Psychology Ethics: Psychology as a Science

What is studied and the method of research used.

IS - Lab experiments, quantitative data, brain scans, are objective with limited bias and are measurable
IS NOT - Case studies, field experiments, observations provide qualitative data that is more susceptible to bias and harder to replicate

A model, theory, or area which follows the same principles.

IS - Psychologists have the same goals of explaining and predicting human behaviour
IS NOT - There is a lack of paradigm in explaining behaviour in different areas of psychology (e.g. Biological states aggression can come from evolution or an imbalance of hormones)

All sources of bias that can affect data are minimised with subjective ideas being eliminated.

IS - Quantitative methods operationalise concepts in order to measure them and methods such as inter-rater observations minimise bias
IS NOT - Humans studying humans can never be objective (e.g. observations of behaviour can be subjective to individual)

A hypothesis should able proved to be false not true.

IS - Approaches that are more scientific (e.g. Bio) have methods that are easier to falsify
IS NOT - Easy to find supporting studies and the psychodynamic approach cannot be falsified

The ability to accurately reproduce something correctly, such as the procedure of a study in order to compare to test for consistency.

IS - Lab experiments use high control and standardised procedures that can be easily replicated (e.g. Baddeley 1966)
IS NOT - Some methods are more complex that cannot be standardised (e.g. unstructured interviews in clinical)

A variable that is kept the same throughout the experiment.

IS - Lab experiments have high control due to standardised procedures formed from testing variables
IS NOT - Methods such as observations have EVs that aren’t always identifiable and controllable, control limits realism, correlations have no control, difficult to operationalise abstract concepts

A hypothesis that is formed based on an assumption of what will happen that is tested to get disproved.

IS - Theories can lead to real life predictions (e.g. Behaviourism with conditioning techniques)
IS NOT - Too general for precision (e.g. Schizophrenia has many predictions such as dopamine hypothesis and social causation)

Applying a behaviour to a wider population.

IS - Random sampling methods avoid bias (e.g. stratified)

IS NOT - Pre 1970s studies were often white middle class men from USA (e.g. Milgram's 1963) with current ones often being volunteer students (e.g. Loftus and Palmer 1974)

IS - Evolution of species led to similar traits with rats having similar CNS to humans
IS NOT - Evolutionary discontinuity for animals due to humans being further along with different environments

Accurately measuring what is sought to be measured with research.

IS - Lab experiments are more objective and free from bias with the ability to establish cause and effect (e.g. Meltzer’s 2014 use of placebos)
IS NOT - DCs can cause bias as well as other forms of bias such as researcher

Simplifying something to certain aspects of it that is used to explain the whole.

Looking at something as a whole to explain something.

IS - Reductionism makes concepts measurable

IS NOT - Holism is less measurable and less scientific

• Lab experiments (e.g. Milgram 1963) have high control for cause and effect

• Abstract concepts such as prejudice can be operationalised to become reductionist and measurable like friendship levels (e.g. Sherif 1954)

• Questionnaires and surveys can be replicated easily and use quantitative objective

• 2 researchers can can agree on themes for thematic analysis

• Field experiments have low control over extraneous variables so cannot establish cause and effect (e.g. Sherif 1954)

• Questionnaires are open to bias from social desirability, acquiescence bias, and researcher bias from interpreting open questions

• Humans studying humans can never be objective

• Lab experiments (e.g. Baddeley 1966) use standardised procedures that can be replicated easily

• Reductionism of memory is scientific (e.g. Baddeley 1966 with word lists)

• Memory is universal and so has a unified paradigm

• Brain scans for memory impairments being due to brain impairment

• Evidence of a hypothetico-deductive model in Baddeley 1966 who adapted his experiment to separate the STM and LTM encoding

• There isn't a unified paradigm due to no definitive test for memory

• Case studies (e.g. KF suffered memory impairment from motorcycle accident and had impaired verbal recall as opposed to visual)

• Bartlett's research created his theory which is not a hypothetico-deductive model

• Brain scans (e.g. Raine 1997 using a PET scan) are scientific

• Can generalise animal studies on the CNS of rats, etc to humans (Wagner 1979 with castrated rats)

• Lab experiments have high control for cause and effect and standardised procedures for replication

• Reductionist when looking at areas of the brain being responsible for certain behaviours

• Correlations have no control over variables and cannot establish cause and effect

• Case studies (e.g. Charles Whitman) have complex procedures involving different methods

• No unified paradigm as there are multiple explanations for aggression

• Twin studies never have 100% concordance rate

• Evolutionary discontinuity means can't generalise animal research on aggression as different social environments

• Lab experiments (e.g. Watson and Rayner 1920) have standardised procedures to allow high replication

• Unified paradigm on seeking to find how behaviours are learnt with different theories

• High predictions from learning theories for the development of things like anorexia

• Animal research on behaviour (e.g. Thorndike 1898 and his cats)

• Inter-rater reliability with methods like observations (e.g. Bandura 1961)

• Observations (e.g. Bandura 1961) can be subjective when interpreting behaviour as humans studying humans can never be objective

• Field studies have low control over EVs meaning lower validity with cause and effect of measuring what is sought to be measured

• Animal studies on behaviour are not fully generalisable due to qualitative differences in the brain as a result of evolutionary discontinuity

• Behaviours can be hard to operationalise

• Uses biological explanations to explain criminal behaviour which use scientific methods such as brain scans (e.g. brain structure)

• Lab experiments have high control (e.g. Loftus and Palmer to test EWT) and are standardised

• Social explanations for criminality are reductionist (e.g. labelling theory and SFP and so easily measurable

• Case studies (e.g. psychological case formulation) have complex procedures that are harder to replicate

• Case studies about unique criminals are harder to generalise to other people due to their uniqueness

• Field experiments in EWT have lower control over variables that can affect what is being measured

• CBT is more holistic in its approach by combining biological and social factors for behaviour and so less sicentific

• DSM and ICD are objective with the standardised manuals to follow when diagnosing

• Brain scans are used to compare the brains of patients with mental illnesses

• Animal research is often done to test the effects of drugs on rats with symptoms of mental disorders which can be generalised to humans

• Explanations for disorders are reductionist

• Accurate predictions can be made from learning theories, etc on the development of a disorder (e.g. Nasser 1986 Cairo vs London for Egyptian people)

• Diagnostic interviews are unstructured and gain qualitative data which is harder to replicate and measure

• Subjective interpretation of symptoms

• Interviewer bias on what the patients disclose

• Case studies (e.g. Luk and Agoha 2014) are harder to replicate due to their unique and complex nature with a more holistic approach

• Correlational data doesn't establish cause and effect (e.g. social causation hypothesis)

• Carlsson 2012 used secondary data so can't be sure all methods were standardised or free from bias