National Physical Therapy Exam (NPTE) Performance Analysis: Physical Therapist (PT) and Physical Therapist Assistant (PTA) Pass Rates and Trends
The National Physical Therapy Examination (NPTE) serves as the benchmark entry-to-practice credential for Physical Therapists (PT) and Physical Therapist Assistants (PTA) across the United States. This project establishes an interactive, data-driven framework analyzing institutional performance, class-size dynamics, regional variances, and pass-rate trajectories across 683 academic programs.
By engineering a consolidated data model in Google BigQuery and designing an interactive Tableau Storyboard, this case study enables educational leaders, accrediting bodies, and health system recruiters to benchmark institutional outcomes, identify top-tier programs, and pinpoint institutions requiring targeted academic resources.
Data was sourced directly from the Federation of State Boards of Physical Therapy (FSBPT) across four distinct reporting datasets:
Two-Year First-Time Pass Rates by School (2024–2025) — PT and PTA
Two-Year Ultimate Pass Rates by School (2022–2023 & 2023–2024) — PT and PTA
Pre-cleaning Steps:
Standardization: Manually populated missing state abbreviations and uniform column naming conventions across all four raw data tables
Data tagging: Created additional column (Program_Type) to explicitly distinguish PT vs. PTA track credentials and facilitate ease of joining all four source tables together later in SQL
NULL Values Handling: Handled redacted values (*) inserted by FSBPT for low-volume cohorts to protect test-taker confidentiality, converting them to clean NULL values for accurate statistical analysis
To formulate a master data-table across academic program type without data duplication, Google BigQuery (SQL) was used to join all four source tables after introductory cleaning via Google Sheets.
Only one query was needed for the purposes of this project, which was written to perform full outer joins within academic program domains (PT and PTA) on ‘School_Code’ to align metrics across multiple years and test outcomes. Additionally, the query focused on cleaning and standardizing text using REGEXP_REPLACE to ensure proper formatting of academic program institutional names. The query can be viewed below:
Query 01_combined_pass_rates_table_creation
CREATE OR REPLACE TABLE `npte-case-study.npte_pass_rates.combined_pass_rates` AS
WITH pt_combined AS (
SELECT
'PT' AS Program_Type,
COALESCE(pt_ultimate.State, pt_first.State) AS State,
COALESCE(pt_ultimate.School_Code, pt_first.School_Code) AS School_Code,
REGEXP_REPLACE(COALESCE(pt_ultimate.School, pt_first.School), r'\s*\((PT|PTA)\)$','') AS School,
pt_ultimate.Weighted_Average_Ultimate_Pass_Rate_2022_2023,
pt_ultimate.Graduates_who_took_NPTE_2022_2023,
pt_ultimate.Weighted_Average_Ultimate_Pass_Rate_2023_2024,
pt_ultimate.Graduates_who_took_NPTE_2023_2024,
pt_ultimate.Test_Takers_From_2022,
pt_ultimate.Test_Takers_From_2023,
pt_ultimate.Test_Takers_From_2024,
pt_first.First_Time_Pass_Rate_2024_2025,
pt_first.Graduates_who_took_NPTE_2024_2025
FROM `npte-case-study.npte_pass_rates.pt_ultimate_pass_rate` AS pt_ultimate
FULL OUTER JOIN `npte-case-study.npte_pass_rates.pt_first_time_pass_rate` AS pt_first
ON pt_ultimate.School_Code = pt_first.School_Code
),
pta_combined AS (
SELECT
'PTA' AS Program_Type,
COALESCE(pta_ultimate.State, pta_first.State) AS State,
COALESCE(pta_ultimate.School_Code, pta_first.School_Code) AS School_Code,
REGEXP_REPLACE(COALESCE(pta_ultimate.School, pta_first.School), r'\s*\((PT|PTA)\)$','') AS School,
pta_ultimate.Weighted_Average_Ultimate_Pass_Rate_2022_2023,
pta_ultimate.Graduates_who_took_NPTE_2022_2023,
pta_ultimate.Weighted_Average_Ultimate_Pass_Rate_2023_2024,
pta_ultimate.Graduates_who_took_NPTE_2023_2024,
pta_ultimate.Test_Takers_From_2022,
pta_ultimate.Test_Takers_From_2023,
pta_ultimate.Test_Takers_From_2024,
pta_first.First_Time_Pass_Rate_2024_2025,
pta_first.Graduates_who_took_NPTE_2024_2025
FROM `npte-case-study.npte_pass_rates.pta_ultimate_pass_rate` AS pta_ultimate
FULL OUTER JOIN `npte-case-study.npte_pass_rates.pta_first_time_pass_rate` AS pta_first
ON pta_ultimate.School_Code = pta_first.School_Code
)
SELECT * FROM pt_combined
UNION ALL
SELECT * FROM pta_combined
ORDER BY State ASC, Program_Type ASC, School_Code ASC
To allow stakeholders dynamic control over metrics without cluttering the UI, several custom parameter-driven calculated fields were implemented:
Universal Parameter Controls:
Dynamic Pass Rate: Dynamically routes between reporting windows (2022–2023 Ultimate, 2023–2024 Ultimate, and 2024–2025 First-Time) based on single-parameter user selection.
