Summerizing Data Flashcards
Two major types of data
Quantitative
- numerical data that does NOT describe a cause and effect or comparison
Qualitative
- non-numerical data that is described in words or letters and often presents a comparison/ratio
5 scales of data ranked in order of most to least information content
Continuous ratio
- number scale that has a true zero as a possible answer
- highest information content
Continuous interval
- number scale (such as temp) with NON-true zero
Ordinal data
- scales from 1: whatever
- ex: on a scale from 1-6 what is the pain?
Binary data
- yes or no and direct percentages
Nominal data
- just a value w/ no correlation
- a name/ a blood type/ a country/ etc.
- lowest information content
Continuous data explained
Generally described w/ an estimate of the shape and center of the distribution and variability of the data
- always produces an average however the average value alone is not enough
Three common estimates of the typical patient (estimates of the average)
1) mean
2) median
3) mode
Is the mean the most important statistic?
No not always, it depends on the distribution of the data (or shape of the data)
- can use for symmetric data curves
Right or left skewed shaped histogram uses what estimate of center and variability
Median and interquartile range
In right skewed data, the mean is generally larger than the median and the mode is less than the median
- mean > median > mode
In left skewed data, the mode is generally larger than the median and the mean is less than the median
- mode > median > mean
- note that the data is named for where the tail is!*
Bimodal shaped histogram uses what estimate of center and variability
Mode and interquartile range
Symmetrical shaped histogram uses what estimate of center and variability
Mean and standard deviation
Standard deviation
Is the average deviation from the mean
Normally distributed symmetric shaped statistics information
1) is symmetric
2) centered at the mean
3) 3 standard deviations will encompass 99.7% of the data
- 1 deviation = 68%
- 2 deviations = 95%
- 3 deviations = 99.7 %
4) anything not in the 3 standard deviations are outliers
What do you use to report ordinal (scaled) data to patients
Same as the continuous rotation (based off the shape of the data)
However if there are not enough unique levels (usually <5 ) just use median and IQR
What is variance?
Standard deviation squared
- used to calculate data but is almost never reported