Hawkin Dynamics’ Perspective on Normative Data

What is normative data and why is it useful? 

A simple way of remembering the key applications of force assessment is through the acronym “BRIEF” (Figure 1). Although it is not an exhaustive list, the BRIEF lists benchmarking and talent identification, return to sport, injury risk profiling, exercise biofeedback, and fatigue and readiness monitoring as being considered among the primary applications of force assessment involving various sports populations. As can be also seen in Figure 1, the ability to effectively apply force assessments across the BRIEF spectrum depends on the knowledge of quality baseline normative data, specific to the athlete population, which includes understanding or quantifying the scale of typical scores for key assessment metrics (data distribution) and their measurement error (reliability bandwidth). This allows the practitioner to more accurately interpret individual athlete’s assessment scores within a given session and monitor how they change between sessions.  

Figure 1. The BRIEF framework, highlights primary applications of force assessments in sport. 

It is unsurprising, therefore, that we often get asked about normative data for a range of force plate test metrics. Normative data for each test metric of interest technically should be taken from a large enough sample to ensure it truly represents the full range of scores that can be obtained by the population group. Some research articles on this topic suggest 50-75 participants (i.e., athletes, when applied to sport) are needed for accurate normative data compilation whereas others suggest at least 85 participants to generate stable means and standard deviations regardless of the level of skewness that might be present in the dataset (Piovesana and Senior, 2016). In other words, adding more than 85 participants’ worth of data to the dataset often does not change the associated mean and standard deviation values and so a decision needs to be made as to whether adding more participants’ data is worth it from a cost-benefit perspective.  

The reason it is important to generate stable means and standard deviations for each test metric is because these are used to create Z-scores or T-scores which are the most common form of standardized scores used in sport science for creating benchmarks (see a short video explanation of these standardized scores here). They can also be used to compare an athlete’s present data to their historical data as part of ongoing monitoring. An example of a single squad versus multiple (four) squads of athletes from the same league included in a dataset on the total data distribution (including mean and standard deviation) and calculated T-scores is shown in Figure 2. As you can see, using multiple squads’ worth of athlete data filled the gaps in the data illustrated when only a single squad’s worth of athlete data was included. Think of this as increasing the number of pixels that a camera can capture on the accuracy of the images it produces. In addition to being useful for key output metrics like jump height (as used in the example shown in Figure 1), normative data can also be used to create meaningful bilateral force asymmetry cut-offs, as explained here.  

Figure 2. Importance of normative data sample size on stable data distribution and T-scores.  

So, how can practitioners source normative data that is relevant to their athletes? Well, there are two main approaches to this that we encourage practitioners to consider. Firstly, normative data that has been published in scientific journals can be valuable in this regard. Hawkin Dynamics force plates have been used in several recent studies from which normative or descriptive data for specific sports populations can be sourced (e.g., Berberet et al., 2024; Soriano et al., 2024). Alternatively, practitioners can compile their own normative data over time. Even if practitioners work with small groups of athletes, patiently accruing data over time, such as across successive seasons, particularly around new athlete intake periods and at matched time points (e.g., pre-season versus in-season), will allow them to understand what is typical for their athletes. Importantly, the do-it-yourself approach to compiling normative data will best reflect the practitioners’ and athletes’ unique environment and culture within which they operate and, therefore, will likely be the best source of normative data. We know the latter approach takes some time, but it is a worthwhile investment for those who take force assessment seriously. It can also be done quickly if practitioners collaborate with other trusted practitioners who work with equivalent population groups (see a soccer collaboration example here).  

Important advice on accessing or producing normative data 

An understanding of the involved testing procedures is of paramount importance when it comes to accessing or producing normative data. Practitioners should critically evaluate the source of any normative data that they plan to compare their own athletes’ data to, to avoid potentially erroneous decision-making on athlete training interventions. Poorly sourced normative data can not only lead to wasted time for staff and athletes but also sub-optimal training programs that inadequately prepare athletes for the physical demands of competition. Below is some important advice for practitioners. 

