Descriptive Statistics Flashcards
What do descriptive statistics include?
Average (mean), mode and median, ranger, interquartile range, standard deviation.
What do descriptive statistics measure?
measures of central tendency - a summary of data
The mean - how do you work it out?
- SUM of the values / by how MANY values (N) = the mean
With what data would the mean be used?
in RATIO data with SMALL samples.
Large samples need a computer
What is the mean strongly affected by and what should be used instead?
Extreme values.
You should use the median because it is less senstitive.
The median - how do you work it out?
Put numbers into NUMERICAL order and select central number
If there is an even number of values, you find the mean (or half way point) of two central numbers.
what is the median used for?
Ordinal and ratio data
What are you trying to find with the median?
The MIDDLE number - RANK is important not absolute value (less affected by extreme values)
The mode - how and what are you working out?
The mode is the most frequently occuring value.
NOMINAL data can ONLY use mode out of the desciptive stats.
Mode can be used in ordinal and ratio data as well.
The range - what is this and how do you work it out?
- The range is the numerical value of difference between the maxium and minimum values.
- The HIGHEST value - LOWEST value = the range
- It is a simple way of experessing THE SPREAD OF VALUES
What is the range sensitive too?
EXTREME VALUES. Instead we use the interquartile range.
Interquartile Range - How do you work this out?
Split the data into 4 equal parts (quartiles).
UPPER quartile - LOWER quartile = interquartile range.
What other descriptive statistic is the interquartile range used in assosiation with and why.
The median. They have the same principle strategy:
10 12 12 15 16 19 22 23 28 30 31
^ ^ ^
19 = median
12 + 28 = quartiles. (28-12=16 (the int. range)
write down the pairs of measures of variability you would report together.
interquartile range + median Standard Deviation (SD) + mean
Standard Deviation (SD) - what is this and what data do you use it on?
- Takes into account ALL data points not select ones.
- calculated from VARIANCE (s2), a measure of variability.
- Used in ratio data
- The LARGER the SD the MORE SPREAD OUT the data is.
What symbol can the Sandard Deviation be indicated by?
A greek sigma or s (‘o-‘ a smidge like this)
What can the mean be depicted by?
a mu (‘mew’) like this: u
What does frequency distribution show us?
the SHAPE of the data through histograms
how does normal distribution appear on a histogram?
A bell curve.
What information can be taken from looking at a bell curve?
Normal distribution
The central peak is the mean
What types of skewed distributions can you get?
Negatively skewed (to the right) Possitiveky skewed (to the left) Bimodal (2 peaks or MODES - i.e commonest values)
What discriptive statistics would you report when data are skewed?
interquartile range and the median.
What descriptive statistics would you report when data are normally distributed?
Mean and SD
What statistics would you use on skewed or normal distribution?
skewed - non-parametric)
Normal -parametric (because it assumed the data is normally distributed)