CASE [Select Metric and Year]
WHEN '2022-2023 Ultimate Pass Rate' THEN [Weighted Average Ultimate Pass Rate 2022 2023]
WHEN '2023-2024 Ultimate Pass Rate' THEN [Ultimate Pass Rate 2023-2024]
WHEN '2024-2025 First-Time Pass Rate' THEN [First Time Pass Rate 2024-2025]
END
Dynamic Graduate Count: Dynamically updates cohort volume counts matching the active pass-rate metric (Ultimate vs. First-Time pass rates in given year)
CASE [Select Metric and Year]
WHEN '2022-2023 Ultimate Pass Rate' THEN [Graduates who took NPTE 2022 2023]
WHEN '2023-2024 Ultimate Pass Rate' THEN [Graduates who took NPTE 2023 2024]
WHEN '2024-2025 First-Time Pass Rate' THEN [Graduates who took NPTE 2024 2025]
END
Aggregation:
Weighted Pass Rate Field: Calculates state-level averages weighted by graduate count/cohort-volume:
// Weighted Average Pass Rate by State
SUM([Dynamic Pass Rate] * [Dynamic Graduate Count]) / SUM([Dynamic Graduate Count])
Comparative Analytics
Pass Rate Lift: Isolates performance change between most recently established ultimate pass rate and the most recent current first-time pass rate, identifying programs where there is a large gap in ultimate pass rate and first-time pass rate.
[Ultimate Pass Rate 2023-2024] - [First Time Pass Rate 2024-2025]
PT / PTA / Total Program Counts: Uses distinct count logic to count institutions by credential type
PT Program Count
COUNTD(IF [Program Type] = 'PT' THEN [School Code] END)
PTA Program Count
COUNTD(IF [Program Type] = 'PTA' THEN [School Code] END)
Total Program Count
COUNTD([School Code])
User Interface: Features a top-right parameter menu allowing users to toggle between First-Time pass-rate (2024–2025) and ultimate pass rate (2022–2024) metrics. Clicking any state filters the entire storyboard to that region. Hovering over a state reveals total program volume and program track breakdowns.
Key Findings: Ultimate pass rates remain consistently high nationwide, typically nearing or meeting 100%. First-time pass rates exhibit significantly higher regional variation, particularly among Physical Therapist Assistant (PTA) programs.
User Interface: Employs color encoding (Blue for PT, Orange for PTA) and trend lines for regression analysis. Hovering over individual scatter points reveals program-level cohort sizes, pass rates, and location details.
Key Findings: Class size shows minimal linear correlation with ultimate or first-time pass rates. However, cohort volume heavily dictates variance: smaller programs (primarily PTA tracks with 10–40 students) exhibit pronounced pass-rate volatility, whereas larger PT cohorts maintain tight clustering near upper thresholds.
User Interface: Combines box plots (distribution spread) and a ranked dumbbell plot (institutional detail) displaying the recovery gap between initial first-time performance and ultimate pass rates.
Key Findings: Across all institutions, programs achieve an average recovery gap of 13.9 percentage points. PTA programs demonstrate a wider spread and a higher average recovery gap (15.5%) compared to PT programs (11.8%).
