Statistics

Question 1

What does a chi squared result measure?

Answer 1

The difference between the expected results compared to the actual results

Question 2

How can you determine the results of a chi squared test? When do you reject or accept H0?

Answer 2

Compare to critical value at the ( number of categories - 1), Degrees of freedom
This tells us the max expected value expected to be calculated for a certain percentage

If the chi squared test is larger than the D of F value, we reject Ho as there is a significant difference. There is less than an X% chance of these results occurring from chance

Question 3

What is the general rule from stats test to reject H0?

Answer 3

If the calculated value from tests is larger than the D of Freedom value, we reject Ho
There is less than a X% chance of this happening from chance
So there is a significant difference / correlation

Question 4

When do you use T-tests?

Answer 4

When comparing the means of two samples within an experiment

Question 5

How do you calculate the degrees of freedom in t tests?

Answer 5

N1 + N2 - 2
So add the number of categories for the first sample and the second, then subtract two

Question 6

What test do you perform when looking at correlation?

Answer 6

Spearman’s Rank

Question 7

What test do you perform when comparing frequencies?

Answer 7

Chi squared

Question 8

How do you use standard deviations with graphs?

Answer 8

As error bars, 1 standard deviation above and below the the value (normally the mean)

Question 9

What do large standard deviations suggest about errors?

Answer 9

Random errors have occurred, usually due to low resolution in equipment

Question 10

What do 6 markers about graph analysis want you to say?

Answer 10

Start with supporting the statement, use data
Then disagreeing with the data, using stats and figures

Large majority: lack of validity in results and general errors, could include:
- lack of scale/numerical values
- no error bars, means, statistical tests
- no percentage change
- no method given for results
- no control variables givens
- what was recorded is not specified
- abiotic factors used
- lack of controls and repeats
- lack of consideration of other factors linking to the factor being investigated

Question 11

What can you infer from spearmans rank in relation to the critical value?

Answer 11

Whether the result is significant, so so whether the correlation is significant