research methods midterm

why conduct a literature search before designing a study?

To learn what is already known and refine the research question/hypothesis

what is the difference between a resource and source?

a resource helps you find research, a source is the actual research itself

primary source

A source reporting original research conducted by the authors

secondary source

a source that summarizes or discusses research conducted by OTHERS

what is the purpose of review paper

to summarize the current state of research in an area

what is the purpose of a theory paper?

to propose a theory that explains existing findings

what are the 5 sources of research questions?

intuition, everyday life, researcher, discussions, real world, problems, existing research

what does it mean that science is cumulative

new research builds on existing research while making a new contribution

what is a hypothesis

SPECIFIC and FALSIFIABLE prediction about the relationship between variables

what does falsifiable mean?

there are possible results that do not support the prediction

hypothesis vs theory

hypothesis = specific prediction

theory = broad explanation that generates predictions

(THEORY generates multiple HYPOTHESES)

research question vs hypothesis

research question = asks what will happen in a relationship

hypothesis = makes a specific prediction

theory vs law

theory explains/predicts many relationships

law is assumed to apply to ALL situations

conceptual vs empirical variables

conceptual = abstract construct

empirical = how its actually measured

what is an operational definition

the empirical implementation of a conceptual variable

(how a researcher turns an abstract concept into something that can actually be manipulated or measured)

in an experiment how are the IV and DV treated?

IV = manipulated

DV = measured

what is a nominal measure

categories with no meaningful order, NAMES

(categories)

what is an ordinal measure

ORDERED values in which the distances between values aren’t necessarily equal

( categories + order )

what is an interval measure

equal INTERVALS between values but no meaningful zero.

(categories+ order + equal intervals )

what is a ratio measure?

equal intervals and a meaningful zero point

( categories + order + interval + REAL zero)

experimental IV manipulations are usually measured at what level?

nominal/categorical

what are the 4 major types of measures

self report, performance, behavioral , psychophysiological

what is reactivity

when being studied/measured influences participants responses or behavior

what is social desirability

responding in ways that make oneself appear more socially acceptable

what is acquiescent responding?

The tendency to agree with statements, regardless of their content (yeah-saying)

what is the good participant effect and #&*% YOU effect ?

good participant = changing behavior/responses to match what the participant thinks the researcher expects

#&*% YOU ! = the opposite, a participant may intentionally resist what they think the researcher wants rather than cooperating with the perceived purpose of the study

what is reliability?

reliability = CONSISTENCY of a measure, free from random error

what is validity?

whether a measure actually measures what it claims to measure, free from systematic error

how does random error affect reliability?

higher random error = lower reliability

lower random error = higher reliability

random error vs systematic error ?

random error → reduces RELIABILITY

systematic error → reduces VALIDITY

reliability vs validity

reliability = Consistency

Validity = measures what it claims to

item vs scale

item = individual question/measure

scale = multiple items (questions) measuring the same conceptual variable, usually combined into one score

what is test retest reliability?

The consistency of scores on the same measure across two different times

( same measure, different times)

how does test-retest reliability differ for traits vs states?

traits - generally show high test-retest reliability ( traits remain stable, don’t usually change)

states - may have low test-retest reliability because they naturally change over time.

what is a retesting effect?

taking the same measure previously influences response responses/performance when taking it again

what is equivalent-forms reliability?

The consistency between scores on two different but equivalent versions of a measure

(different equivalent versions, different times)

why use equivalent forms instead of test retest?

to reduce retesting effects

what is internal consistency?

The extent to which items within a multi scale consistently measure the same conceptual variable ( do items within one scale correlate with one another )

what is split half reliability?

correlation between scores on two halves of the same scale; measures internal consistency

what does Cronbach’s alpha measure?

internal consistency; how strongly items within a scale correlate with one another ( measure the same construct)

what is interrater reliability?

