Skills and techniques for geographical investigations



Skills and Techniques for River Investigation (Section C)

A- Making sense of graphs

o Graphical method should be appropriate and justified. [Why do you use this method?]

o Graphs illustrate data in a visual way as an aid to interpretation.

o Similar types of information may be plotted on the same set of axes.

o Two or more graphs should be placed on the same page where information is to be compared and interrelationships are being investigated.

o Data must be presented in a variety of ways.

o All illustrations should use appropriate annotations.

o Text may be replaced by detailed and effective annotated diagrams, particularly if the world count is in danger of being exceeded.

o Diagrams, graphs, maps and photographs are included for a purpose, which should be clear from the annotations.

o Each item should be numbered in sequence and carry a reference to it in the main text.

o Primary data: data, which your group collected.

o Secondary data: data from all groups.

B- Types of graphs

1- Line graphs

o Most common ways of representing data which shows change over time or space.

o Effective way of comparing trends in similar data on one graph.

o Describe the general trend shown by the data and identify any anomalies i.e. any odd features which do not fit the general pattern [annotate the graph with this information].

o Use data from the graph to illustrate what you mean in your answer.

o Look out for positions on the graph where the line changes, i.e. the rate of change varies and use this for your description.

o Suggest reasons for the pattern you have identified. Draw answer on the theory and general knowledge of the subject.

2- Scatter graphs

o Quick and easy way to show a correlation between two variables.

o E.g. discharge (independent variable on X-axis) and width (dependent variable on Y-axis) of the channel.

o Draw best-fit line.

o Annotate diagram to show main trend and anomalies. Add data! (Missing from example below)

o Suggest reasons for the pattern you have identified, main trend e.g. positive correlation and anomalies.

3- Bar charts

o Easy to construct and use.

o Quantitative y-axis

o May be simple, i.e. represent one set of data

===================== Other methods of presentation========================

1 The 3 cross-sections have to be on the same graph paper/ same scale/ neat in black ink/ don’t forget scale and title, Figure number. Annotate to show changes between cross-sections.

2 Field sketch to highlight most important geographical features.

o Shows the necessary information e.g. narrow/wide channel, banks made of clay and pebbles from river deposition [alluvium], bed load (boulders), vegetation.

o Annotate your sketch. All annotations should be relevant to the aim of the sketch.

3 Annotated photos, for instance to suggest reasons for findings.

4 Original method to present findings

B-Making sense of statistics

Statistical methods: Justify and add formula

1- Descriptive statistics (How?): Mean. Useful start to comparing sets of data, more manageable form by summarizing a large amount of data. However, it is distorted by extreme values.

2- Investigate relationships between sets of data (Why?):

Spearman’s Rank Correlation coefficient. Click here for formula etc…

The Spearman’s test indicates the strength of the relationship as a numerical value between -1 and +1. The closer the answer to -1, the greater the inference of a negative relationship and vice versa. The closer the answer to 0, the greater the inference of no relationship or a random pattern.

❑ Number that summarises the relationship between twos sets of data.

❑ Used to investigate whether or not there is a linear relationship between two sets of data.

❑ Should be at least 10 pairs of data.

❑ Draw a scatter graph of the sets of data studied to get a visual impression of the relationship before you begin the calculation. Determine from this whether there appears to be a linear relationship.

❑ May be positive or negative and the calculated figure must be between +1 and –1.

Spearman’s Rank Correlation coefficient shows whether the relationship has occurred by chance, or whether it would be replicated if different values are used. It focuses on how reliable the results may be.

Use an online calculator! Click here!

Critical values for the significance of correlation coefficient:

95% significance level: 0.56. There is 5 per cent likelihood of the relationship occurring by chance. This is the limit of risk.

Below this value there is too much risk of the relationship occurring by chance to be able to make reliable statements about the relationship( rejection level.

99% significance level: 0.75

The value of 99% is within acceptable limits for definitive statements to be made about the relationship. The reading shows a high level of confidence and is very significant. There is only 1 per cent likelihood of the relationship occurring by chance.

(See table of critical values on blog)

To add to section C when you use the coefficient for the first time.

The Spearman’s test indicates the strength of the relationship as a numerical value between -1 and +1. The closer the answer to -1, the greater the inference of a negative relationship and vice versa. The closer the answer to 0, the greater the inference of no relationship or a random pattern.

Critical values: The coefficient must exceed the critical value for the test at the 95 per cent level (also expressed as the 0.05 level of significance)

Significance: statistically, significance is only attached to our findings if we exceed the commonly accepted 95 per cent level of confidence that the result was not a statistical chance or fluke. Any less than this we would reject the findings.

[pic]

Critical values

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download