Tests Utilized in Choosing to Report Average Prices
The following tests were applied to price data where 10 or more but 25 or less prices were obtained for a given product:
- Coefficient of Variation (CV) test:
Where less than 16.5% prices are reported without an asterisk.
- Those with a coefficient of variation between 16.6% and 33.3% are reported with an asterisk. Caution is to be used in reporting on these prices because of higher levels of sample variance.
- Where CV was greater than 33.3% data was not reported.
- Skewness test: the absolute value Hildebrand measure was used:
Absolute (estimated. mean — estimated median divided by estimated. standard deviation).
- where less than .2, no asterisk is added.
- where equal to .2 but less than .4 an asterisk is added. This indicates that the data has a moderate skew.
- Where the measure is greater than .4, the statistic is not reported.
Note: using the Pearsonian measure of skewness and values of absolute 1 and 2 yield similar results.
- Kurtosis Test:
- Where between -2 and 2, no asterisk is added.
- Where greater than 2 but less than 4 an asterisk is added indicating that the distribution is flatter than a normal distribution.
- Where greater than 4, the mean is not reported, since there is not a high degree of "central tendency" in the distribution.
Tests were applied to the price distributions of the following products:
- Text Telephone Device (with display only)
- Vibro Tactile Full Function Alarm
- Shoulder-Elbow-Wrist-Hand-Finger Orthosis
- All Prostheses
- Scooter/Wheelchair Acid Battery
- Porch Lift
- Chair Lift
- Elevator
- Bi-level CPAP Device
- Glucometer Test Strips
Because of the nature of the data collected, the tests could not be applied to van/automobile conversion data or data concerned supports for the blind/persons with severe sight impairment.