1. Understanding Test Methodology Quality 
   It’s essential to understand how tests are conducted, including details such as the hardware (including force plate size, model generation etc.) and software used, the verbal cues given, athlete familiarization with the protocol, the testing surface, the zeroing process, and the methods applied to name some. In case you didn’t know, some force measurement software providers offer practitioners different calculation options that can affect data consistency, which creates challenges when compiling normative data. Variables such as weighing duration, sample frequency, filtering methods, and event thresholds (e.g., the onset of force production or reduction, take-off, and landing in vertical jumps) need to be standardized to ensure data integrity. This is why Hawkin Dynamics provides only the best data analysis options and doesn’t allow methods to be altered by users
2. Avoiding Duplicate Athlete Data 
   Normative data must only include unique athlete entries. If the same athlete’s data is entered multiple times into a dataset, it can skew the calculated means and standard deviations (from which the Z-scores and T-scores are derived). For example, an experienced athlete’s vertical jump height may not vary significantly over time. If data is populated based on repeated samples recorded for the same athlete rather than unique athletes, it risks over-representing individuals, thus compromising the representativeness of the dataset for a specific sport or competitive level.  
   
3. Representativeness of Normative Data 
   Normative data should accurately reflect the population it is supposed to be representing. This means accounting for factors like sport, sex, position, level of competition, and, where relevant, age or biological maturity (e.g., through bio-banding). Ensuring the dataset is representative of the population is crucial to making meaningful comparisons and anybody who uses the normative data should have full access to this information on demand. If the population information cannot be accessed, then the result can be akin to comparing apples to oranges. 
4. Ethical and Compliant Data Collection 
   Ethical considerations are paramount in any data collection planning. Data must be collected under formal agreements, with individuals directly or whole organizations (or both), ensuring compliance with data security and ethical standards. For example, Hawkin Dynamics does not access or use our customers’ data without explicit written consent. Our approach to normative data compilation is rooted in formal agreements, often established through direct collaborations with organizations or sponsorships of PhD students. These agreements ensure that de-identified data is collected and used responsibly and ethically, with a complete understanding of points 1-3 above. 
5. Transparency and Diligence 
   Our customers and users can trust that the normative data we share has been collected under rigorous, vetted standard operating procedures. Our science and innovation team at Hawkin Dynamics ensures every step adheres to high scientific and ethical standards. The population, methodology, and data collection process are fully documented, ensuring our normative datasets are robust, representative, and reliable. Practitioners should be empowered to ask their sports technology providers how any normative data that they share has been sourced and processed and receive a full and open response.  

Normative Data – A Customer Example 

Hawkin Dynamics wireless dual force plates are being used in a landmark study across the British Army, led by the Defence Medical Rehabilitation Centre (DMRC) and supported by Hawkin Dynamics’ UK team. This research is referred to as “The STRONG Study”. It has been commissioned to address one of the UK Defence Medical Services’ key strategic aims of preventing and treating musculoskeletal injury (MSKI) among military personnel (Ladlow et al., 2023). The overarching aim of the study is to assess around 2000 active British Army members, from a range of ground close combat (GCC) and non-GCC roles, on both the isometric mid-thigh pull (IMTP) and countermovement jump (CMJ) using Hawkin Dynamics’ dual force plates across multiple military bases around the UK. The study will provide the British Army with accurate IMTP and CMJ normative data according to job role which will better equip the DMRC staff who are responsible for expediently returning injured British Army to full duty. Specifically, DMRC will have better insight into what a typical score is for select IMTP and CMJ metrics for members of a range of regiments across the British Army which they can then use to set benchmarks to 1) identify if and for which “groups” strength training interventions are required to help reduce MSKI risk and 2) more effectively guide the return of personnel with MSKI back to active duty. 

Summary 

Good quality normative data is very useful to practitioners. We strongly encourage our customers to critically evaluate any normative data that they use as part of their decision making across the BRIEF applications of force assessments (Figure 1). Customers should ideally use normative data from trusted sources only, including from peer-reviewed studies (although, these studies should still be critically evaluated, of course) and from their own specific data collection that they know has been collected by adhering to strict standard operating procedures. We don't own our customers’ data, which is why we don’t use it to provide normative datasets without an agreement in place and the required supporting information. Having an agreement in place and conducting a full audit of data collection processes and athlete population characteristics overcomes the inherent but important blind spots that arise from using data collected within different contexts to create normative datasets. It becomes even more complex when compiling normative datasets involving injured athlete populations, as even more information about the population characteristics is needed for it to be useful, but that’s beyond the scope of this piece.   

I hope this short article has got you thinking about normative data, from how it is sourced to how it is applied. If you want to learn more about normative data at Hawkin Dynamics, please reach out to me directly at john.mcmahon@hawkindynamics.com.

Hear more from Dr. John McMahon on normative data, check out Strength Coach Network's podcast: 

 YOUR Team's Data is the ONLY Data That REALLY Matters