Data Type |
Description |
Discrete/Continuous |
Qualitative/Quantitative |
Non-Parametric/Parametric |
Nominal |
categories without order (eye colour, marital status) |
Discrete |
Qualitative |
N |
Ordinal |
Ordered categories (Ficat classes) |
Discrete |
Qualitative |
N |
Integer |
Number of counts (papers) |
Discrete |
Quantitative |
P/N |
Ratio |
value independent of units |
Continuous |
Quantitative |
P/N |
Interval |
Distances betw. units are known (hours spent studying) |
Continuous |
Quantitative |
P/N |
Parametric Test = assume that the data were sampled from a particular form of distribution, such as a normal distribution; Non-Parametric = makes no such assumption.
Always plot data. If data is presented always ask if it has been tested for normality, skewness & kurtosis (to determine need for parametric or non-parametric analyses)
Measures of Central Tendency:
- Mean - average
- Median - central value; use for ordinal data
- Mode - value with most frequency; use for nominal data
Normal (Gaussian) Distribution:
- Continuous symmetric distribution that follows the familiar bell-shaped curve
- The distribution is uniquely determined by its mean and variance
- Allows for Parametric tests (which are more powerful)
- Rarely occurs in Orthopaedics
- Skewed distribution:
- Non-symmetrical, with a tail
- easily seen by plotting data
- use median or mode rather than mean
- if in doubt assume non-normality & use non-parametric tests.
- Kurtosis:
- is a measure of the heaviness of the tails in a distribution, relative to the normal distribution.
- A distribution with negative kurtosis is light-tailed relative to the normal distribution
- A distribution with positive kurtosis is heavy-tailed relative to the normal distribution