User Interface: Features box plots paired with interactive 5% histogram bins. Clicking any bin (e.g., <70% or 100%) filters underlying worksheets to inspect institutional distributions.
Key Findings: Ultimate pass rates are heavily left-skewed, with 40.2% of programs (249 schools) achieving a perfect 100% ultimate pass rate and 86.8% scoring above 90%. Conversely, First-Time Pass Rates follow a standard normal curve centered at 80.9%, with only 4.9% of programs (33 schools) reaching a 100% first-time pass rate.
User Interface: Incorporates a global search bar and column header sorting for Rank, Pass Rate, and Graduate Volume. Dynamically highlights performance metrics based on active parameter selections.
Key Findings: Serves as a granular benchmarking tool across all 683 programs, empowering stakeholders to identify top-tier institutional models alongside programs that may benefit from targeted board prep interventions.
While this project does not evaluate specific academic curricula, it establishes a data-driven framework to identify geographic regions and institutions experiencing difficulty producing first-time and ultimate NPTE passers. It also provides insight to institutions that might be seeing gaps in ultimate pass-rate and subsequent first-time past rates.
Targeted Academic Resource Allocation: Educational administrators and accrediting bodies should consider prioritize academic intervention and board preparation support for programs demonstrating significant first-time pass rate dips (histogram bins <70%).
Specialized Support for PTA Cohorts: Given that PTA programs display higher pass-rate volatility and wider recovery gaps (15.5% vs. 11.8% in PT), state systems should evaluate remediation strategies specifically tailored to associate-level health science curricula.
Benchmarking Successful Programs: Institutional leaders should study structural methodologies from high-volume, high-pass programs to build scalable board preparation frameworks
Data was obtained from the Federation of State Boards of Physical Therapy (FSBPT) using its publicly available school-level NPTE pass-rate reports.
Four source datasets were used:
PT Two-Year Ultimate Pass Rates by School
PTA Two-Year Ultimate Pass Rates by School
PT Two-Year First-Time Pass Rates by School
PTA Two-Year First-Time Pass Rates by School
The analysis includes three reporting measures:
2022–2023 Weighted Average Ultimate Pass Rate
2023–2024 Weighted Average Ultimate Pass Rate
2024–2025 First-Time Pass Rate
FSBPT defines a first-time pass rate as the percentage of graduates who passed the NPTE on their first attempt.
An ultimate pass rate represents the percentage of graduates who eventually passed the NPTE, regardless of how many attempts were required. FSBPT calculates its published multi-year ultimate rates as weighted averages based on the number of graduates testing during each included year.
Because first-time and ultimate rates answer different questions, they are analyzed separately throughout the project.
The four FSBPT datasets were first imported into Google Sheets and preserved as separate raw-data tabs. Corresponding working tabs were then created for cleaning so the source data remained unchanged.
The working workbook contained:
PT Ultimate Pass Rate – Raw
PTA Ultimate Pass Rate – Raw
PT First-Time Pass Rate – Raw
PTA First-Time Pass Rate – Raw
PT Ultimate Pass Rate – Working
PTA Ultimate Pass Rate – Working
PT First-Time Pass Rate – Working
PTA First-Time Pass Rate – Working
In the source reports, a state abbreviation was generally displayed once for a group of programs, with subsequent rows in the same state left blank.
Those missing state values were populated in the working tables so that every program had an explicit state designation. This was necessary for reliable filtering, grouping, and geographic visualization later in Tableau.
Column names were standardized across the four datasets before import into BigQuery.
Examples included:
School_Code
Program_Type
Weighted_Average_Ultimate_Pass_Rate_2022_2023
Graduates_who_took_NPTE_2022_2023
Weighted_Average_Ultimate_Pass_Rate_2023_2024
Graduates_who_took_NPTE_2023_2024
First_Time_Pass_Rate_2024_2025
Graduates_who_took_NPTE_2024_2025
Using consistent field names simplified the SQL join and reduced the need for additional renaming after the datasets were combined.
A Program_Type field was added to each working dataset and populated with either:
PT
PTA
This ensured that the two credential types remained identifiable after they were consolidated into one table.