The degree to which multiple coders/observers agree in their judgments

( do different observers/ coders agree?)

why should cutters be blind to the hypothesis and experimental condition?

to prevent their expectations from biasing their judgments

why should cutters be blind to the hypothesis and experimental condition?

to prevent their expectations from biasing their judgments

why are surveys better at telling us ‘what’ than ‘why’?

people may not accurately know what causes their own dots or behavior

population vs sample

sample = the people you studied/ who actually participated

population = the larger group you actually want to learn about and generalize your findings too

parameter vs statistic

Parameter - describes Population

Statistic - describes Sample

what is convenience sampling?

recruiting participants because they’re easy or readily available

what is snowball sampling?

using existing participants to help recruit additional participants

what is non-probability sampling?

sampling where population members don’t have a known chance of being selected

frequency distribution

shows how frequently different values occur in a dataset

central tendency vs dispersion/variance

central tendency = center (mean,mode,median)

dispersion = spread ( range, standard deviation)

mean vs median vs mode vs range

mean = average

median = middle

mode = most common

range = max - min

positive vs negative skew

positive = tail right

negative = tail left

confidence interval

range used to estimate a population parameter from sample data

how does sample size affect confidence intervals ?

larger sample → narrower confidence interval → more precise estimate

naturalistic observation vs participant observation

naturalistic = observe without participating

participant = researcher joins group

why define behavioral categories before observation?

to operationalize behavior so they can be observed and recorded consistently

what is the main limitation of case studies?

findings may not generalize to the larger population

archival research vs content analysis

Archival = Analyzes existing records/data

Content analysis = Codes existing material into categories

what are unobtrusive measures?

measures behavior without participants knowing, reducing reactivity

main strength vs limitation of naturalistic methods?

STRENGTH- high external validity/real world behavior

LIMITATION- less control, cannot establish causation

naturalistic vs structured observation

naturalistic- observes behavior and it’s normal setting

structured - creates a situation, then observes behavior

reactivity vs observer bias

reactivity- participants change because they are being observed

observer bias- researcher expectations affect observations

time sampling vs event sampling

time - observe during set time intervals

Event - record Each occurrence of a target behavior

null vs alternative hypothesis

null (H0) - no effect/relationship

alternative (H1) - an effect/relationship does exist

p value, and how is it interpreted

p value - the probability of getting the observed results if (H0) the null is true

if the p is LOW the null must GO ( p<a) reject null (H0), results = SIGNIFICANT

if the p is HIGH, the null will FLY ( p>a) fail to reject null (H0), results NOT significant

type 1 vs type 2 error

type 1 = false positive ( find an effect that isn’t real)

type 2 = false negative ( miss a real effect)

statistical significance vs effect size

significance - evidence an effect exists

effect size - how large/strong the effect is

what is statistical power ?

ability to detect a real effect

higher power → Fewer type two errors

( larger samples increase power)

what is alpha (a)?

cutoff for significance, usually .05

lower a = stricter cutoff + lower type 1 error risk

one tailed vs two tailed hypothesis

one tailed - predicts direction of effect

two tailed - predicts an effect, but not its direction

what does a confidence interval tell us?

A range of values used to estimate the true value in the population from the sample data

smaller sample → wider CI= less precision

larger sample → narrower CI=more precision

why is replication important?

repeating a study test, whether it’s findings are reliable and can be reproduced

publication bias

significant/positive findings are more likely to be published than null findings, potentially distorting the evidence

replication vs Meta-analysis

replication - repeat a study to see if the findings reproduce

Meta-analysis- statistically combines results from many studies

positive vs negative correlation

positive - variables move in the same direction

negative - variables move in opposite directions

the closer ( r ) is to -1 OR 1 the stronger the correlation

closer to 0 = weaker

directionality vs Third variable problem

directionality (reverse causation problem) = unclear which variable causes which

third variable = another variable may cause both

what is the coefficient of determination? (r²)

shows how much the differences in one variable are associated with differences in the other (Proportion of shared variance)

what is restriction of range?

when you only obtain a limited range of scores, so the correlation may appear weaker than it really is

what is a curvilinear relationship?

One variable changes with another up to a point, then the pattern changes direction

( ex increasees, peaks, drops off)

how can outliers affect a correlation?

Extreme scores can distort r, making the relationship appear stronger or weaker

predictor vs Criterion variable

predictor- Used to make a prediction

Criterion - Outcome being predicted

what is multiple regression?

uses multiple predictor variables together to predict one outcome

cross-sectional vs longitudinal research

cross-sectional- Different groups at one time

Longitudinal- Same people followed overtime

what is a spurious relationship?

when two variables seem related, but their relationship is actually explained by a third variable

(The relationship looks real, but something else explains it)

what is self selection in correlational research?

participants already differ on the predictor variable before the study so researchers cannot fully control those differences

self selection → makes it difficult to establish causation