# BIOL132 - L3

0.0 / 5

- Created by: Katherine
- Created on: 07-06-16 17:13

What is a sample?

A representative of a subset of the population

1 of 41

Why does a sample need to be represnetative?

There are all sort of variation in whatever we are investigating - people, animals, plants, cells etc

2 of 41

What might the variation cause?

It might affect our findings. If our sample is different in these factors from the population, our findings might only apply to our unusul smaple - so we can't generalise

3 of 41

Why else?

Unsystematic variation can also interfere with our interpretation of differences across experimental groups/condiitons. In experimetns, we are testing for systematic variation across experimental groups/conditions.

4 of 41

How do we get a sample that is representative of the population?

A random sample of the population. Evens out all the sources of unsystematic variation within the population. Need good information on the population to generate a random sample.

5 of 41

What are the negatives of randomisation?

It can miss relatively rare but potentially important sources of unsystematic variation

6 of 41

What are not good options of participants?

Self selection or volunteers (not representative). Convenience sampling, haphazard sampling.

7 of 41

What is convenience sampling?

Choosing whatever is easier for you without any reference to possible sample biases.

8 of 41

What is haphazard sampling?

Sampling with your eyes shut is not random sampling.

9 of 41

What is likely if you use these methods?

Your sample is likely to be biased.

10 of 41

How can you minimise the chances of unwanted bias or difference in your sample across experimental groups?

Randomly assign. Any individual has the same chance as being chose as anyone else in the group.

11 of 41

What does randomisation work best in?

Large numbers, and there may be some factors you know are important and you want to make sure are balanced/matched across your groupp

12 of 41

What is a population?

The entire set of participants of interes

13 of 41

What is generalizability?

The ability to infer population characteristics based on the sample

14 of 41

What is probability sampling?

The liklihood of any member of the population being selected is known

15 of 41

WHat in nonprobability sampling?

The liklihood of any member of the population being selected is unknown

16 of 41

What is sampling error?

Difference between sample and population characteristics

17 of 41

What is the goal of any technique?

Reducing sampling error

18 of 41

What happens as sample size increases?

Sampling error decreases

19 of 41

Which samples are more representative?

Big

20 of 41

What are the flaws of big sample sizes?

More expensive, more recruitment, longer times. Larger samples ignore the power of scientific inference

21 of 41

When is a larger sample needed?

When variability within each group is great, when differences between groups are smaller.

22 of 41

Why might this be the case?

Because, as a group becomes more diverse, more data points are needed to represent a group. As the differencce between groups is smaller, more participants are needed to reahc critical mass to detect difference.

23 of 41

What is non-sampling bias?

It is present even if sampling and analysis is done correctly. Would still be present if survey measured outcome in entire sampling frame.

24 of 41

What is the type of non sampling bias?

Sampling frame is not equal to population to which you want to generalize. It may be out of data, or there may be non-response among sampling units in sampling frame.

25 of 41

What is measurement error?

Tape incorrectly fixxed to height board.

26 of 41

What are some sources of non-sampling bias?

Sampling frame out of data, non response and measurement error

27 of 41

What is prevention of sampling frame out of date?

Use current sampling frame, limit generalizations.

28 of 41

What is prevention of sampling non response?

Minimize non response. Use various statistical methods to weight data.

29 of 41

What is the preventino of measurement error bias?

Standardize instruments. Write clear and simple questions. Train survey workers. Supervise survey workers

30 of 41

What is sampling bias?

Selection of nonrepresentative sample, the likelihood of selection not equal for each sampling unit. Failure to weight analysis of unequal probability sample.

31 of 41

What is sampling bias essentialy?

You have not sampled data with equal probability, and you have no accounted for this in your analysis

32 of 41

What is an example of sampling bias?

Non representative sample e.g. convenience sampling in biology. (e.g. almost every model system that is used - cell, insects, mice ect.) Failure to do statistical weighting

33 of 41

WHat are examples?

Non representative sampling, failure to do weighting

34 of 41

What is the prevention of nonrepresentative sampling?

Calculate the probabilities of selection.

35 of 41

What is the prevention of failure to do weighting?

Apply appropriate statistical weights if selection probabilities unequal

36 of 41

What is sampling error?

Difference between survey result and population value due to random selection of sample.

37 of 41

What is samplnig error influenced by:

Sample size, sampling scheme

38 of 41

Can sampling error be predicted?

Yes, it can be calculated and accoutned for.

39 of 41

What are measures of sampling error?

Confidence limts, standard error, p values, others

40 of 41

What do we use these error to do?

To determine how sure we are of result after analysis, calculate sample size prior to sampling.

41 of 41

## Other cards in this set

### Card 2

#### Front

Why does a sample need to be represnetative?

#### Back

There are all sort of variation in whatever we are investigating - people, animals, plants, cells etc

### Card 3

#### Front

What might the variation cause?

#### Back

### Card 4

#### Front

Why else?

#### Back

### Card 5

#### Front

How do we get a sample that is representative of the population?

#### Back

## Similar Biology resources:

0.0 / 5

0.0 / 5

0.0 / 5

0.0 / 5

0.0 / 5

0.0 / 5

0.0 / 5

0.0 / 5

0.0 / 5

0.0 / 5

## Comments

No comments have yet been made