Statistics & Presentation Of Data Flashcards
What are flow diagrams used for
Flow diagrams are used for non-numerical data to show processes such as the global hydrological cycle and the farming system
What are logarithmic graphs used for
Logarithmic graphs are used for data sets that have large ranges such as population and GDP. It allows for all data to be graphed together.
What are Kite diagrams used for
Kite diagrams are used for presenting the percentage of vegetation found along a transect.
- useful to display several observations made at each of the sample point along the transect.
- illustrates change diagrammatically.
What are polar graphs are used for
Polar graphs are used to illustrate data combined with compass direction
What are triangulation graphs used for
Triangulation graphs are used to show data sets across three categories – for example primary, secondary and tertiary services. This data must be in percentages.
What are dispersion diagrams used for
Dispersion diagrams are used to show if data is clustered or dispersed.
They make it easy to draw comparisons – for example, life expectancy of two areas
What is bipolar analysis used for
Bipolar analysis is useful for visually comparing subjective results from a questionnaire or an environmental quality survey
How to do map work question
Always check the scale
Pick your location first then Blu-Tack the tracing paper down. Then draw with a ruler the building or route
Standard deviation model answer
Standard deviation is an important descriptive statistics as it gives a more accurate figure than the range or interquartile range as it takes into account all figures.
It is not affected by extreme values .
It measures the spread of data from the mean while the range measures the two extreme values in the data set .
SD shows the extent of the difference from the mean .
Using this statistics allows us to compare how the spread of the data around the mean has changed
Chi Squared - model answer
Chi squared can test the association between variables.
It is one of the most widely used and versatile tests of association.
The sample should include 20 observations in each area.
The data however must be in the form of frequencies and must have precise numerical value, for example you cannot use percentages.
The expected frequency in any category must be greater than five or else the statistic will not work.
The data must be organised into categories or groups.
It can be compared with significance tables to confirm whether the difference between the observed data and the expected data is a chance effect or has statistical significance.
Nearest neighbour analysis model answer
NNA can be used to identify a tendency towards clustering or dispersion of data sets.
Clustered: RN=0, all dots occur close to the same point.
Random: RN=1, there is no pattern in data
Regular: RN=2.15, there is a perfectly uniform pattern
It cannot always distinguish between single and multi-clustered distribution.
I
I
I
I
I
I
I
An index value of 1 does not always mean at totally random distribution. It may be related to a second unmapped factor
Spearman’s model answer
Spearman’s test is relatively quick and easy calculation in comparison to the Pearson’s Product Moment Correlation.
The test uses data which can be ranked but this means that the test loses some of its accuracy as it is not using the actual values .
Can only test for linear relationship is so scatter graph could be drawn to see if this is the case.
It requires a sample size for at least seven observations, however the larger the sample size are more reliable the result.
Pearsons model answer
Pearsons is the most widely used correlation method.
It is a more powerful method than Spearman’s rank because it uses the actual values instead of ranks in its calculations.
It Tests the strength of correlation between two variables and Ideally needs 12-15 data values to be an effective statistic.
You can also draw a scatter graph of the results to visually confirm whether there is a correlation between your two data sets.
Linear regression model answer
This statistical technique is the most commonly used method of predictive analysis to determine the linear relationship between one independent variable and a dependent variable.
It allows us to understand how much the dependent variable will change when we change the independent variable.
Regression equation allows us to mathematically calculate and plot the best fit line on a scatter graph.
By calculating this line of best fit mathematically – it is more certain than subjectively plotting it by eye.
It also allows us to protect trends and future values.
Proportional symbol maps
Proportional symbol maps use symbols to represent data that varies from place to place.
These maps have a good visual impact as the symbol varies in size depending on the figure that has been represented.
A common problem with proportional symbol maps is symbol congestion/overlap, especially if there are large variations in the size of symbols or if numerous data locations are close together.
It may also be difficult to calculate an exact values figures around in a key.
