STATS - PAPER 3 GEOG

Question 1

Standard deviation

Answer 1

•Is a more reliable measure or dispersion than IQR

LARGE SD: The numbers in the data set are spread out around the mean

SMALL SD: The numbers are bunched closely around the mean

Question 2

Spearmans rank PART 1

Answer 2

Is a test to find out whether two sets of numbers are correlated. E.g. GDP per capita and life expectancy
•The number you get is ALWAYS between -1 and + 1.
-POSITIVE NO: Means the variables are positively correlated - as one variable increases, so does the other. •(THE CLOSER TO 1=STRONGER CORRELATION)

-NEGATIVE NO: Means the two sets of variables are negatively correlated - as one variable increases, the other decreases. •(CLOSER TO -1=STRONGER CORRELATION)
•If coefficient=0, there probably isn’t much of a relationship between figures
EXAMPLE: 0.95= close to 1, therefore positive strong correlation

Question 3

Spearmans rank PART 2

Check correlation for significance

Answer 3

A SR may tell you that the 2 sets of numbers are correlated. But you need to check is there is evidence for a genuine link between the two.
(you sometimes get correlations by chance)
•You can check for evidence of genuine link by looking at probability that a correlation would happen by chance. If there’s a 5% (or higher) probability that a correlation is cos of chance, then it’s not very significant.
•If there’s 0.1% or less chance, then it’s very significant evidence for a link.

Question 4

Chi-Squared rest

PART 1

Answer 4

Tells you whether two variables are linked
•Start by making a hypothesis and a null hypothesis
•Use the NULL hypothesis to predict a result - EXPECTED RESULT
•Next, the experiment is carried out and actual result is recorded - OBSERVANT RESULT
•Chi squared is then carried out, and outcome either supports the null or allows you to reject it.

Question 5

Chi squared PART 2

Answer 5

Compare your results to the critical value
•The X2 value shows whether there is a significant difference between your observed and expected results. If so u can reject null
•To find out if there is significant difference between observed and expected - u need to compare x2 value to a critical value
•Critical value is the x2 value that corresponds to a 5% level of probability that the difference between O and E is due to chance
•If x2 value is SMALLER than critical value = no significant difference between O and E = NULL
•If x2 value is LARGER than critical value = significant difference between O and E results = null rejected
•Example: CV of 5.99. The x2 value of 11.37 is bigger than 5.99, so there is a significant difference between O and E results = Null rejected
BE CAREFUL: your x2 value is evidence supporting ur hypothesis that there is a link - HOWEVER: it doesn’t prove this link - there could be other factors involved