Some FSBPT records contain suppressed pass-rate values for programs with limited reporting data or small cohorts.
These nonnumeric values were converted to NULL rather than being treated as zeros.
This distinction is important:
0 would represent a measured 0% pass rate.
NULL represents a pass rate that was not available for analysis.
Retaining those values as missing prevents suppressed results from artificially lowering averages or affecting Tableau distributions.
The working datasets contained:
305 PT programs in the ultimate-pass-rate file
377 PTA programs in the ultimate-pass-rate file
300 PT programs in the first-time-pass-rate file
373 PTA programs in the first-time-pass-rate file
The program lists were not perfectly identical between reporting periods. This became an important consideration when the files were combined in SQL.
After spreadsheet preparation, the four working datasets were imported into Google BigQuery.
The goal was to create a single row for each program containing the available ultimate-pass-rate, first-time-pass-rate, cohort-size, state, and program-type information.
A FULL OUTER JOIN was used separately for PT and PTA programs.
This was chosen instead of an INNER JOIN because some programs appeared in one reporting period but not another. An INNER JOIN would have removed programs without a matching record in both tables.
School_Code served as the program-level join key.
The resulting PT and PTA tables were then combined with UNION ALL.
The final table contained 683 programs: 305 PT and 378 PTA programs.
CREATE OR REPLACE TABLE
`npte-case-study.npte_pass_rates.combined_pass_rates` AS
WITH pt_combined AS (
SELECT
'PT' AS Program_Type,
COALESCE(
pt_ultimate.State,
pt_first.State
) AS State,
COALESCE(
pt_ultimate.School_Code,
pt_first.School_Code
) AS School_Code,
REGEXP_REPLACE(
COALESCE(pt_ultimate.School, pt_first.School),
r'\s*\((PT|PTA)\)$',
''
) AS School,
pt_ultimate.Weighted_Average_Ultimate_Pass_Rate_2022_2023,
pt_ultimate.Graduates_who_took_NPTE_2022_2023,
pt_ultimate.Weighted_Average_Ultimate_Pass_Rate_2023_2024,
pt_ultimate.Graduates_who_took_NPTE_2023_2024,
pt_ultimate.Test_Takers_From_2022,
pt_ultimate.Test_Takers_From_2023,
pt_ultimate.Test_Takers_From_2024,
pt_first.First_Time_Pass_Rate_2024_2025,
pt_first.Graduates_who_took_NPTE_2024_2025
FROM
`npte-case-study.npte_pass_rates.pt_ultimate_pass_rate`
AS pt_ultimate
FULL OUTER JOIN
`npte-case-study.npte_pass_rates.pt_first_time_pass_rate`
AS pt_first
ON pt_ultimate.School_Code = pt_first.School_Code
),
pta_combined AS (
SELECT
'PTA' AS Program_Type,
COALESCE(
pta_ultimate.State,
pta_first.State
) AS State,
COALESCE(
pta_ultimate.School_Code,
pta_first.School_Code
) AS School_Code,
REGEXP_REPLACE(
COALESCE(pta_ultimate.School, pta_first.School),
r'\s*\((PT|PTA)\)$',
''
) AS School,
pta_ultimate.Weighted_Average_Ultimate_Pass_Rate_2022_2023,
pta_ultimate.Graduates_who_took_NPTE_2022_2023,
pta_ultimate.Weighted_Average_Ultimate_Pass_Rate_2023_2024,
pta_ultimate.Graduates_who_took_NPTE_2023_2024,
pta_ultimate.Test_Takers_From_2022,
pta_ultimate.Test_Takers_From_2023,
pta_ultimate.Test_Takers_From_2024,
pta_first.First_Time_Pass_Rate_2024_2025,
pta_first.Graduates_who_took_NPTE_2024_2025
FROM
`npte-case-study.npte_pass_rates.pta_ultimate_pass_rate`
AS pta_ultimate
FULL OUTER JOIN
`npte-case-study.npte_pass_rates.pta_first_time_pass_rate`
AS pta_first
ON pta_ultimate.School_Code = pta_first.School_Code
)
SELECT *
FROM pt_combined
UNION ALL
SELECT *
FROM pta_combined
ORDER BY
State ASC,
Program_Type ASC,
School_Code ASC;
FULL OUTER JOIN
Preserved programs that appeared in only one of the available reporting datasets instead of dropping them from the analysis.
