Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


Reliability and Validity

Examining the concepts of reliability and validity in relation to psychological research

Denise Hunt

on 19 November 2012

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Reliability and Validity

VALIDITY one of the most important concepts in research methods INTERNAL VALIDITY EXTERNAL VALIDITY What caused the effect on the DV? Was it the IV, or something else, such as EV's? This type of validity is concerned with what goes on inside the research. This is related to generalising. Can we apply, or generalise, our findings from a piece of research to situations beyond the study? This is concerned with what goes on outside the research. Participant Effects Demand Characteristics: people look for cues on how to behave. Certain situations 'demand' certain behaviour, but in research people are often uncertain how to behave - so they look for the cues. These cues are the demand characteristics.

'Social Desirability' is the result of participants wanting to please the researcher and be helpful
'Screw-you Effect' is the result of participants deliberately behaving in a way to spoil the study. what affects it? Dealing with Participant Effects SINGLE BLIND METHOD DOUBLE BLIND METHOD The participant does not know the true aims of the study. There is often deception involved. Both the participant and the experimenter are blind to the aims, or the conditions that the participants are allocated to, so the experimenter cannot produce any cues about what they expect what else affects internal validity? Investigator Effects Direct effects of the investigator lead to demand characteristics, but there are also indirect effects:

They could operationalise the variables in such a way that the desired result is more likely

Or, they might not give standardised instructions which leaves room for the results to be influenced by the researcher EXTRANEOUS VARIABLES SITUATIONAL VARIABLES Time of day, temperature, noise etc.
Order Effects (practice, boredom, fatigue)
Investigator effects
Demand Characteristics PARTICIPANT VARIABLES Any individual difference issue will affect the internal validity IF an independent measures design is used. If a repeated measures design is used then participant variables are controlled for. Ecological Validity: the ability to generalise to different places or settings

Population Validity: the ability to generalise to different people or populations

Historical Validity: the ability to generalise to different times (e.g. 1990's to 2012)

Mundane Realism: does the task in the study reflect real-world or everyday experiences VALIDITY RELIABILITY Checking for Validity of Research... Lie-Scale: if it is a self-report method, you can include questions to assess if people are telling the truth or not in the study (e.g. "Are you always happy?" A truthful answer would be no).
Face Validity: does the self-report measure look like it is measuring what the researcher claims to be investigating? This only requires measurement by intuition.
Concurrent Validity: can be checked by comparing performance on a new questionnaire with a previously validated one on the same topic. Participants are given both questionnaires at the same time, then the scores on each compared. If they are both similar then the measure is said to have concurrent validity. Any tool that is used to make measurements must be reliable. If you used a ruler to measure the size of a box, you would expect the box to measure the same the next day. This would make the ruler reliable as it gave consistent measurements. The same is true for psychological measures. We want them to be consistent and reliable. EXTERNAL One single ruler should give the same measurement of something even if there is a day or a month between making the measurements. One questionnaire should have the same outcome if it was given to a participant one day and also a week later. Two interviewers/observers should also produce the same findings.
This is called INTER-RATER RELIABILITY This refers to whether something is consistent INTERNAL this is a measure of the extent to which something is consistent within itself (e.g. are ALL the questions in an interview measuring the same thing, or are all the conditions of the IV measuring the same IV. checking reliability internal... SPLIT-HALF METHOD: in an interview or questionnaire you split the questions in half and compare them to each other checking reliability external... TEST-RETEST METHOD repeating the test with the same person on two separate occasions to see if the same results are obtained. If the results are the same then it has HIGH external reliability, but if they are not then it has LOW external reliability. the time between the first and second test/interview/questionnaire must be long enough for the participant not to remember their responses/answers
...BUT not too long that their opinions, abilities or behaviours might have changed. The two scores can be checked for consistency using a correlation Improving low internal reliability Improving low external reliability in the test/questionnaire, after using the split-half method to CHECK the reliability, then select the items that are most consistent to improve low external reliability we should standardise our measures, make sure there is no ambiguity and operationalise all variables and hypotheses. When evaluating any method or piece of research, you should ALWAYS consider it's reliability and validity, and comment on it. RELIABILITY G541: Practical Investigations
Full transcript