Normative data - FAQs

What are normative percentiles?

Normative percentiles represent the position of an individual's test result within a reference population, indicating how their performance compares to others. They are integrated into VALD Hub to provide practitioners with real-time contextual insights.

How do I interpret the percentile values?

Percentile values indicate the percentage of individuals in the reference population who scored lower than a particular test result. For example, if your result is in the 75th percentile, that means it is higher than for 75 percent of your cohort.

Should the results be considered "good" or "bad"?

The interpretation of results as "good" or "bad" depends on the context and purpose of the assessment. For example, in the case of Countermovement Jump landing force, higher values are generally considered better. However, there are other metrics where lower values or average values are preferred.

Additionally, certain metrics may not have a clear interpretation as good or bad. In VALD Hub, we have made efforts to provide normative percentiles within a relevant context, and we encourage practitioners to use their domain expertise to interpret the results effectively.

How are norms generated?

The norms are generated by analysing hundreds of thousands of tests in VALD's databases. This provides reference values across multiple test types for more than 100 metrics.

But what algorithm was used?

The LMS method was used to construct the percentile curves across age and sex ranges for each test type, sex and metric. This method is well-established in the literature and has been widely used for age-based growth curves, including by the CDC.

Can norms be used to predict injuries?

At VALD, we aim to be as transparent as possible about our data processing methods and never suggest that data from our products can be used to predict injury. Instead, we recommend practitioners use our data in combination with other information about the individual to estimate injury risk and develop an appropriate training intervention (e.g. exercise prescription) to reduce that risk. See VALD's injury prediction statement for more information.

How much data has been used to generate the norms?

There are hundreds of thousands of ForceDecks tests in the normative dataset, ensuring thousands of tests for every test type, sex, and age range. This robust sample size provides statistically significant results.

What happens if there are no tests in a cell?
Example: 68-year-old male performing a hop test

The LMS method produces smoothed percentile curves across the entire age range. This means that all ages are treated along a continuum, so in the above example, the normative values for 68-year-olds would take into account the data from adjacent ages.

How does VALD ensure data quality?

Data quality is ensured through standardized testing protocols, in-app data validation at the point of capture, and additional data cleansing conducted by VALD's data scientists. This process helps to eliminate any anomalous test results or incorrect dates of birth from the normative dataset, ensuring its reliability and accuracy.

Has the normative data been validated?

The norms undergo a validation process to ensure their accuracy and reliability. They are compared with existing literature, and are also reviewed by VALD's team of exercise science professionals.

What is the reference population?

The reference population is the general population aged 11 and above. Notably, test results from VALD's performance clients are excluded because the are not representative of the general population.

Can individuals be identified from the normative data?

No, individuals cannot be identified from the normative data as it is anonymized and aggregated, protecting the privacy and confidentiality of participants. The smoothed curves ensure that it is impossible to infer individual results, even if there is only one test result for a given age and sex.

Why should other organizations benefit from my data?

The benefit to your organisation of accessing rich normative data greatly outweighs the cost of providing it. The normative data set is built from hundreds of thousands of tests uploaded by more than 5,000 organizations. VALD's policy dictates that organizations must consent to including their data in the norms reporting process to gain access to the valuable insights derived from the data set.

How often is the normative data updated?

The norms are periodically reviewed and updated to reflect VALD's growing repository of musculoskeletal testing data.

Can I see normative data for other populations?

We are always extending our normative data to account for other metrics, test types, products, and populations. You can also obtain downloadable and interactive normative data reports in VALD Hub by following this guide.