Selecting participants and research strategies – Research Methods

Target population vs. Accessible population

Target population: the specific group the researcher wants to understand

Accessible population: the portion of the target population that the researcher can actually reach

what’s selection/sampling bias?

this occurs when the sampling procedure favours certain individuals over others

ex: studying adults but only recruiting uni students

what is attrition?

the loss of participants during a study

• attrition can Mae a sample biased if certain types of participants are more likely to drop out

• researchers must report how many participants started and finished the study

What are the two main categories of sampling?

• Probability sampling

  • participants are chosen by chance, using a random role

• non-probability sampling

  • chosen by choice or convenience

what are the 3 conditions of probability sampling that must be met?

1. it is possible to identify every individual in the population

2. each individual in the population has a known, specific probability of being selected

3. selection must occur using a random, unbiased process so that every member of the population has an equal chance of being selected

what are the 5 different types of probability sampling?

• simple random sampling

• systematic sampling

• stratified random sampling

• proportionate stratified random sampling

• cluster sampling

what is simple random sampling?

a probability sampling method in which every individual in the population has an equal chance of being selected

• equal probability of selection

• selection is based on chance, not researcher choice

• minimizes selection bias

• in the long run seems to produce representative samples

however Random ≠ guaranteed representative in the short run 

sampling with vs without replacement

WITH: after you pick someone for your study you then put them back into the population which makes propability math easier

WITHOUT: you dont put them back into the populations so there are no duplicates

what is systematic sampling?

a random starting point from the population is chosen and then every nth individual is selected

• n = a fixed number

  • you select every n-th person

    • ex: every 5th person

What is stratified random sampling? and what is proportionate stratified random sampling?

the population is divided into specific sub-groups (strata), and random samples are taken from each sub-group because they matter to the study

• sample sizes can be different or equal and not necessarily proportional

• ensure all important subgroups are represented

proportionate stratified random sampling: a type of sampling where the proportions of subgroups match the size of the population

What is cluster sampling?

you divide the population into pre-existing clusters (naturally occurring groups), then randomly select the entire clusters, and include all members of the selected clusters

what are the two types of non-probability sampling?

convince sampling: selecting some of the most easily obtained individuals from the population to participate in a particular study

• based on availiibilty and willingness to participate

• no random selection

quota sampling: a convenience sample that includes a pre-determined number of participants from specific sub-groups or categories within the population and you pick those who are most available and willing

• quota is a set number of people you need from a specific subgroup

what is a research strategy?

the general approach you take to answer your research question

• it depends on the type of question (description, relationship, cause) and answer you want (correlation, average, explanation)

What is the descriptive research strategy?

its focus is to describe individual variables and is NOT concerned with the relationships between the variables

what is correlational research?

measures two variables and tries to determine if there is a relationship

• numerical scores for each

what is the experimental research strategy?

goal is to find out the cause and effect, it uses control and random assignment to rule out their explanations

what is a quasi-experimental strategy?

tries to measure cause and effect bur cannot be fully controlled

• there is no random assignment and the groups already exist

what is the non-experimental research strategy?

its goal is to show a relationship exists but not the cause

• there is no manipulation of variables

• no random assignment

non-experimental vs. correlational

both show relationships but not causes

correlational: sees if the two variables are related

nonexperiemental: sees if one variable has a relationship with the group

research design vs research procedures

research design: is the general plan for the study

research procedure: is the step by step instructions for the study

what are the 5 ways external validity can be threatened?

1. selection bias

2. volunteer bias: students who volunteer may be more motivated

3. novelty effects: behaving differently because its a new thing

4. multiple-treatment interference: prior conditions affect performance (fatigue)

5. experimenter characteristics: different exoeriementers may influence results

extraneous variables vs confounding variables

extraneous: variable in the study other than the independent or dependant variables that can influence the results

confounding: type of extraneous variable that systematically varies wth the independent variable and can provide an alternative explanation for the results

• threatens internal validity