What experimental design means

1. Experimental design is the way a researcher decides which participants take part in which condition of a study.
2. Two designs cover it: repeated measures and independent measures.
3. The choice is about who does the conditions, not about how many conditions there are.
4. The design sets how many people are recruited, how one score is set against another and which problem has to be guarded against.

Repeated measures design

1. In a repeated measures design the same participants take part in every condition.
2. Each person's score in one condition is set against their own score in the other, so the comparison happens inside one person.
3. In a memory study the same 20 pupils would learn one 20-word list with music playing and a second 20-word list in silence.
4. That gives two scores per participant, which is what a difference hypothesis needs.
5. Strengths of repeated measures.
   1. Participant variables are removed, because the very same people do both conditions, so a difference in the scores is more likely to be produced by the independent variable.
   2. Removing that alternative explanation raises the validity of the study.
   3. About half as many participants are needed for the same amount of data, so the study is quicker and cheaper to run.
   4. A smaller sample is easier for another researcher to gather when they repeat the study to check its reliability.
6. Weaknesses of repeated measures.
   1. Order effects build up as participants work through the conditions.
   2. Seeing both conditions makes the aim easy to guess, so demand characteristics can change what participants do in the second one.
   3. Anyone who drops out before the second session is lost from every condition rather than from just one, which shrinks the data twice over.
   4. Two sets of materials are needed, such as two 20-word lists matched for difficulty, which takes preparation.

Order effects

1. An order effect is a change in performance that comes from the sequence of the conditions.
2. A practice effect raises the second score, because the participant has already worked through a similar task once.
3. Fatigue and boredom lower the second score, because the participant is tired of the task by the time they reach it.
4. An order effect gives a second explanation for the difference, so it costs the study validity in the same way an uncontrolled extraneous variable would.
5. Leaving a gap of a day between the two conditions cuts the effect down.
6. Using two sets of materials matched for difficulty, such as two 20-word lists of equally common words, also cuts it down.
7. Order effects arise only in a repeated measures design, because they need the same person to do more than one condition.

Independent measures design

1. In an independent measures design different participants take part in each condition, so each person meets only one level of the independent variable.
2. The mean of one group is compared with the mean of the other, so the comparison happens between two sets of people.
3. In a learning study one tutor group would revise a topic from a video while a different tutor group revised it from a booklet.
4. Both groups then sit the same test under the same standardised instructions, so the only planned difference is the independent variable.
5. Strengths of independent measures.
   1. There are no order effects, because nobody ever does a second condition.
   2. Participants see one condition only, so they are far less likely to work out the aim and produce demand characteristics.
   3. The same materials can be used in both conditions, so there is no need to match two word lists or two tests for difficulty.
   4. It is the workable choice when doing one condition would spoil the other, such as a study where a leading question would give the neutral version away.
6. Weaknesses of independent measures.
   1. Participant variables differ between the groups, so a gap in the scores may come from the people rather than from the independent variable, which lowers validity.
   2. Recruiting participants who are similar on the variable that matters most, such as reading age, reduces that problem.
   3. Roughly twice as many participants are needed, which takes longer and makes a representative sample harder to reach.
   4. An uneven split between the groups can hide a real effect, so a weak result is hard to read.

Choosing between the designs

1. A repeated measures design fits when the same people can safely do both conditions, such as learning two matched word lists a day apart.
2. An independent measures design fits when doing one condition would give the other away, such as a leading question followed by a neutral one.
3. An independent measures design also fits when the conditions are groups that already differ, such as two age groups.
4. The reason for the choice is given in terms of the study in front of you, naming its task and its participants.
5. The design of a study that is already described is worked out from whether the same people appear in both conditions.

Turning this into marks

1. When a design and a reason are both asked for, make them two separate points, one naming the design and one naming what it does for this study.
2. Name the order effect as practice, fatigue or boredom, because "the order changed things" does not secure the knowledge mark.
3. A strength has to belong to the design chosen, so tie it to the task and the people described rather than to the design that was rejected.
4. The mark most often lost is giving participant variables as a weakness of repeated measures, when the same people doing both conditions is what removes them.