# BIOL132 - L6

HideShow resource information
• Created by: Katherine
• Created on: 07-06-16 21:47
What is a z test?
We use a z test is our sample is very large, such as greater than a couple hundred subjects.
1 of 37
When do we use a z test if our sample is small?
If our estimate of the standard error of the population mean is a known parameter.
2 of 37
When do we use a t test?
When our sample is small and the standard error of population mean must be estimated from our sample statistics
3 of 37
Do we use z test if we know the standard deviation and the sample is about 30?
Yes
4 of 37
Do we use a z test if we don't know the standard deviation and the sample size is below 30?
No - we use a t-score
5 of 37
What do descriptive statistics do>
They provide basic measurs of a distribution of scores.
6 of 37
What do interential statistics allow?
Allow inferences to a larger population from the sample
7 of 37
How does inference work>
Representatvie sample from two groups are selected. Variables are tested. Means from each group are compared. Researchers conclude that measured differences between either result from chance or, reflect true difference.
8 of 37
What is a conclusion drawn regarding...
The role group membership plays in observed differences
9 of 37
What does statistical testing do>
It looks are the probabiltiy that this result occurred by chance - that if we ran this experiment 100 times we'd find that really there is no diffeerence between treatment or none.
10 of 37
What is the p-value of 50/100?
p=0.5 (50%)
11 of 37
What is the p value of 10/100?
p=0.1 (10%)
12 of 37
When does convention occur? When there is an accepted probabilty of at least...
p
13 of 37
Who decided on the 0.05 rule?
Ronald Fisher
14 of 37
What are related samples?
Those where the same experimental subjects provide two sets of scores e.g. without any training in maths and then later after training.
15 of 37
What are independent samples?
Where independent groups of experimental subjects, e.g. comparing males and females, subjects in group A or B.
16 of 37
What sort of sampe was the rat experiment>
Independent
17 of 37
When should parametric tests be used>
When the level of measurement is continous/interval or ratio (more than ordinal), the distribution of the scores are normal, calculations assume normal distribution, the variances of both variables are roughly equal or homogeneous.
18 of 37
What is a histogram?
It is a graph
19 of 37
What does parametric mean>
It means the the test uses statistics based on the normal distribution.
20 of 37
When comparing treated vs untreated rats, can we use parametric>
Yes - independent samples, the level of measurement os continuous/interval/ration/ and the data will be relatively normally distributed
21 of 37
What are the types of tests used to test differences between 2 samples?
Independent samples t test, mann whitney test, related samples test, wilcoxon test.
22 of 37
If I have independent samples which are parametric, what test should i used?
Independent samples t test
23 of 37
If I have independent samples non parametric data, what test should i use?
Mann whitney test
24 of 37
If I have related samples parametric data, which test should i use?
Related t samples t test
25 of 37
If I have related samples non parametric data test, which test should i use?
Wilcoxon test
26 of 37
What is independent sample?
If the values in one sample reveal no information about those of the other sample, then the samples are independent.
27 of 37
What is a related sample?
If the values in one sample affect the values in the other sample, then the samples are dependent.
28 of 37
What is the effect of the t value on the p value?
The larger he value of t, the lower the p value, the more likely we reject H0.
29 of 37
What is it called when we don't know whether the treatment may harm, help or have no effect, when we test for difference between means>
two tailed t test
30 of 37
How many ways are there to state out alternative hyptohessi>
2
31 of 37
What is the first way?
h0: control mean = treatment mean and H1: control mean not equal to treatment mean
32 of 37
What is the second wat>
h0: control mean = treatment mean 1: control mean < treament mean h1: control mean > treatment mean
33 of 37
What do we expect in a non directional hypothesis?
Extreme results that are some distance greater than or less than the controll mean. But were not sure which.
34 of 37
What do we expect in the second hyptohesis>
A directional hypothesis, one trailed test - we expect our result to be greater than the control mean, or we expect out result to be less than the control mean. we only test on of these two outcomes.
35 of 37
What is the critical value?
0.05 level.
36 of 37
The one taield p value is ... the 2 tailed p value?
Half
37 of 37

## Other cards in this set

### Card 2

#### Front

When do we use a z test if our sample is small?

#### Back

If our estimate of the standard error of the population mean is a known parameter.

### Card 3

#### Front

When do we use a t test?

### Card 4

#### Front

Do we use z test if we know the standard deviation and the sample is about 30?

### Card 5

#### Front

Do we use a z test if we don't know the standard deviation and the sample size is below 30?