke, brain injury, and spinal cord injury, where progress is highly individual and standardized tests often miss small but meaningful gains. In pediatric therapy, it measures children’s progress toward specific goals in occupational therapy, physical therapy, and speech therapy. It’s been validated as a reliable outcome measure for children with sensory processing challenges receiving occupational therapy interventions.
In education, GAS has been used to measure student progress toward Individualized Education Program (IEP) objectives. A teacher and student set goals at the start of the school year, define the five levels, and score the outcome at the end. This gives a quantifiable result for goals that might otherwise be described only in narrative terms. Palliative care is another growing area of use, where the goal isn’t recovery but maintaining quality of life, something generic health scales are poorly equipped to capture.
Strengths and Limitations
The biggest strength of GAS is its flexibility. It works for any goal you can define in observable terms, from physical function to social participation to pain management. Because the patient helps set the goals, it increases engagement and ensures the treatment plan reflects what they actually care about. And because outcomes convert to a standardized T-score, you can compare results across very different patients and programs.
The biggest limitation is that same flexibility. Because goals are individualized, two patients with the same condition will have different scales, making direct comparisons less straightforward than with a standardized test. There’s also a risk of bias: if the clinician setting the goals is the same person scoring the outcomes, they may unconsciously set easy targets or score generously. Training helps, and studies have shown acceptable inter-rater reliability when clinicians follow structured protocols, but it remains a concern in settings without formal GAS training.
Writing good scales also takes time and skill. Defining five distinct, observable, evenly spaced levels for each goal is harder than it sounds, especially for goals involving social or emotional functioning where progress is less concrete. Poorly written scales, where the gap between −1 and 0 is much larger than the gap between 0 and +1, for example, undermine the validity of the final score.
Digital Tools for GAS
Historically, GAS was done on paper forms, which made the scoring formula cumbersome and limited real-time tracking. That’s changing. Researchers at Western Michigan University have developed software that implements GAS digitally, allowing clinicians to update goals and criteria from a laptop, tablet, or mobile device. The software provides real-time analytics and statistical reports, so progress can be monitored at any point during treatment rather than only at the end. It also includes communication features between patient and clinician, making it useful for telehealth. The tool is currently in beta testing, but it signals a broader shift toward integrating GAS into electronic health records and therapy platforms.