Reliability and Validity

Get Started. It's Free
or sign up with your email address
Rocket clouds
Reliability and Validity by Mind Map: Reliability and Validity

1. Validity

1.1. External validity

1.1.1. External validity is concerned with how well the results of a study can be generalised beyond the study itself - studies have high external validity when Ps behaviour reflects realistic behaviour

1.1.2. Population validity The degree to which the results can be generalised from the limited sample of the study

1.1.3. Ecological validity The degree to which the findings can be generalised to other situations, places and conditions

1.2. Internal Validity

1.2.1. The ability of the study to test the hypothesis that it is designed to test - are we measuring the effects of the IV on the DV?

1.2.2. Threats to internal validity Demand characteristics Experimenter bias Order effects Participant variables

1.2.3. Improving internal validity You should remove all confounding variables - this is done through the use of controls Single blind technique Double blind technique Counterbalancing Use matched pairs design or repeated measures instead of independent measures

1.3. Face validity

1.3.1. This is simply a judgement about whether a test seems valid

1.4. Criterion validity

1.4.1. Looks at whether a test of a particular construct relates to other measures it Concurrent A tes shows concurrent validity if it shows similar findings to another existing measure Predictive Measured by how well a test predicts future performance

1.5. Internal vs external validity

1.5.1. The more that confounding variables are controlled (internal validity) the less the more the study becomes artificial thus reducing external validity - vice versa

1.5.2. Deciding whether internal or external validity is more important depends partially upon the purpose of the study If a study is designed to test the detail of a theory then internal validity is more important If the study is designed with intention of applying results to the real world (e.g. effectiveness of CIT) then external validity is more important

2. After we study data we have to decide how useful the data from the study is likely to be

2.1. Is it reliable?

2.2. Is it valid?

2.3. Are there any biases due to sampling?

3. Reliability

3.1. External reliability

3.1.1. This is the most important for the majority of studies

3.1.2. This is the ability to reproduce the same results every time the test is carried out

3.1.3. Typically assessed using the test-retest method (doing it again and seeing if results are the same)

3.1.4. One way of getting a measure of reliability is to do a test of correlation between two sets of scores - a high correlation coefficient indicates the test is reliable

3.2. Internal reliability

3.2.1. Usually associated with measures such as attitude scales or psychometric tests e.g. personality tests

3.2.2. Is concerned with the consistency within a test

3.2.3. Typically assessed using the split half method (compares one half of the test with the other to check the scores of the variable are consistent)

3.3. Improving reliability

3.3.1. Take multiple measures from each P e.g. do the test 3 times and take an average score

3.3.2. Use a pilot study to check that the proposed method of measurement works properly and that Ps can use apparatus

3.3.3. Standardised procedure for investigators where observations or interviews are used in order to improve inter-rater reliability

3.3.4. Careful copying of data to ensure things are not copied or recorded incorrectly

4. Issues of sampling

4.1. A key issue when choosing a sample is whether it has population validity - this is increased when the sample is representative of the target population - this in turn increases external validity

4.2. Volunteer

4.2.1. Research has found that those who volunteer are not representative of the whole population as they re generally more stable and outgoing than those who do not volunteer

4.3. Random

4.3.1. Has greatest population validity as it is the most representative

4.3.2. Rarely used as psychologists cannot force anyone to take part (ethical guidelines)

4.4. Oppportunity