COALESCE
Selected the available state, school code, and school name when a program appeared in only one side of the join.
REGEXP_REPLACE
Removed the (PT) and (PTA) suffixes contained in the ultimate-pass-rate source files because Program_Type already stored this information in a separate field.
UNION ALL
Combined the PT and PTA results while retaining every program record.
The consolidated BigQuery output was imported into Tableau Public.
A parameter-driven structure was used so users could switch between multiple pass-rate reporting periods without requiring a separate dashboard for each measure.
The main Tableau parameter allows the user to choose among:
2022–2023 Ultimate Pass Rate
2023–2024 Ultimate Pass Rate
2024–2025 First-Time Pass Rate
This parameter controls both the displayed pass rate and its associated cohort size.
Dynamic Pass Rate
CASE [Select Metric and Year]
WHEN '2022-2023 Ultimate Pass Rate'
THEN [Weighted Average Ultimate Pass Rate 2022 2023]
WHEN '2023-2024 Ultimate Pass Rate'
THEN [Ultimate Pass Rate 2023-2024]
WHEN '2024-2025 First-Time Pass Rate'
THEN [First Time Pass Rate 2024-2025]
END
This field allows the same visualization to display different NPTE measures depending on the user's selection.
Dynamic Graduate Count
CASE [Select Metric and Year]
WHEN '2022-2023 Ultimate Pass Rate'
THEN [Graduates who took NPTE 2022 2023]
WHEN '2023-2024 Ultimate Pass Rate'
THEN [Graduates who took NPTE 2023 2024]
WHEN '2024-2025 First-Time Pass Rate'
THEN [Graduates who took NPTE 2024 2025]
END
This ensures that the cohort size shown in the dashboard corresponds to the selected pass-rate reporting period.
A simple average would give a program with 15 graduates the same influence on a state average as a program with 150 graduates.
To better represent the number of graduates contributing to each state's result, Tableau calculates a graduate-weighted pass rate:
SUM(
[Dynamic Pass Rate] *
[Dynamic Graduate Count]
)
/
SUM([Dynamic Graduate Count])
This allows larger graduating cohorts to contribute proportionally to the state-level result.
COUNTD(
IF [Program Type] = 'PT'
THEN [School Code]
END
)
PTA Program Count
COUNTD(
IF [Program Type] = 'PTA'
THEN [School Code]
END
)
Total Program Count
COUNTD([School Code])
These fields allow the geographic dashboard and tooltips to display the number of PT programs, PTA programs, and total programs represented in each selected area.
The following calculated field compares the most recent ultimate-pass-rate reporting window in the project with the most recent first-time-pass-rate reporting window:
[Ultimate Pass Rate 2023-2024]
-
[First Time Pass Rate 2024-2025]
A positive value indicates that the reported 2023–2024 ultimate pass rate is higher than the reported 2024–2025 first-time pass rate.
This calculation does not represent a true same-cohort recovery rate.
The ultimate metric covers 2023–2024 graduates, while the first-time metric covers 2024–2025 graduates. Although the reporting periods overlap, they do not represent identical cohorts.
The field is therefore used as a descriptive comparison between two reporting windows, not as proof that a specific group of initially unsuccessful candidates later passed the NPTE.
The Tableau project contains five story sections designed to move from national patterns to individual program-level detail.
The geographic view allows users to compare state-level NPTE performance and identify regional patterns.
Interactive U.S. map
Parameter control for reporting metric/year
Graduate-weighted state pass rates
PT and PTA program counts
State-level filtering
Hover tooltips with additional program information
Selecting a state allows users to narrow the analysis to programs within that location.
Ultimate NPTE pass rates are consistently high across much of the country.
First-time pass rates show substantially more variation, particularly among PTA programs.
Because the state values are weighted by the number of graduates testing, states with several large programs are represented differently from states consisting primarily of small cohorts.
This scatterplot examines whether programs with larger graduating cohorts consistently achieve higher pass rates.
