# Interpreting Geographical Data

Year one statistics exam

• Summation sign
• Rounding
• Central tendency
• Variability
• Boxplots
• Standard deviations
• Normal distribution
• Sampling
• Reliability and standard errors
• Confidence intervals and t-distribution
• Colomn, charts and tables
• Hypothesis testing and one sample t-test
• Two sample t-test
• F-test
• Anova I
• Anova II
HideShow resource information
• Created by: Sophie
• Created on: 04-01-15 13:38

## 1. How do we calculate the F test?

• You have two double the P value as we need to allow for the fact that either of the variances could have been bigger. "=FDIST(0.53/0.197,99,99)*2"
• F = variance 1 ÷ variance 2. That is, F is simply the ratio of the variances. For this reason, the F-test is often called the ‘variance ratio test’. You use the degrees of freedom samples of the numerator and denominator
• F= variance 1 x variance 2
• F = mean 1 / mean 2.
1 of 20

## 2. Which is not a characteristic of an ANOVA I within country extreme example?

• When we can account of overall means as each country means, the country means are not significantly different, but are instead the same.
• When we can account for all the variance by using country means instead of overall means, the country means are significantly different.
• All the variation (this time SST=1) is within-country variation (SSE=1) and there is no between-country variation (SSM=0).

## 3. Which statement is incorrect about error bars on Excel?

• Excel only knows the value of the mean and not the other data sets
• It will either add the wrong error bars or add none at all as does not obtain the whole dataset
• You have to do is calculate the SD, SE or CI yourself and then enter it in this ‘custom’ option, specifying your correct value as both the upward-extending bar (‘positive’) and the downward one (‘negative’).
• You can use the option of standard error or standard deviation in Excel
• Make colomn first and then add the error bars yourself

## 4. Definition of central tendancy

• When cannot measure a value perfectly so automatically introduces uncertainty
• All different values
• A typical value
• A value that lies within the values

## 5. Which statement is incorrect about the FITTED VALUE?

• Our model uses this information, as best it can, to produce a prediction (fitted value) for any rice grain
• These fitted values can be compared with the actual (or OBSERVED) values.
• The best our model can do is predict that any given rice grain’s length will not be the mean grain length of rice from the same country – as judged by the grains in our sample.
• These values come from the best fit procedure, the mean is the single value that best describes the data, it fits the data better than any other single value