Evaluating Fitness Tests Flashcards
(10 cards)
1
Q
validity and validity coefficient
A
- ability of test to measure accurately, a specific physical fitness component compared to gold standard
- r = validity coefficient, relation between predicted score and reference score, the closer to 1 the greater the validity, r>0.8 is more acc
2
Q
line of best fit
A
regression line for relation between measured and predicted scores
3
Q
standard error of the estimate (SEE)
A
how far away the predicted datapoints are from the line of best fit, close to line lower error
4
Q
sensitivity
A
probability of correctly id individuals with risk (low false neg)
5
Q
specificity
A
probability of correctly id individuals without risk (low false pos)
6
Q
reliability
A
consistent and stable over time, can’t exceed 1; reliability affect validity since unreliable tests don’t produce consistent results, can have reliability w/o validity
7
Q
objectivity
A
intertester reliability, can’t exceed 1
8
Q
considerations when using prediction equations
A
- is ref measure the gold standard
- larger samples are more valid
- less variables more accurate
- what is the total error (average of individual diff)
- id how good equation is at estimating individual ref value using Bland-Altman plot
9
Q
line of id and total/prediction error
A
- Line of identity assumes perfect match (r=1)
- check if individual data points lie around the line equally with low total error for valid test, if the points mostly lie above then predictive equation overestimates, if points mostly under the line of id then predictive equation underestimates
10
Q
Bland-Altman plot
A
- compares the differences in score to the average values of reference and the actual value
- If difference scores are negative then the reference value is lower than the predicted, overestimating the actual value
- Narrow intervals indicate reliable tests since low variability
- Closer to the line more accurate
