Research Methods 2

correlational research

statistical relationship between 2 quantitative variables

strength and direction

what relationship does correlational research have?

pearson’s r

statistic that quantifies the strength of a correlation between quantitative variables

linear

what kind of relationships is pearson’s r good for?

-1 to +1

what is pearson’s r range?

positive correlation

as one variable increases, the other variable increases

negative correlation

as one variable increases, the other variable decreases

-1 or +1

what is strong correlation closest to?

correlational research

in what kind of research is there no manipulation of the independent variable?

no

can we make causal conclusions from correlational research?

causal directionality and confounds

what two aspects are not proven when a correlational relationship is shown?

external validity

the extent to which results of a study are generalizable to a population

internal validity

the extent to which the study establishes a causal relationship as opposed to a relationship due to confounding variables

inaccurate correlational coefficients

what does a restriction of variable range result in?

spurious correlation

caused by the presence of a confounding variable; occurs when two variables appear to be related but are not

observational research

research where behavior is watched and recorded; nothing is manipulated

naturalistic observation

researcher is passive; takes place in a natural setting where behavior normally occurs

structured observation

researcher observed specific behaviors in a controlled setting

observer bias

observers see what they are expecting to see

observer effects

participants behave differently because they are aware they are being observed

well-trained observers, measure inter-rater reliability, used blind research designs

how can we prevent observer bias?

unobtrusive observations or habituation

how do we reduce reactivity/observer effects?

habituation

allows participants to get used to being observed

participant observation

researchers become active participants in the group or situation theyt are studying

changed social dynamic; influencing the group; less objective researcher

what are the drawbacks to participant observation?

structured observation

researcher makes observations in a setting that is less natural than real-life; quantitative data is gathered

quantitative

what kind of data is gathered in structured observation?

coding

clearly defining a set of target behaviors

case study

in-depth examination of an individual; typically more qualitative in nature; useful when studying rare conditions

people like to take shortcuts (cognitive misers) and will be truthful

what are the two main assumptions in survey research?

open-ended question

respondent can answer however they want

closed-ended question

respondent chooses from a limited set of answers

simple questions. know audience, be specific, void problematic answers, consider response sets

what are the guidelines for writing surveys? (5)

double-barreled question

asks about two or more ideas in one question

leading question

question that sways people to respond in a certain way

loaded question

assumes something about the respondent

negative wording

survey question that is difficult to interpret

response sets

respondents respond a certain way because of social desirability or have extreme responses

number of participants / number of people selected

how do you calculate response rate?

more than 70%

what is an ideal response rate?

50%

what is the acceptance rate for validity?

determine population, conduct first draft, pre-test survey, administer, and analyze

what are the steps in survey research? (5)

how well sample is selected + size

what determines sample accuracy?

population

entire group or type of people you are studying

sample

a smaller part or subset of a population

census

data collection that involves every member of the population of interest

survey

data collection that invovles a subset of the population

probability sampling

each person in the population has an equal chance of being selected

simple random sampling

everyone has an equal chance of being selected

stratified sampling

population is broken into subgroups and then sampled from those groups

probability

what kind of sampling is simple random sampling?

probability

what kind of sampling is stratified sampling?

non-probability sampling

the chance of someone being selected is unknown

self-selected sampling

people actively seek out participation in research; biased

convenience sampling

recruitment of those who are easily accessible; biased

purposive sampling

recruitment of those who fit specific criteria and intentionally seek them out; targets unique populations

snowball sampling

current participants recruit others to participate in the study; targets hard-to-reach populations

probability sampling

what’s the idea sample method?

experiment

designed to see whether there is a causal relationship between two variables

IV is manipulated and others are controlled for

what are the 2 fundamental factors of an experiment?

control variables

variables that experimenter is holding constant between groups

control groups

group that does not get the IV or gets some standard value of the IV

random assignment

control technique to equate groups of participants; ensures every person has an equal chance of being assigned to every group

yes

does random assignment eliminate systemic differences?

no

does random assignment eliminate the influence of extraneous variables?

one-way design

one variable is manipulated; requires 2 levels of the IV to compare effects on the DV

between-subjects design

different people are compared to each other; each participant is only in one experimental condition

post-test only design

participants undergo the manipulation and then complete measures

pre-test post-test design

participants complete measures, then manipulation, then measures again

no selection bias, less noise

what are the advantages to pre-test post-test?

within-subjects design

the same participants are compared to themselves; all participants experience all conditions

no selection bias, no noise, doubled sample size

what are the benefits of within-subjects designs? (3)

order effects

what are the drawbacks of within-subjects designs? (1)

order effects

occurs when experiencing one condition changes how participants react to another

fatigue effect

participants get worse at measures; boredom, distraction, etc.

practice effect

participants get better at measures; learning

carryover effect

effects of one condition contaminate later responses

counterbalancing

presenting the levels of the IV to participants in different sequences to control for order effects

counterbalancing

how can we prevent order effects?

demand characteristics

features of the study that influence participants to become aware of the purpose of the study

general description, deception, filler questions, asking afterwards

how can we void demand characteristics?

straightforward manipulations

manipulating the IV w/ written instructions, verbal materials, and/or visual materials

staged manipulations

manipulating an entire setting; involved confederates

confederates

people who appear to be participants or bystanders but are part of the experiment

ethics, realism, and resources

what should researchers consider when conducting an experiment?

manipulation checks

measures used to determine whether the IV manipulation was effective

floor effect

scores are clustered at the low end of the measure

ceiling effect

scores are clustered at the high end of a measure

debriefing

explaining the purpose of a study

causal diagrams

visual representations of relationships between variables

mediator

explains how the main cause influences the main effect

common causes

make it appear like there is correlation when really there is not

yes

should common causes be controlled for?

noise variables

alternative causes of an apparent relationship that researchers do not know about

temporal precedence

the cause must occur before the effect

statistical relationship, temporal precedence, and controlled confounds

what are the 3 requirements to establish a causal link?

yes

can both experiments and observational research establish statistical relationships?

single-blind

either the participant or the researcher does not know the experimental condition

double-blind

neither the participant nor observer knows the experimental condition

selection bias

participants in each group may have pre-existing differences