Target populations and samples

1. The target population is the whole group a piece of research is about, such as all Year 11 pupils in the UK.
2. Testing all of them is almost never possible, so the researcher selects a sample, the smaller group who actually take part.
3. The way that sample is picked is the sampling method, and three of them are random, opportunity and self-selected.
4. The population is named precisely, because "teenagers" and "Year 11 pupils at one school in Cardiff" are very different claims.
5. The sample is what gets studied, and the conclusion drawn from it is about the whole population.

Representativeness and generalisability

1. A representative sample holds the same mix of people as the target population, with age, gender, background and ability in roughly the right proportions.
2. Generalisability is how far a finding from the sample can be applied to the whole target population.
3. A sample that leaves out a whole group is unrepresentative however large it is.
4. When a sample is unrepresentative, the study loses population validity, because the result may not hold for the groups it missed.
5. Gender bias, cultural bias and age bias all start at the sampling stage, and can be built in without anyone intending it.

Random sampling

1. Random sampling gives everyone in the target population an equal chance of being picked.
2. It starts from a full list of the population, with every name numbered, and numbers are then drawn from a hat or produced by a random number generator until the sample is full.
3. Strength: it is the least biased of the three methods, so it usually gives the most representative sample and the safest generalisation.
4. Strength: the researcher has no say in which names come out, which rules out their own preferences shaping the sample.
5. Weakness: a complete list of everyone in the population is needed, and for something like all UK teenagers no such list exists.
6. Weakness: the people drawn can still refuse, and those who refuse may share a feature that then goes missing from the sample.
7. Weakness: chance alone can produce an odd mix, such as 18 girls in a sample of 20 drawn from a mixed school.

Opportunity sampling

1. Opportunity sampling uses whoever is available and willing at the time.
2. In practice the researcher goes to one place at one time and asks the people there, such as the pupils in the canteen at lunch.
3. Strength: it is the quickest and cheapest method and needs no list of the population.
4. Strength: it fits a small study that has to be finished inside a single lesson.
5. Weakness: it reaches only whoever happens to be in that place at that time, so the sample is tied to one setting and one moment.
6. Weakness: the researcher chooses who to approach, so their own preferences can shape the sample.
7. Weakness: a narrow sample lowers population validity, so a finding from 20 pupils in one canteen says little about Year 11 nationally.

Self-selected sampling

1. Self-selected sampling means participants put themselves forward, which is why it is also called volunteer sampling.
2. The researcher advertises, perhaps with a poster on the form room door or a message in a tutor group chat, and uses whoever replies.
3. Strength: it takes little effort from the researcher, since the participants come forward on their own.
4. Strength: volunteers are willing, so fewer of them drop out part way through a long study.
5. Strength: it can reach people who would be hard to find any other way, such as anyone who has trouble getting to sleep.
6. Weakness: volunteers tend to be one type of person, often more confident, more interested or keener to please the researcher.
7. Weakness: that shared feature lowers representativeness, so the finding applies to keen volunteers more than to the population named.
8. Weakness: a keen volunteer wants the study to work, which makes demand characteristics more likely.

Sample size

1. A larger sample is more likely to be representative, because one unusual participant shifts the mean less.
2. With 5 participants one very high score drags the mean upwards, and with 50 the same score barely moves it.
3. A big sample costs more time, more money and more materials, so the size chosen is the largest one the study can actually run.
4. A large sample drawn from one narrow group is still biased, so size is judged separately from representativeness.
5. Sample size can be given as a fraction of the population, so 20 pupils out of a year group of 300 is 20 over 300, which cancels to 1/15, or about 6.7%.

Principles of sampling in science

1. The sample is chosen so that the results stand for the population rather than only for the people tested.
2. The sampling method is written down in enough detail for another researcher to copy it exactly.
3. Running the study again with a fresh sample from the same population is how psychologists check that a finding is reliable rather than a one-off.
4. Sampling sets how far a finding can be generalised, and the measure used sets whether the study captured the right behaviour.

Describing a sampling method

1. Name the target population first, as in "Year 10 pupils at one school".
2. Give the method and the number, such as an opportunity sample of 20 pupils from that year group.
3. Set out the actual steps, such as standing by the canteen door at lunch and asking the first 20 pupils who agree.
4. Choose a method that can be carried out with the time, the space and the people available.
5. Give one weakness of the method chosen and say what it costs the generalisation.
6. When judging a sample that is already described, use its own numbers to show how far it matches the population.

Turning this into marks

1. When a method and a weakness are both asked for, keep them as two separate points, one naming the method and one naming what it costs the generalisation.
2. Name the population, such as Year 11 pupils at one school in Cardiff, because "teenagers" does not secure the knowledge mark.
3. Describe the steps, such as numbering all 300 names and drawing 20 of them from a hat, rather than giving the label on its own.
4. The mark most often lost is calling an opportunity sample random, when nobody outside the canteen had any chance of being picked.