Fitness testing Flashcards
What is fitness & why do we test
- Fitness = range from peak condition - injured
- Highlight strengths/weaknesses, develop profiles, evaluate training, talent ID, motivation, monitor rehab
Lab based data collection pros/cons
Pros - high levels of interval validity, control all variables, only measure relevant aspect
Cons - low ecological validity, expensive equipment
Field based data collection pros/cons
Pros - mimics perf environment = ecological validity ^, cheaper, more accessible
Cons - Can’t control all variables, difficult to claim interval validity
What makes a good test
- Validity
- Reliability
- Objectivity
- Specificity
Types of Validity
- Content validity (extent an assessment represent all facets of tasks within the assessed domain)
- Criterion-related validity (extent to which the output of test correlates to the criterion measure (the gold standard))
- Concurrent validity (relating test results of criterion test to different test shortly after)
Types of Reliability
- Repeated research would produce same results
- Test-retest reliability (same tester & individual & same results on number of occasions)
- Inter-rater reliability (ability for multiple researchers to produce same results with same individual)
Observed score
True score (actual perf) + Error score (other factors eg wind speed)
Improving test validity/reliability - Performers
- No training on the day
- Familiarisation
- No smoking/drinking on day
- Similar footwear/clothing each test
Improving test validity/reliability - Test admin
- Same environmental conditions
- Same social conditions
- Same protocol, warm-up, technique
- Practice test trials
Improving test validity/reliability - Instrumentation
- Ensure familiarity w/ test equipment
- calibrate test instruments
- Have spare batteries/equipment ect
Improving test validity/reliability - Test delivery
- Know how to measure each test
- Unbiased scoring
- Report scores to precision of the whole test procedure not just equipment
Interpreting testing errors
- Coefficient of variance (CV) (Observed score doesn’t always = true score (score is in a range dependant on reliability))
- Confidence level predicts chances of incorrect interpretation
- Objectivity (is data collection independent of persons involved in data collection - eliminate human error, defined & strictly enforced test finishing criteria