Experimental Design

• Experimental design is how we allocate participants to conditions.
• Repeated measures: Definition

  Each participant is tested in both conditions. Participants are tested against themselves.
• Repeated measures: Example

  Participants run on a treadmill and then came back a week later to do it again.
• Repeated measures: Strengths

  As the same participants are used in each condition, participant variables are reduced and fewer people are needed as they all take part in all conditions: saves time.
• Repeated measures: Weaknesses

  Control: you need to change materials, you cannot do the same test twice and one may be harder than the other.

  Order effect: order that conditions take place could affect results (overcome with counterbalancing). C2 is often ruined so people usually do better or worse in C2: lack of consistency as you roughly know what to expect.
• Repeated measures: Overcoming limitations

  Counterbalancing:

  1. Turn all names into numbers and randomly allocate half the numbers to A and half to B. A do C1 first and B do C2 first.
  2. Do not allow A and B to discuss after finishing their condition.
  3. Give them a rest.
  4. B then do C1 and A do C2.
• Independent groups: Definition

  Different participants in each condition. Participants only experience one condition.
• Independent groups: Strengths

  Avoids order effects (such as practice or fatigue) as people participate in one condition only.
• Independent groups: Weaknesses

  Participant variables/ individual differences. You need more participants than repeated measures to end up with the same amount of data.
• Independent groups: Overcoming limitations

  Random allocation:

  1. All names into numbers
  2. All numbers into random number generator
  3. 1st half into condition 1
  4. 2nd half into condition 2.
• Matched pairs: Definition

  Involves matching participants on the key characteristics of the study and placing them in each condition. (Match and pair participants and then randomly allocate one of each pair to each condition).
• Matched pairs: Strengths

  Reduces participant variables because the researcher has tried to pair up the participants so that each condition has people with similar abilities and characteristics.
• Matched pairs: Weaknesses

  Time consuming- trying to match participants

  Can only match variables known to the researcher, there could be hidden variables.

  If one participant drops out, you lose 2 participants' data.
• Matched pairs: How to match participants

  1. Identify a key variable
  2. 'Test' them on that key variable
  3. Analyse the results t match the pairs.
  4. Randomly allocate one of the pairs to C1 and the other to C2.
• Correlations

  Designed to look at the strength (correlation coefficient, runs from 1 ->0 -> -1 )and direction (positive or negative)of relationships between co-variables.
• Experiments

  Look at the difference between two conditions of an IV.
• Co-variables

  Variables investigated in a correlation. Not referred to as independent or dependent variables because the study is investigating their relationship.
• Positive correlation

  Both co-variables increase with eachother (same direction).
• Negative correlation

  One CV increase and the other decreases (opposite directions).
• Correlation coefficient

  The further from zero, the stronger the correlation. (-1 = perfect negative correlation, +1 = perfect positive correlation). Between -0.3 and +0.3 there is no correlation.
• Strengths of correlations

  No manipulating anything, more ethical.

  Allows predictions to be made.

  Can be used to build on experiments.
• Weaknesses of correlations

  Cannot infer causation, not carried out under controlled conditions.

  Two variables might be linked but caused by a third variable. (E.g. authoritarian personality + fascism).

  Coefficient only works for linear relationships.
• Correlations show how much data you have, you cannot ignore data.