general evaluation limitations

correlational

• fundamental weakness is cannot establish causal relationship between the co variables

• as there is no deliberate manipulation of an IV, the researchers cannot establish the direction of [topic]

• relationship may also be purely coincidental due to the third variable problem where an unmeasured, intervening variable is responsible for the change in both co-variables like [topic]

• means findings lack internal validity making the research less useful for [topic]

self report

• undermined by response bias threatening internal validity

• rely on participants’ honesty making it susceptible to social desirability bias to make answers look ‘good’ or ‘normal’ rather than truthful

• may fall victim of acquiescence bias or the interviewer effect

• data may be a reflection of how they wish to be perceived rather than reflection of true behaviour so lacks scientific accuracy

retrospective study

• reliance on retrospective recall is prone to inaccuracy and distortion

• human memory is a reconstructive process - memories can decay, be interfered or participants may engage in effort to alter meaning

• issue for internal validity leading to false conclusions

lab studies

• low ecological validity and mundane realism

• strictly controlled and artificial environments dont reflect complexities of real life settings, and tasks are often contrived

• can trigger demand characteristics

• findings struggle to generalise beyond lab setting limiting real world applicability

animal study - ethics

• speciesism argument by singer

• suggests inflicting suffering on animals is a form of discrimination, especially when findings may not generalise to humans due to significant biological and cognitive differences

• research fails to meet a positive cost benefit ratio, leading to a call for more human centric alternatives

• researchers must adhere to the BPS 3Rs framework - replace animals with computer models where possible, reduce no. animals used to minimum, refine procedures so least suffering

animal study - generalisability

• issue of extrapolation

• profound qualitative differences in brain structure, cognitive processing and social complexity

• humans have more developed prefrontal cortex, complex language abilities and more heavily influenced by conscious intentions and cultural norms - factors absent in other species

• lack population validity when applied to humans limiting usefulness by providing a misleading or incomplete picture of the multifaceted nature of humans

case studies

• unrepresentative - focus on one individual or experience meaning the findings are hard to generalise to the wider ‘normal’ population

• researcher bias - LT clinical relationships can lead to the researcher losing objectivity, potentially interpreting data to fit their own theory

longitudinal studies

• attrition rates - participants often drop out over time and those staying until the end might be a specific type of person (ie more motivated) which biases the final results

• cohort effects - results might only apply to that specific generation or group of people because of the unique historical time they grew up in

meta-analysis

• file drawer problem, publication bias - if it only looks at published studies, the conclusion may be biased

• method consistency - studies may use slightly different procedures so the data may not be standardised

lacks generalisability final sentence

caution is needed when applying the findings to the general population, limiting external validity

determinism/reductionism final sentence

…… underestimates the complexity of human behaviour by failing to acknowledge the role of …… in determining how we ……

lab studies final sentence

they dont capture the complexity of every day behaviour as ……. may behave/be different in natural settings