RM from t

• Objectivity

  When studying measurable phenomena scientists must approach their investigation in an unbiased way not allowing their personal opinions and biases to affect the way they collect or interpret data about ppt behaviour.
• How do you improve objectivity? (x4)

  - collecting empirical data

  - operationalisation

  - double blind trials

  - training researchers to not let opinions affect data gathered
• The empirical method


  gathering observable data through senses in order to test hypothesis
• advantages of empirical method (x2)

  1) facilitates gathering of objective data as you are less likely to rely on your own opinions when gathering data

  2) facilitates operationalisation which facilitates replicability
• Replicability

  findings must be shown to be consistent (reliable) even in different circumstances in order to trust scientific theory
• What is needed for replicability? (x4)

  - standardised procedure

  - controlled variables

  - operationalised variables

  - empirical measurement
• Why is replicability important? (x3)

  - confirms validity of findings

  - expands generalisability

  - provides confidence in findings
• Falsifiability

  Scientific theories should be testable using hypothesis and proven false, in order for this theories need to be measurable - if they are not then they can be considered a psuedo science according to Karl Popper
• What happens when falsifiability is tested?

  Theories which survive attempts to be falsified have the strongest validity, falsified theories are abandoned or refined and tested further to get a true understanding of behaviour
• What is a theory?

  a set of laws/principles that explain a particular behaviour
• What are the steps in theory construction? (x3)

  1) observe phenomena in the world/past research

  2) create theory based on what has been observed

  3) use theory to make predictions - hypothesis
• What is required in hypothesis testing? (x2)

  - operationalised DV

  - specifically describe how the IV changes
• Paradigm

  theoretical framework and set of agreed assumptions about the explanation of behaviour which is accepted by a majority of scientists in that particular field of study
• Paradigm shift


  a change in paradigm and shared set of assumptions from one way of thinking to another as more evidence accumulates over time to suggest the current paradigm is either invalid or inadequate. Scientists in a field collectively change their way of thinking
• Internal Reliability

  when all the items/questions within the method (usually questionnaire or interview) provide consistent, similar results

  
  

  NOT IN AO1: example - from q1-10 ppt answers shows symptoms of OCD
• How do you test internal reliability?

  Split Half Method

  
  

  1) Once ppts have answered all questions split all their responses into 2 halves

  2) conduct a correlation analysis where 1 variable is the total score from one half and the 2nd variable is the total score from the other half

  3)if the correlation analysis shows a strong positive correlation (co-efficient ≥ 0.8) we can conclude there is internal reliability
• External reliability

  when ppts produce similar results over time

  
  

  NOT IN AO1: example - ppts provide same results in experiment 1 and a replication of experiment 1 (experiment 2)
• How do you measure external reliability?

  Test Re-Test Method

  
  

  1) test ppts in an experiment

  2) test same ppts again using the same standardised procedure

  3) conduct a correlation analysis comparing each ppt's data from the first test to the second test

  4) if there is a strong positive correlation (co-efficient ≥ 0.8) we can conclude that there is external reliability
• Inter-observer reliability

  When two observers using the same behaviour category sheet to observe the same participants for the same duration at the same time produce consistent results
• How do you measure inter-observer reliability?

  1) two observers record data independently

  2) conduct correlation analysis where one variable is the total result (tallies) from one observer and the second variable is the total score from the other observer

  3) strong positive correlation (co-efficient ≥ 0.8) we can conclude that there is inter observer reliability
• How do you improve reliability in self report measures? (x4)

  - avoid asking ambiguous questions to avoid different interpretations of the question which would lead to varying responses

  - use structured interviews as it has a standardised procedure allowing for replicability

  - use closed questions as it allows for objective analysis of responses

  - same neutral tone in all questions during an interview
• how do you improve reliability in lab experiments? (x5)

  - use a standardised procedure to allow for replicability which allows for reliability

  - control conditions so you can recreate the setting

  - control of EVs so they do not occur in one condition and not another condition

  - operationalisation of variables

  - use pilot study to find EVs and cut them out
• How do you improve reliability in observations? (x4)

  - operationalised behavioural categories and ensuring categories don't overlap

  - define behaviour specifically in behaviour so observers do not have different interpretations of the same behaviour

  - control of EVs

  - training observers to observe w/o missing out things or letting opinions affect findings
• internal validity

  the extent to which the method/research tool is measuring what it intends to measure
• what is internal validity in an experiment?

  being confident that changes in DV is due to changes in IV and not due to EV

  
  

  (eliminating EV provides strong cause and effect relationship making internal validity stronger)
• How can you use face validity to assess internal validity?

  assess at first glance whether a research method measures what it intends to measure

  
  

  take research method to the expert in the field and ask them to rate from a scale of 1-5 how well it measures what it intends to measure. High rating shows face validity
• How can concurrent validity be used to assess internal validity?

  1) ppts complete newly established method

  2) the same ppts complete already established method that measures the same behaviour

  3) complete a statistical analysis - correlation analysis: where one variable is the total score from the newly established method and the second variable is the total score from the already established method.

  
  

  - strong positive correlation (co-efficient ≥ 0.8) we can conclude that there is internal validity
• How to improve internal validity (x9)

  1) Pilot study - find and control for EVs which may affect DV

  2) High levels of control - strong cause and effect relationship

  3) Ensure ppts do not know the aim of the study - prevents them showing demand characteristics

  4) Double blind trials

  5) Random allocation

  6) Counter balancing

  7) Train researchers to treat ppts equally/use a neutral tone and have an unbiased interpretation of behaviour

  8) Standardised procedure - no ppts have an unfair advantage

  9) Matched pairs design
• External validity

  the extent to which findings can be generalised beyond a study
• What are the three types of external validity

  population, ecological, temporal
• How do you measure population validity

  do experiment with ppts who are different from the original sample, but still lie within the target population
• How do you measure ecological validity?

  do experiment in a real life setting and see if you can get similar results
• How do you measure temporal validity?

  repeat experiment from the past and see if you can get similar results
• how do you improve ecological validity?

  conduct experiment in real life setting with the experimental task representative of something that would happen in real life
• how do you improve population validity?

  use a sample which is representative of the target population
• Case studies

  An in depth investigation based on a specific individual or a small group of people
• Strengths of case studies (x2)

  - provides in depth data on ppts & gives a deeper insight on causes of ppt behaviour INCREASING VALIDITY

  - allows investigation of rare phenomena that you cannot ethically induce in ppts
• Weaknesses of case studies (x3)

  - based on individual/small group with unique characteristics or situations, cannot generalise findings to target population or everyday circumstances

  - studies based on one-off cases - difficult to replicate the investigation as the characteristics cannot be controlled DIFFICULT TO TEST RELIABILITY

  - researcher may form bond with object of study as research often takes place over a long duration, causes bias in the interpretation of behaviour - REDUCES OBJECTIVITY & UNDERMINES THE VALIDITY
• Qualitative data


  non-numerical, descriptive data
• Strength & weakness of qualitative data

  - in-depth, more detail about ppt behaviour and reasons for their behaviour

  - difficult to conduct statistical analysis - relies on the subjective analysis of researchers