Methods

• Methods of assessing validity

  Face validity - Whether a test, scale, or measure appears to measure what it is intended to measure

  
  

  Concurrent validity - Whether the findings of a particular test/scale demonstrated is close to those obtained in another recognised test
• Methods of improving validity - Experiments

  - Control group means there is a better assessment of if the changes in the DV are due to the IV.

  - Standardisation minimises the impact of participant reactivity and investigator effects

  - Single & double blind
• Methods of improving validity - Questionnaires

  Many incorporate a lie scale within questions in order to assess the consistency of answers. Anonymity allows for the reduction of social desirability bias
• Methods of improving validity - Observations

  - Minimal intervention by researcher means for greater ecological validity

  - Covert observations

  - Broad behavioural categories negatively impact validity
• Methods of improving validity - Qualitative methods


  - Higher ecological validity as these often have depth and reflect reality

  - Interpretative validity - Interpretations of events matching participants

  - Validity is increased via triangulation (the use of different sources for evidence)
• Methods of assessing reliability

  Test-Retest reliability:

  Giving the same test or questionnaire to the same person on different occasions. Reliability is seen if the same or similar answers are given every time. Must be decently far apart to avoid recall.

  
  

  Inter-Observer reliability:

  Established as a way of moderation, pilot studies are used to check behavioural categories applied in the same way. Recorded behaviour is easy to do this
• Methods of improving reliability - Questionnaires

  Reliability of these are tested through Test-Retest. The correlation of these studies must be +0.80, if failing they must be deselected or rewritten.

  Split half method - Comparison of half the questions to the other half for similarities
• Methods of improving reliability - Interviews


  Same interviewer used every time, they are trained thoroughly to repeat in same ways and ask in same ways
• Methods of improving reliability - Experiments

  Lab experiments are strong for reliability as they allow for the control over some aspects of procedure allowing them to be consistent
• Methods of improving reliability - Observations


  Behavioural categories must be operationalised correctly, measurable and self evident. These should not overlap with one another.
• Features of science - Paradigms

  Kuhn - A shared set of assumptions and disciplines which are shared by members of its community.

  
  

  For: Many psychologists have an agreement regarding the definition of psychology

  
  

  Against: Kuhn suggests psychology doesn't have a universally agreed paradigm (alternative views on mental illness) and therefore its best seen as a "pre-science"
• Features of science - Theory construction and hypothesis testing

  Theory - General laws or principles with the ability to explain events, constructed through observations.

  Should be precise and clear predictions on the basis of this theory. Hypothesis testing is essential for a theory to be scientifically tested.

  
  

  For: Most psychologists develop a hypothesis prior to conducting their study

  
  

  Against: Many hypothesis testing is on a small sample which is barely representative, may be influenced by social and cultural norms.
• Features of science - Falsifiability


  Popper - Genuine scientific theories should hold themselves up for hypothesis testing and the possibility of being proven false.

  If it cannot be falsified, it is seen as a pseudoscience.

  
  

  For: Aspects of psychology such as ones involving biology and cognitive approaches can be falsified

  
  

  Against: Other aspects such as psychodynamic approach cannot be falsified. Many bio theories use causal correlations
• Features of science - Replicability


  Scientific study is trusted if findings are repeatable across different contexts. Replicability affects validity and thus studies must be reported with precision to verify their work

  
  

  For: Many psychological findings are replicated in supporting research (Hofling repeating Milgram)

  
  

  Against: Not all findings are replicated such as controversial ideas (Freud)
• Features of science - Objectivity and the empirical method

  Researchers must keep a critical distance during research and must not allow for personal opinions or biases to affect data or influence data.

  
  

  For: Experimental method and observational method are good examples of the empirical method

  
  

  Against: Some methods are more subjective and open to interpretation. Much subject behaviour in psychology cannot be directly observed.
• Correlations


  A mathematical technique in which the researcher investigates an association or relationship between 2 variables which are known as "co-variables"

  
  

  Correlation coefficient - A numerical value between -1 (negative correlation) and +1 (positive correlation) which represents the direction and strength of a relationship between co-variables. Closer to 1 = Stronger correlation.

  Correlation coefficients are denoted by the letter R
• AO3 - Correlations


  Allows to investigate naturally occurring variables which may be unethical or impractical to test by an experiment

  Allows to clearly and easily see if there is a relationship between variables in a graphical form

  
  

  Doesn't imply causation even if there is a strong association between the two variables

  Doesn't allow us to go beyond the data given and gives us no understanding as for reasons behind it
• Case studies


  A detailed and in-depth analysis of an individual, group, institution or event, typically focussing on the "abnormal" circumstances.

