psych analysis (8) - sampling for canvas

external validity

extent to which our study generalizes to a population

potential sampling strategies

• emailing

• purchasing

• posters

• survey

two types of sampling

• probability sampling - everyone in population has an equal chance of being selected

• non probability - some people in population are left out/less likely to be selected

random selection vs random assignment

Random selection is randomly choosing participants from a population to participate in a study, while random assignment is randomly placing participants into different groups within a study.

probability sampling

everyone in population has an equal chance of being selected

non probability sampling

some people in population are left out/less likely to be selected

simple random sampling (probability sampling)

every member of a target population has an equal, independent, and non-zero chance of being selected for a psychological study

systematic random sampling (probability sampling)

• start at x point on the list

• select every nth element

cluster random sampling (probability sampling)

divide a large population into smaller, naturally occurring groups (clusters) (classrrooms, city blocks, etc), then randomly select clusters, collect data

multistage random sampling (probability sampling)

• divide into clusters

• break down into even smaller groups

• pick participants from the most specific groups

stratified random sampling (probability sampling)

• target population is split into mutually exclusive subgroups (age, gender, etc) then random selection

• direct proportional stratified sampling - mutually exclusive subgroups, random samples drawn in proportion to their size in total population

• disproportional stratified sampling - mutually exclusive subgroups, random sampling NOT drawn in proportion to their size in total population

strata

demographic groups

direct proportional stratified sampling (stratified sampling) (probability sampling)

mutually exclusive subgroups, random samples drawn in proportion to their size in total population

disproportional stratified sampling (stratified sampling) (probability sampling)

mutually exclusive subgroups, random sampling NOT drawn in proportion to their size in total population

convenience sampling (non probability sampling)

researches select participants because they’re easy to reach, available, and willing to take part

snowball sampling (non-probability sampling)

use social networking

purposive sampling (non-probability sampling)

researchers purposely select participants based on specific traits, experiences or knowledge relevant to a study’s objective

quota sampling (non probability)

divide population into mutually exclusive subgroups, select participants non-randomly until you have enough people (non random stratified sampling)

self selection

individuals independently choose to participate in a study, and the traits driving that choice correlate with the study’s focus

• ethical implications

• bias

• remedies

  • incentives

  • reduce stress

  • convinience

  • appeals