Cohort size on one axis
Selected NPTE pass rate on the other
PT and PTA programs shown separately by color
Program-level tooltips
Tableau trend lines
The data do not show a strong linear relationship between cohort size and NPTE pass rate.
PT programs generally occupy the larger-cohort portion of the chart and tend to cluster near higher pass-rate levels.
PTA programs include more small cohorts and display a wider range of observed pass rates.
Small cohorts also produce more percentage volatility because the outcome of a small number of candidates can have a larger effect on the program's overall rate.
For this reason, extreme values from small programs should be interpreted cautiously rather than treated automatically as evidence of exceptionally strong or poor educational performance.
This section compares the 2023–2024 ultimate pass rate with the 2024–2025 first-time pass rate for programs with both values available.
Box plots summarize the distribution for PT and PTA programs, while a program-level dumbbell visualization shows the magnitude of the difference for individual schools.
Across the 619 programs with both measures available, the average difference was:
13.9 percentage points overall
11.8 percentage points for PT programs
15.5 percentage points for PTA programs
PTA programs therefore demonstrated a wider average separation between the two reported measures.
This difference illustrates how much higher ultimate-pass-rate performance can be than first-time performance at the program level.
However, because the two measures come from overlapping but non-identical reporting periods, the difference should not be interpreted as the percentage of candidates who failed once and later passed.
A future version of this project could use matched graduation-year first-time and ultimate data to directly study same-cohort progression.
This section moves beyond averages and shows how programs are distributed across the full range of reported NPTE outcomes.
Interactive histograms and box plots allow users to examine both the center of the distribution and programs at the upper and lower ends.
Among the 619 programs with a reported 2023–2024 ultimate pass rate:
249 programs, or 40.2%, reported a 100% ultimate pass rate
537 programs, or 86.8%, reported an ultimate pass rate of 90% or higher
The overall mean was approximately 95.7%
The distribution is therefore concentrated heavily toward the upper end of the scale.
Among the 667 programs with a reported 2024–2025 first-time pass rate:
The mean first-time pass rate was approximately 80.9%
The median was approximately 83.9%
33 programs, or 4.9%, reported a 100% first-time pass rate
First-time performance shows substantially greater variation than ultimate performance.
This difference is expected in part because ultimate rates allow candidates additional opportunities to pass after an unsuccessful first attempt.
The final section provides the most detailed program-level view.
Rather than requiring users to locate individual schools through the other charts, the leaderboard allows direct searching, sorting, and comparison.
Users can:
Search for a specific institution
Sort programs by pass rate
Compare graduate cohort size
View PT and PTA programs
Change the active pass-rate measure
Compare programs within a state or across the country
The leaderboard gives users a direct way to move from national patterns to individual programs.
For the 2023–2024 reporting period, 86.8% of programs with available data reported ultimate pass rates of at least 90%, and 40.2% reached 100%.
This indicates that most graduates represented in these programs eventually pass the NPTE.
The 2024–2025 first-time pass-rate distribution is substantially wider than the ultimate-pass-rate distribution.
The overall mean was 80.9%, compared with approximately 95.7% for the 2023–2024 ultimate metric.
This illustrates the importance of distinguishing initial exam readiness from eventual licensure-exam success.
PTA programs generally have smaller cohorts and display a wider spread of pass-rate outcomes than PT programs.
They also showed a larger average difference between the two reporting measures examined in this project:
PTA: 15.5 percentage points
PT: 11.8 percentage points
This difference should be used to generate further questions rather than treated as evidence that PTA programs are inherently less effective.
The scatterplots do not show a strong linear relationship between the number of graduates testing and program pass rates.
Large programs can perform either strongly or weakly, and small programs can do the same.
Cohort size is nevertheless important when interpreting percentages because smaller denominators make program-level rates more sensitive to the outcome of individual candidates.
A first-time pass rate reflects whether graduates successfully passed on their first attempt.
An ultimate pass rate reflects whether graduates eventually passed after any number of attempts.
For education programs, both measures provide useful but different perspectives:
First-time pass rate: near-term readiness for the licensure examination
Ultimate pass rate: eventual success reaching the licensure threshold
Neither measure independently describes every aspect of program quality.
