Abernethy, P., Wilson, G., & Logan, P. (1995). Strength and power assessment. Sports Medicine, 19, 401-417.

This review article considers the factors associated with strength and power testing. It highlights shortcomings and problems associated with this form of assessment. An underlying theme that is verified by its exhaustive review is that much strength and power testing is invalid and unreliable. The reasons for this status are primarily a failure to scientifically develop testing protocols and validities and then to cross-validate the utility of tests with each sport and level of participant.

Generally, there is little consistency between laboratories in terms of the rationale for, or execution of, strength and power assessments. This means that results will be dependent upon who does the testing. The tenuousness of those results is compounded further when unsubstantiated validity, reliability, and accuracy are considered.

Strength and power are assessed for four main purposes:

• to quantify the relative significance and contribution of strength and power to various athletic events;
• to identify specific deficiencies and prescribe corrective programs to produce appropriate changes (diagnosis);
• to identify talented individuals who may be suited to particular sporting activities; and
• to monitor the effect of various training and rehabilitation interventions.

Current methods of dynamometry can measure power associated with tasks in which either the load or movement velocity are held constant. This contrasts with the dynamic nature of sporting activities. This discrepancy between sports and laboratory activities is one of the major challenges for sport scientists.

Testing is further confounded when the type of contractions, movement paths, and nature of the testing activity are at variance with what occurs in the sporting situation. These problems create difficulties in mimicking sporting movements. Consequently, many testing protocols are limited to nonspecific tasks. The difficulty presented with "general" tests is their lack of sensitivity to the requirements of and changes in specific sporting activities. Important movement, capacity, and adaptive subtleties are likely to be lost in nonspecific tests.

The failure to address these problems because of expediency and the "need" for testing does not make testing any more valuable or useful. It serves to perpetuate problems and increases the likelihood of inappropriate information being used to make coaching decisions. The current trend to perform testing for testing sake is the symptom of this expediency.

There are three modes of dynamometry:

Isometric dynamometry measures the amount of strength that can be exerted against an immovable object (MVC - maximal voluntary contraction). The important characteristic of isometric contractions is the rapidity with which maximum force can be developed (RFD - rate of force development). Proponents of this form of testing assert there is a high level of assessment control. Detractors assert that isometric tests bear little semblance to the dynamic nature of sporting tasks.

Isometric assessments usually display high test-retest reliabilities. However, reliability varies between the muscle groups and the parameter (RFD or MVC) being assessed. Different tests are appropriate for athletes, nonathletes, women, men, and children. Differences in the posture of actions also change reliabilities.

To achieve some potential for acceptable reliability, it is suggested that:

• actions be performed as hard and fast as possible,
• the testing posture mimic that of the sport (since strength is context specific), and
• tests be performed at the joint angle associated with peak force during the target athletic movement.

Isoinertial dynamometry involves natural activities. Previously this was called isotonic (constant tension) work, but the nature of dynamic work is hardly that and so the "new" label is more appropriate. Isoinertial means constantly resistant to motion, and in most activities resistances such as gravity, water, air, and equipment are always impeding progress and performance. Maximum successful exertion for one repetition of a task (1RM) is the usual measure of maximum isoinertial strength and is used commonly in sport profiling.

However, isoinertial movements also bear little resemblance to sporting activities. They are usually performed relatively slowly and are trained by a low number of repetitions. This nonspecific measure and training mode is questionable for its relevance and transfer value to actual sporting movements. Some argue that the dynamic accelerative motion associated with these tests more closely approximates movements in many sports. Those who argue against this form of assessment emphasize the potential for athlete injury and poor reliability and objectivity due to intertrial, interathlete, and interlaboratory variations. Much of the gains in these forms of tests results from learning to do the movements of the tests rather than reflecting training effects from sporting participation.

There may be a load threshold beyond which reductions in reliability of this form of movement seriously compromise its validity. For example, several acute variables, including preloading and recovery between 1RM efforts, may affect 1RM strength. The number of trials to achieve maximum and the type of movement will also affect reliability. Further, it is also very possible that maximum values of exertion may not be the best measure of a strength and power capacity in a number of sports.

Isokinetic dynamometry is perceived to be the easiest assessment over which to achieve reliability and objectivity. However, the movement loading pattern bears little relationship to most sporting activities. That weakness makes inferences from results to actual activity tenuous at best.

Reliability is affected by the selected speed of the movement. Usually fewer trials are needed for slow movements while more are needed for fast movements. The order of testing also influences results. It has been shown that reliability of eccentric and concentric assessments is reduced when tests for highest speed precede those of lower speeds. Further, results for one joint should not be considered characteristic of other untested joints. It is incorrect to infer strength status between joints and movements, particularly in highly trained athletes. As with other forms of strength assessment, test positioning and stabilization affect reliability.

