Can someone explain how to "comment on histograms" and "compare histograms A and B" ?
Cheers
So firstly, a histogram is a great way of displaying univariate, numerical data in a really economical way, when you have more than 40 data values. When looking at histograms, you can gather four summary statistic values:
(a) Shape
(b) Centre
(c) Spread - in terms of range.
(d) Outliers (if any)
Shape:
* The shape of the histogram can be:
- Positively skewed (median < mean)
- Negatively skewed (median > mean)
- Approximately symmetric (median = mean)
Outliers:
Outliers are data values that lie a significant distance away from the main body of data under analysis. You may or may not have any outliers in the data set.
For example: The distribution of 12 pre-schoolers' head circumferences is approximately symmetric, with no apparent outliers.Centre:
The data distribution is centered at the median value.
You find this by calculating
. Sometimes, in a data set you are given, your central point of tendency (i.e. median) will be found somewhere in a particular interval. This is usually the case with continuous, numerical data.
For example: The head circumferences of the preschoolers is centered somewhere in the interval 22 - 26cm, the median value.Spread:
Usually, when we analyse a data set, we look at spread in terms of Range (Max - Min) and IQR (Q3 - Q1). However, because we don't have enough information given to us or displayed on a histogram, we only use the Range as a measure of spread.
For example: The spread of the distribution of preschoolers' head circumferences, as measured by the range, is 12cm (30 cm - 18 cm).So, altogether:
The distribution of head circumferences of preschoolers is approximately symmetric, with no apparent outliers. The distribution of head circumferences is centered at somewhere in the interval 22-26cm, the median value. The spread of the head circumferences, as measured by the range, is 12cm (30cm - 18cm).
_________________________________________________________________________
I think we've established that. Comparing two histograms is also easy
The distribution of head circumferences of preschooler boys is approximately symmetric with no apparent outliers, while the distribution of head circumferences of preeschooler girls is positively skewed with no outliers. The distribution of male preeschoolers' head circumferences is centered at a higher value than the distribution of female preschoolers' head circumferences, 25cm to 20cm, respectively. The spread of the distribution of male preschoolers' head circumferences, as measured by the range, is greater than that of spread of the distribution of female preschoolers' head circumferences, 12cm (30 - 18) to 8 cm (27-17) respectively. Overall, the head circumferences of males tend to be greater and more variable than females' head circumferences in the distribution.