Statistics

Question 1

What is the main objective of statistical tests?

Answer 1

To determine whether a set of data differs significantly from outcomes that may be produced by chance

Question 2

What is a null hypothesis?

Answer 2

States there will be no significant difference / correlation between variables.

Investigations aim to disprove the null hypothesis

Question 3

What is a significant difference / correlation?

Answer 3

One unlikely to have occurred due to chance.

Works on a significance level (probability it is due to chance) - usually 0.05.

Question 4

How would you determine which statistical test to use?

Answer 4

• If you are collecting frequencies (counting number of individuals in categories), use the Chi-Squared test
• If collecting measurements, are you looking for correlations or differences between means?
• If looking for a correlation between variables of the same sample, use the Correlation coefficient test
• If looking for differences between two means, use the Student-t-test

Question 5

What is the Student-t-test?

Answer 5

Determines if there is a significant difference between means of sets of measured data.

The number of subjects doesn’t need to be the same in each sample, but should be at least 15

Question 6

How do you calculate degrees of freedom for the student-t-test?

Answer 6

(n1-1) + (n2-1)

where n1 is the number of subjects in sample one and n2 is the number of subjects in sample two

Question 7

Describe what the Spearman’s rank correlation coefficient values mean?

Answer 7

1 = a strong positive correlation
-1 = a strong negative correlation
0 = no correlation

Question 8

Describe the correlation coefficient graphs

Answer 8

A value of 1: y = x

A value of -1: a negative line in the positive quadrant

A value of 0: a horizontal line in the positive quadrant

Question 9

What is the Chi-squared test?

Answer 9

Looks at differences in frequencies between groups / categories and determines whether they are statistically different to what you would expect.

Question 10

Give the structure of a biological conclusion

Answer 10

• The test statistic is (less/more) than the critical value
• We (accept/reject) the null hypothesis
• There is a (more/less) than 0.05 probability the difference is due to chance.
• There (is not/is) a significant difference/correlation between (x) and (y)