Sampling, credibility, bias and generalizability are vital aspects in assessing the quality of research studies in psychology. Understand these principles and apply them to achieve higher grades and evaluation marks in your IB exam papers.
Below is a summary/ TLDR of textbook pg 7-17.
Refer to definitions as needed.
| SAMPLING | Sampling is the process of recruiting individuals for the study. Types of sampling: Quantitative: Random, Stratified, Self-selected, Opportunity Qualitative: Snowball, Convenience, Quota *Sampling methods are chosen based on efficiency/ time/ cost |
| CREDIBILITY | Credibility refers to how trustworthy findings are, for example, do results reflect reality? For example, in Draganski et al. (2004) experiment on whether learning processes can affect the amount of grey matter found in the brain, he compared brain scans on an experimental group which learnt juggling, and a control group. He found that learning (IV) had in fact affected the amount of grey matter (DV) evident in the experimental group. However environmental influences may impact accuracy of results, ie. some participants may have exposure to learning new things at work, hence increasing grey matter, that had nothing to do with the experiment. In quantitative studies, credibility can be strengthened by controlling confounding variables, keeping constant in all conditions. In qualitative studies, credibility can be strengthened with triangulation, ie. relying on multiple sources to build a richer context; establishing rapport with participants to gain their trust and allow them to open up; iterative questioning that could mean repeated questions whether in same or different phrasing, to test consistency and lastly, to encourage participants to provide rich descriptions of their experiences. |
| BIAS | A bias is a tendency, inclination, or prejudice toward something, someone. This can threaten internal validity. For quantitative studies, threats to internal validity include selection; testing effects; regression to the mean; experimenter bias and demand characteristics. For qualitative studies, threats to internal validity include participant bias; dominant respondent and social desirability, which is the desire to appear good to others, ie. misreporting smoking frequency to look good to others. |
| GENERALIZABILITY | Generalizability refers to whether results can be applied to a wider context/ population? If yes, that’s high external validity If results can be generalized to wider population that’s high population validity, usually if population size is wide enough and encapsulates all races, cultures, genders, ages. If study proximate real-world settings, that’s high ecological validity. For example, memory studies are oft criticized for lacking eco validity as they often take place in laboratory settings and because the stimuli used, ie. recalling trigrams, random number sequence are often artificial. If research measures accurately assesses what it’s supposed to, it has high construct validity. |
Note, generalizability is usually an easy score as an evaluative point for for IB psychology essay, ie. if participants sampled were all of the same race, results are ethnocentric, hence not generalizable to the wider population.
And if research only saw to males being sampled, findings are androcentric.
Detailed Notes on Research Methods