The dashboard is most useful as a screening, benchmarking, and exploratory tool.
Program directors or education leaders could use it to:
Compare their program with national or state-level patterns
Identify changes in pass-rate performance worth reviewing
Examine first-time and ultimate outcomes together
Identify programs that consistently perform well for potential benchmarking
Explore whether unusually high or low rates may partly reflect small cohort sizes
Generate questions for deeper internal analysis
A low pass rate should not, by itself, be interpreted as proof of poor instruction.
Likewise, a high NPTE pass rate does not establish that a program performs strongly on every educational outcome.
Programs with lower first-time pass rates may warrant additional review, particularly when similar patterns occur across several reporting periods.
Internal review could examine areas such as:
Board-exam preparation strategies
Timing of NPTE preparation
Student academic support
Remediation processes
Curriculum-to-exam alignment
Differences between first-time and eventual exam success
A single low reporting period, particularly in a small cohort, should not be treated as sufficient evidence of a program-level problem.
PT and PTA programs differ substantially in typical cohort size and show different pass-rate distributions in this dataset.
Separating the two program types provides more appropriate comparisons than treating all physical therapy education programs as one uniform population.
A 20-percentage-point change in a program with a small cohort may represent only a few additional candidates passing or failing.
Program-level rates should therefore be interpreted alongside the number of graduates contributing to each measure.
Programs with consistently strong first-time and ultimate outcomes may provide useful candidates for further study.
However, this dataset does not contain information about curriculum structure, admissions criteria, faculty resources, student characteristics, board-preparation programs, or other factors that could explain performance.
Additional research would be required before identifying specific educational practices as causes of higher pass rates.
The most recent ultimate and first-time metrics used in this project cover different two-year periods:
Ultimate: 2023–2024
First-Time: 2024–2025
Their difference is therefore a cross-period comparison rather than a same-cohort measure of retesting success.
The FSBPT data are reported at the academic-program level.
The dataset does not include individual candidate characteristics such as:
Academic performance
NPTE preparation methods
Number of attempts
Demographics
Prior educational background
Personal circumstances
Timing of exam attempts
Individual-level explanations therefore cannot be evaluated.
Some results are suppressed or unavailable for programs with limited reporting information or small cohorts.
These values were maintained as NULL rather than converted to zeros.
As a result, the number of programs included in individual calculations varies depending on the selected metric.
Percentages from small cohorts can change substantially based on the outcomes of only a few candidates.
This is especially important when interpreting PTA programs, many of which have smaller graduating cohorts than PT programs.
NPTE performance is an important education outcome, but it does not capture every component of program effectiveness.
The analysis does not evaluate:
Graduation rates
Student retention
Clinical education quality
Employment outcomes
Tuition or financial burden
Faculty characteristics
Curriculum structure
Student satisfaction
FSBPT also cautions users against treating differences in first-time NPTE pass rates as direct measures of instructional quality.
This project identifies patterns and associations in publicly reported data.
It does not establish that cohort size, geography, program type, or any other observed characteristic causes changes in NPTE performance.
Several additions could expand this project:
Match first-time and ultimate rates for identical graduation cohorts to directly measure eventual improvement after initial attempts
Analyze multiple consecutive reporting periods to identify persistent versus temporary performance changes
Add CAPTE or program-characteristic data where appropriate
Compare results by region rather than only individual states
Examine whether program size is associated with the variability of pass rates over time
Compare PT and PTA trends using separate statistical models
Evaluate changes in both pass rates and cohort sizes longitudinally
Explore other publicly available educational outcomes alongside NPTE results
This project converts several separate FSBPT NPTE reporting files into a single interactive analysis of 683 PT and PTA education programs.
The data show a clear distinction between first-time and eventual NPTE success: ultimate pass rates are concentrated near the top of the scale, while first-time pass rates show substantially more variation. PTA programs also display greater variability than PT programs, while cohort size alone does not show a strong linear relationship with pass-rate performance.
The dashboard is best used to identify patterns, compare programs, and generate questions for deeper reviewrather than to label individual schools as high- or low-quality based solely on NPTE pass rates.