With all forms of dynamometry, there are weaknesses that often make results unreliable and therefore, unrelated to sporting pursuits. Poor generalization, particularly in light of strong support for the specificity of training effect, is further compounded by non-standardized and poorly constructed testing protocols. The validity, reliability, and objectivity of strength and power testing procedures has to be established before any credence can be placed in results so that better coaching decisions can be made.

Most assessment procedures offered by sports science services display discrepancies in the logic and practices of their proponents because they are generally governed by belief, preference and/or expediency, rather than objectively verified fact and practice. As such, strength and power testing rarely deserves the status of acceptable sports "science." The following are worthy of note.

• It is unlikely that any one procedure will be able to be used for several major purposes (e.g., diagnosis, talent identification, training effects).
• The usefulness of particular protocols within a sport may differ with the level of athlete targeted for testing.
• Minor changes in protocols are likely to greatly affect the usefulness of a test in a particular context.
• Procedures influenced by assumptions and approaches based upon anecdotal evidence should be questioned.

1. Factors which modify strength and power test results.

• The time of testing after previous competitions or intensive training bouts will influence results. Strength (1RM) can take up to 3 days to recover from similar intensive stimulation, such as that performed by athletes in regular training.
• Strength, particularly isokinetic strength, and power are reduced by exposure to single bouts of intensive endurance activity. Continued endurance training in the absence of strength and power work reduces those capacities even further. It is recommended that prior to assessment, the physical activity of athletes be limited, monitored, and controlled for 72 hours.
• Any strength or power assessment should be substantiated by established validity, reliability, and objectivity demonstrations rather than being assumed or expediently argued as self-evident truths.

2. Correlations between test measures and athletic performance.

• There is only a limited amount of published research relating strength and power assessments to athletic performance.
• Correlations may exist between tests and performances but that does not mean the tests will be able to reflect training effects. For example, power may be shown to relate to sprint swimming performance, but changes in performance speed usually are not reflected in strength/power test results.

3. Discriminations between groups of individuals.

• Generally strength and power assessments demonstrate correlated differences between athletes across a wide range of standards. Also, men usually test stronger than women as do the mature over the young, and the active over the inactive.
• Discriminability is quite weak when groups of like individuals are tested. Inferring any strength/power differences to performance is spurious. For example, the strength of football players is unrelated to playing status and ability.

4. Sensitivity of testing to training effects.

• There is no consistent finding that dynamometric measures are sensitive to the effects of various types of training nor whether all measures are similarly sensitive to training effects. It would be an error to place much credence on one study that supports a discriminatory finding.
• An important question to ask of supposedly training-sensitive strength and power measures is whether existing correlations between tests and performance demonstrate a correlation of a similar or increased magnitude following training or detraining. A marked reduction in correlation would suggest that what was being measured initially has changed and become more unrelated.
• Increments in strength and power following training may be limited to specific movement patterns, contractile speeds, and loading modalities. It is likely that in many studies and practical situations involving seriously training athletes the only power and strength improvements that occur will be in the actual activities used for training. There will be no carry over to a targeted performance activity.
• It is erroneous to believe that all strength and power assessments will be sensitive to the adaptations associated with all forms of strength and power conditioning.
• It is frequently assumed, but more likely unknown, that strength and power measures reflect training changes established in other activities through other modalities.

5. Assumptions of strength and power assessment.

• About the only fact that can be stated accurately about strength and power assessment procedures is that they assess the performance used in their conduct.
• For the large majority of strength and power tests, it is difficult to verify that the mechanisms being tested are in fact being measured.
• In situations of practical expediency, the issues of logical validity and models for the development of strength and power are often confused. A distinction must be made between the two to understand what assessments actually do.

Implications. Practical situations often determine the limits of testing and training. Assessments quite often are restricted to available equipment, whether or not the modality of testing is appropriate for a sport. When this situation exists, a coach or athlete should ask for objective verification of claims of benefit before participating in testing or training. Given the state of scientific knowledge and demonstrations in this field, it would be more prudent for a coach to graciously refuse to accept assistance from "strength and power trainers" if no substantiating evidence can be provided.

Until this area of sports conditioning and assessment is clarified by acceptable research, its claims must be viewed with skepticism. There are too many advocates making too many uncorroborated claims about the value of inferring strength and power assessment results to program presumptions and consequential athlete benefits.

Return to Table of Contents for this issue.