experimental method - extraneous and confounding variables

• Demand characteristics

  A type of extraneous variable which could affect your dependent variable. 

  It occurs when participants become aware of a study's aims due to cues from the researcher or the research design. 

  The participants often: 

  Act in the way that they think is expected and over-perform to please the researcher

  Act deliberately or underperform to sabotage the results of the study

  • Act is a way which the most socially acceptable/makes them look popular = Social Desirability Bias
• Investigator effects

  A type of extraneous variable which could affect your dependent variable. Investigator effects occur when the researcher consciously or unconsciously influences the participants behaviour. 

  Types include: 

  Expectancy effects (giving away what you hope to find)

  Unconscious cues (encouragement, nodding, smiling, frowning, attitude ect)

  • Factors relating to the study's design (participants chosen, materials used, procedure designed)
• Pilot Studies

  Pilot studies are small-scale trial runs of a research study which takes place before conducting the 'real thing'.

  Pilot studies aim to:

  Test whether each part of the design works or not.

  Often identifies extraneous variables which we need to control for. For example: participants do not understand the instructions provided, guess what the study is about, get bored and lose interest.

  • Try before you buy... saves a bigger investment of time and money

  The results from the pilot studies are disregarded, we are just testing the materials/procedure.
• Control groups

  A group which is not exposed to the manipulation of the IV to establish a baseline result. Helps to establish 'cause and effect' as we can compare the results of the control group versus the experimental group and determine whether the IV had a significant effect or not.

  Example:

  When there's an experiment on how chocolate may effect a person, there would be a control group. The group that doesn't have any chocolate would allow us to identify the effect/change.
• Standardisation

  Ensuring that (as far as is possible) all participants within an investigation are subject to the same environment, information, and experience. 

  Example: 

  These participants have the same amount of chocolate.
• Randomisation

  the use of chance used wherever possible to reduce the researcher's influence on the design of the investigation. Aims to control investigator effects.

  Example:

  Put everyone's names into a hat, and the names that are chosen are the ones who get to eat the chocolate so it is not biased or based on a person's appearance.
• Single- blind condition

  When the participants are not told the aim of the investigation before they take part. Other details may also be kept a secret- such as, which condition they have been assigned to or whether there's a 2nd condition at all. Aims to reduce demand characteristics.

  Example: The researchers are the only ones who know what type of chocolate is given and the aim of the investigation sp the participants are honest with the way it affects them.
• Double-blind condition

  When participants AND the researcher who conducts the investigation are not told the aim of the investigation. Often this involves a 3rd party researcher who is hired to carry out the study (whilst those who designed the study analyse the results). 

  Example:

  Both the researcher and the participants would not know who took which chocolate and the aim either.
• Realism

  How realistic the results are to the real world.
• Generalisation (of results)

  The application of the results from a study, to the wider target population.
• Reliability

  A measure of whether something stays the same consistenly.
• Validity

  The extent to which a test measures what it claims to measure.