validity + reliability

• Internal validity

  This refers to whether the effects observed in a study are due to the manipulation of the independent variable and not some other factor. In-other-words is it the IV that is causing the DV to change - or is there something else (an extraneous or confounding variable) that is having the effect?
• External validity

  This refers to the extent to which the results of a study can be generalised to other settings. External validity includes ecological and population validity
• Ecological validity

  This is a type of external validity that refers to how realistic your research setting is - does it represent real life. For example a controlled experiment that is run in a laboratory and maybe artificial is said to lack ecological validity.

  To improve ecological validity you could run your study in a field setting (real life environment) or make the tasks in your study as real life a possible
• Population validity

  This is a type of external validity that refers to how representative the participants are of the target population.
• Experimental realism

  the extent to which situations created in psychology experiments are real and impactful to participants. For example, in Milgram's obedience study did the participants believe the experimental situation?
• Reliability

  This term in psychological research refers to the consistency of a research study or measuring test. For example, if a person weighs themselves during the course of a day they would expect to see a similar reading. Scales which measured weight differently each time would be of little use. If findings from research are replicated consistently they are reliable.

  There are two types of reliability - internal and external reliability.
• Internal reliability

  This refers to the consistency of results across items within a test. This usually applies to a questionnaire used to measure personality or another aspect of behaviour. Do all the questions measure the same aspect of personality?

  You can deal with problems of internal reliability by testing your questionnaire out in a pilot study and take out any questions that seem to cause problems.
• External reliability

  This refers to the extent to which a measure varies from one use to another - if you did your study again with the same participants would you get the same results?

  To deal with external reliability you should make sure the research is run in the same way (consistently) for all participants. E.g. use standardised instructions and standardised procedures.
• Inter-observer (rater) reliability

  This refers to whether two observers (or interviewers) would record the behaviour or responses from the same person in the same way. To improve the reliability of different observers you should train your observers before they go out and collect data, and make sure any checklist of behaviours for the observation is clearly defined.
• Social Desirability bias

  This is something that will affect the validity of a study and is when participants don't answer or behave honestly to put themselves over in a good light.
• Demand Characteristics

  This is something that will affect the validity of a study. These are clues in the experiment or study that suggest to the participant what the aims or hypothesis of the study may be. This means the participants may change their behaviour to "help" the experimenter (please you effect) or possibly behave in a way to spoil the researt (screw you effect)
• Researcher or investigator bias

  The researcher has a hypothesis that they want to prove. This may lead to them intentionally or unintentionally encouraging behaviours from the participants that will support the hypothesis
• Participant variables

  These are differences between people that are not of interest in the study (not the IV but an extraneous one) but may affect the outcome of the study. They could include gender, age, ethnicity, IQ etc. They will effect the validity of the study
• Investigator effect


  This is something that can effect the validity of a study. The characteristics of the investigator may affect the way the participants behave which is an extraneous variable. For example the participants may behave differently if the researcher is more formal or casual, old or young, good looking or not etc