Unit 1: Advanced Statistics / Hypothesis Testing

Question 1

What does the statistic tell you?

p value

Answer 1

The likelihood that differences in data are due to chance

Question 2

What does the statistic tell you?

A high p value

Answer 2

Differences in data are probably due to chance and not the presence / absence of the independent variable;
differences are NOT significant;
Fail to reject the null hypothesis

Question 3

What does the statistic tell you?

A low (below 0.05) p value

Answer 3

Differences in data are probably due to the presence/absence of the independent variable;
differences are significant;
Reject the null hypothesis

Question 4

What is

α level

Answer 4

The p-value that you set before an experiment, which will be used as your cutoff to determine significance

Question 5

What α level is used in biology?

Answer 5

p = 0.05

Question 6

p value > α level

Answer 6

Differences in groups of data are likely due to chance

Question 7

p value < α value

Answer 7

Differences in groups of data are NOT likely due to chance (ergo, are significant)

Question 8

What do you do with a null hypothesis if p < 0.05

Answer 8

Reject the H₀

Question 9

What do you do with a null hypothesis if p > 0.05

Answer 9

Fail to reject the H₀

Question 10

Define / describe

Test Statistic

Answer 10

A calculated value that is used to compare two or more sets of data, and is used for determining p value

Question 11

What are the two test statistics often used in biology?

Answer 11

χ² and t-test

Question 12

What test statistic is used if the dependent variable is continuous?

Answer 12

Student’s t-test

Question 13

What is a

Critical Value

Answer 13

A value that is determined by statisticians (i.e. you do NOT need to calculate these!) which correspond to p values for given statistical tests

Question 14

How do you use critical value to determine p values?

Answer 14

If the test statistic exceeds the critical value, then p ≤ 0.05 (or whatever α value you choose)

Question 15

If test value > critical value

Answer 15

p ≤ 0.05
Data are significant / likely due to relationship with independent variable
Reject the null hypothesis

Question 16

If test value < critical value

Answer 16

p > 0.05
Data are not significant / likely due to chance
Fail to reject the null hypothesis

Question 17

Steps of calculating χ²

Answer 17

1. Determine total number of observed data points
2. Determine expected values. These must be based on total number of observed data points (i.e. total of observed MUST equal total of expected)
3. (o-e) For each category, calculate difference between observed and expected values
4. (o-e)² Square each of those
5. (o-e)²/e Divide each of those by the expected value
6. Σ(o-e)²/e Add all of those values

Question 18

How to find df

Answer 18

If categorical: number of categories minus one
If discrete or continuous: sample size minus number of categories (ex: if n=10 in group a and n=10 in group b, the total is n=20 then minus two)

Question 19

What test statistic is used if the dependent variable is a count?

Answer 19

χ²