- Created by: Kathryn Swarbrick
- Created on: 31-01-12 19:22
- Shows the median between 2 sets of data.
- Good with dealing with skewed data.
- You can decide the boundaries of the 2 groups.
- More appropriate when the data sets are independent of each other.
- More appropriate when both sets of data have the same shape distribution.
- Have to have equal sample sizes.
- Shows the spread from the mean.
- Very visual.
- Gives an indication of the reliability of the data.
- Can work out mean, range, mode, median, lower quartile, upper quartile and interquartile range.
- Can compare graphs easily using the above ^ for analysis.
- Anomolies can be shown.
- Can work out standard deviation.
- Works better with lots of data.
- The standard deviation can easily be manipulated and can be bias.
- Can compare multiple sets of data.
- Lots of data can be put on one graph.
- Individual variables within the diagram can be compared.
- No stats tests can be linked to them.
- Hard to spot anomolies.
- Hard to make a scale suitable.
- Clear and easy to interpret.
- Shows changes over distance.
- Shows density and distribution of variables.
- Not all data can be represented by these charts.
- Time consuming to plot by hand.
Proportional Pie Charts
- Allow fractional and percentage comparison.
- Display approximate proportions of variables throughout the area taken up by the pie chart.
- Visual- can see a general trend.
- Can't use for exact comparisons.
- Impossible to extract specific data.
- Can't represent more than one point at a time.
- May not always be acurate (especially if plotted by hand).
- Overlaps can cause issues if they are used on maps.
Line Graphs and Bar Charts
- Little background knowlegde would be needed to understand the graphs.
- Comparisons can easily be made with other similar graphs or more than one line/chart can be plotted on one graph.
- Anomolies are quite clear.
- Give visual image- shows the general trend/correlation (giving basis for stats test to analyse.)
- Can plot the standard deviation.
- Bar Charts show cumulative data/discrete data which is common so they can be used for many purposes.
- Line Graphs use continuous data which is also common.
- Can be tedious and time consuming to contruct by hand.
- Can be difficult to read accurately.
- Can often require additional information for them to be useful.
- Effective in showing the spatial density.
- Shows variation and pattern.
- Easy to interpret.
- Purpose is easily understood.
- Easy to generate on a computer.
- Actual values can't be seen.
- Dots crowded/can lead to clustering- not very accurate.
- Time consuming if done by hand.
- Small countries arn't represented accurately (imagine dots in the USA compared to the UK- they may have the same amount of dots but the UK will look clustered and suggest lots of something whilst in the USA the dots may look sparse and suggest little of something).
- Easy to make a mistake/be subjective.
- Easy to compare.
- 3 bits of data can be compared at the same time as they use the same scale (are always out of 100).
- By using lots of graphs, comparisons can be made.
- Difficult to construct.
- May be wrongly interpreted.
- Quite difficult to read- have to have the background knowledge of how to use the graph.
- Very visual
- Can represent a large range of data.
- Not dependent on size of the area.
- Difficult to produce.
- Not accurate/ can't extract exact data.
- Overlap can occur making it confusing and difficult to read/interpret.
Flow Lines, Desire and Trip Lines
- Immediate impression- visual
- Can show movements easily such as traffic/migration etc.
- Desire lines show trends in migration of population.
- Gives clear sense of direction.
- Clear locational component.
- Hard to draw.
- Flows can be in the same direction/overlap.
- May be difficult to show meeting point of the wide bands without overwhelming the map.
- Visual impression of change over a space- gives general impression.
- General anomolies can be indentified.
- Easily done by hand or on the computer.
- Doesn't breach data protection.
- Good for data which involves density reading.
- Easy to interpret via a key.
- Gives false impression of abrupt change at the boundaries.
- Variations within each area are hidden.
- Reading exact figures is impossible.
- Drawn easily on computers.
- Can see areas of equal value.
- Can see gradual changes.
- Avoids the problem of boundary lines.
- Do not show discontinuous distributions.
- Only work where there is plenty of data spread over the study area and the changes are gradual.
- Small lines and numbers on graphs can be difficult to read.
- Shows the significance of the data.
- Proves/disproves correlation.
- Allows for further analysis.
- Can be difficult to work out.
- Quite a complicated formula.
- Can be misinterpreted.
- Can test association between variables.
- Identifies difference between observed and expected.
- Can't use percentages.
- Data must be numerical.
- Catagories of 2 are not good to compare.
- The number of overservations must be 20+
- Quite complicated to get right- difficult formula again!