  
  

  Conducting one usually produces qualitative data as they construct a case history using interviews, observations or questionnaires. Though they may and can be subject to experimental or psychological testing which can produce quantitative data.

  
  

  Case studies tend to be longitudinal and may involve in gathering other data from other people involved with the participant such as friends and family
• AO3 - Case studies

  They offer rich, detailed insights which can bring light to unusual forms of behaviour

  Can contribute to our understanding of "normal" functioning

  Can generate hypotheses for future studies as one case study which provides contradictory research can cause the entire theory to be relooked at

  
  

  Lack of external and population validity

  Based on subjective selection and interpretation of the researcher

  Personal accounts may be prone to inaccuracy due to biases and memory decay, affecting internal validity
• What is a content analysis?

  A type of observational research in which people are studies indirectly via communications produced. Wide ranging and may include spoken interaction. Written forms or broader examples from the media. They aim to summarise and describe this communication in a systematic way so that conclusions can be drawn.
• What is coding and quantitative data in relation to content analysis?

  Coding is the initial stage of content analysis.

  Helps deal with extremely large data sets as they categorise the data into meaningful units.

  This may involve the counting of number of times a particular word or phrase appears in the text to produce quantitative data
• What is thematic analysis and qualitative data in relation to content analysis?

  Content analysis may involve generating qualitative data which thematic analysis does. The process of coding and identification of themes are closely linked as themes can only merge when data has been coded.

  
  

  Once satisfied that the themes cover majority of the data, they collect new data to test validity, writing a final support using direct quotes
• AO3 - Content analysis

  They circumnavigate many ethical issues as the text is already in public domain.

  The communication is normally of a more sensitive nature

  These can be flexible for the aims of the research

  
  

  Data is analysed outside of context and thus may be subject to researcher bias

  Lack of objectivity - especially an issue with descriptive forms of thematic analysis
• What is significance


  A statistical term that tells us how sure we are that a difference or correlation exists. A 'significant' result means that the researcher can reject the null hypothesis.
• What is a null hypothesis?

  A hypothesis which suggests no difference in conditions as the IV has no effect on the DV
• What is a type 1 error

  This is when one rejects the null hypothesis when it should be accepted. This is an optimistic error and normally occurs if the significance figure is too lenient
• What is a type 2 error

  When one accepts the null hypothesis when the alt should be accepted. This is a pessimistic error (false negative) and comes about when the significance figure is too strict
• What is the abstract section in a psychological research write up

  - The first section, typically around 150-200 words.

  - Includes all major elements such as aims, hypotheses, methods, results and conclusions
• What is the introduction section in a psychological research write up

  - A literature review on the general area of investigation

  - Details all relevant theories, concepts and studies that are related to current study

  - Should follow a logical progression
• What is the methods section in a psychological research write up

  - Should include sufficient detail for other researchers to be able to precisely replicate

  - Split into sub-sections:

  * Design (independent groups, observations) and justification

  * Sample (How many, context of them), sampling method and target population

  * Apparatus (details of instruments used)

  * Procedure (A "recipe" style list of every thing which happened including verbatim briefings, standardised instructions and debriefings)

  * Ethics and explaining how these were dealt with
• What is the results section in a psychological research write up

  - Summarises key findings from investigation

  - Likely to feature descriptive statistics (tables, graphs and measures of central tendency)

  - Inferential statistics should include reference to choice of stat test, calculated and critical values, level of significance, final outcome and which hypothesis was rejected

  - Raw data and any calculations to appear in an appendix

  - Qualitative methods of research, the findings are likely to involve theme analysis
• What is the discussion section in a psychological research write up

  - Summarisation of findings in verbal form. To be discussed in context of evidence presented in the introduction and other research that may be relevant

  - Be mindful of limitations of current study and discuss these as part of this section. May include reference to method of sample

  - Wider implications are considered. May include real life applications and what contribution the study has to existing knowledge
• How to reference in written up psychological research

  Author, date, title of book (in italics), place of publication, publisher
• AO3 - Observations


  Useful for investigating children

  Typically done in natural environment - ecological validity

  Useful for pilot study

  
  

  No explanation for observed behaviour is gained

  Observer bias

  Difficult to replicate
• Null hypothesis

  I know nothing weird is happening
• One tailed hypothesis 


  I know something weird is happening and it is because of this
• Two tailed hypothesis (non-directional)


  Something weird may happen but it can go either way