Abstract:
Field tests are often the most practical method to assess aerobic fitness, but
demonstrate greater error variability than laboratory tests. Researchers can improve
field tests by identifying factors that contribute to systematic error in VO₂max
estimation. PURPOSE: To examine the validity and reliability evidence of two field
tests of aerobic fitness: Cooper's 12-minute run (12MR) and the multistage shuttle run
(MSR). METHODS: Sixty participants (mean age = 21.8 ± 3.6y), completed three
trials (occasion) of each field test (instrument) for a total of six test trials. To estimate
overall reliability and evaluate possible sources of error in the field tests, a
psychometric statistical tool called generalizability study (G-study) was employed.
This analysis utilized a two-random facet design (occasion and instrument) in a
completely crossed ANOVA. In addition, criterion VO₂max was assessed in a subgroup
of volunteers (n = 21) via an incremental treadmill run and expired gas analysis (TR).
Each participant completed the study within a six-week period. RESULTS: G-study
analysis of the two field tests returned a high reliability coefficient (φ = 0.96), with the
largest amount of systematic error variance (4.3%) attributable to an interaction
between participants and test occasions. This mild interaction suggests certain test
participants demonstrated larger error variability across test occasions than other
participants. The MSR predicted VO₂max values lower than those measured in the
laboratory setting (p < 0.01; paired t-tests), while 12MR and TR scores were not
different (p > 0.05). The 12MR underestimated VO₂max values at lower aerobic fitness
levels and overestimated VO₂max values in individuals demonstrating greater aerobic
fitness, which was not observed in the MSR data. CONCLUSIONS: These results
suggest high reliability for VO₂max field tests in young, healthy individuals. However,
test administrators must use caution when attempting to use field test data to predict
criterion VO₂max scores. While test participants can be expected to attain MSR scores
significantly lower than the criterion value, the consistent mean bias across VO₂max
values makes the MSR a more useful test when comparing test participants.
